Sabine Bahn

Sabine Bahn is a practicing psychiatrist and a professor of Neurotechnology at the University of Cambridge in the UK. She is Head of the Cambridge Center for Neuropsychiatric Research and co-founder of the company Psyomics.

Favorite metabolite
Ceramide

Learn more about Sabine Bahn here.
Cambridge Center for Neuropsychiatric Research | https://ccnr.ceb.cam.ac.uk/
Discover Psyomics here.

Paper discussed in this episode
Metabolomic Biomarker Signatures for Bipolar and Unipolar Depression
Jakub Tomasik, Scott J. Harrison, Nitin Rustogi, et al. JAMA Psychiatry. 2024;81(1):101-106. | doi:10.1001/jamapsychiatry.2023.4096

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alice’s first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Robert Nagourney

Robert A. Nagourney, MD, is Medical and Laboratory Director of The Nagourney Cancer Institute in Long Beach, California, Associate Clinical Professor, at University of California Irvine and founder and CEO of Metabolomycs, Inc.  He is board certified in internal medicine, medical oncology, and hematology.

Favorite metabolite
Glutamine

Learn more about his work
Nagourney Cancer Institute | https://www.nagourneycancerinstitute.com/

Metabolomycs | https://www.metabolomycs.com/

TEDx talk | https://youtu.be/mAGhNhrHMJs?feature=shared

Paper we discuss in this episode
Ovarian cancer | Platinum resistance in gynecologic malignancies: Response, disease free and overall survival are predicted by biochemical signature: A metabolomic analysis

Pancreatic cancer | Diagnostic and Prognostic Performance of Metabolic Signatures in Pancreatic Ductal Adenocarcinoma: The Clinical Application of Quantitative NextGen Mass Spectrometry

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Haley Chatelaine

Haley Chatelaine s a post-doctoral fellow at the NIH’s National Center for Advancing Translational Sciences (NCATS, USA)

Favorite metabolite
Cytosine – but also Vitamin D

Paper we discussed in this episode
Paper on the absorption of fat soluble vitamins in patients with metabolic syndrome | Vitamin A and D Absorption in Adults with Metabolic Syndrome versus Healthy Controls: A Pilot Study Utilizing Targeted and Untargeted LC–MS Lipidomics

Learn more about MANA-ECM
Let Arpana Vaniya and Nicole Prince know you appreciate their effort to promote the (net)work of early career scientists as much as we do!

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Sapna Sharma

Sapna Sharma is group leader at Technical University Munich (TUM) in Germany

Favorite metabolite
lysophosphatidylcholines (LPCs)

Papers we discussed in this episode
Effect of BMI on type 2 diabetes | Metabolic Signatures Elucidate the Effect of Body Mass Index on Type 2 Diabetes
More on the KORA cohort discussed in this episode and in the first paper.

Paper on medication associations with multiple omics | Discovery of drug–omics associations in type 2 diabetes with generative deep-learning models
More on the IMI-direct project

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Karel Kalecký

Karel Kalecký is postdoctoral research associate
at the Institute of Metabolic Disease, Baylor Scott & White Research Institute in Dallas, TX (USA)

Favorite metabolite
Betaine (trimethylglycine)

Papers we discussed in this episode
One-Carbon Metabolism in Alzheimer’s Disease and Parkinson’s Disease Brain Tissue
Targeted Metabolomic Analysis in Alzheimer’s Disease Plasma and Brain Tissue in Non-Hispanic Whites

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Garry Patti

Garry Patti is Professor of Chemistry and of Genetics and Medicine at Washington University in St.Louis MO (USA)

Favorite metabolite
lactate

Paper we discussed in this episode
Isotope tracing in adult zebrafish reveals alanine cycling between melanoma and liver

“Cancer cells demonstrate a remarkable degree of metabolic plasticity.”

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Episode Transcript

Alice: Welcome back the Metabolomist. Today I’m joined by Gary Patti. Hello.

Gary: Good morning.

Alice: Gary, you’re a professor at the School of Medicine at Washington University in St. Louis. Would you like to tell us about your expertise and the topics that your research focuses on?

Gary: Sure. Happy to. First, let me just thank you for having me today. It’s a pleasure to be here and look forward to this interview. My lab is a metabolism lab. We do a lot of work in the metabolomics technology development space, but that space is complimented by a lot of applications. In addition to doing mass spectrometry experiments, we do a lot of biochemistry. We work with various different animal models. We do a lot of classical biochemistry types of experiments.
We’re primarily focusing on two different areas:

1. Trying to understand what the impact of chemicals in our environment is on metabolism. So you’re probably aware that we’re exposed to a lot of chemicals throughout our daily lives, whether it be from food, which are kind of obvious instances but also from somewhat unexpected sources such as hygiene products like shampoos or toothpaste, et cetera. And these contain a lot of chemicals; most of which we really don’t understand how they impact health and disease.
2. Trying to understand the development of cancer and tumors. And really in kind of a broad space – not just one specific kinds of cancer but cancer in general and specifically potentially how some of these exposures in our environment, whether it be dietary exposures or carcinogens to smoke, power plants, all of those types of exposures potentially contribute to the development of cancer.

Alice: Thanks. That’s really interesting. I noticed also in your work that you use a variety of models.
On the podcast, I’ve had a lot of guests who were working primarily with human samples, either blood samples or maybe tissue samples. I noticed in your lab you work also with cells and with animal models. In the way that you use them, do you have any pros and cons on using some of those models for different applications?
How do you choose which one you will use for a specific study?

Gary: Yeah, it’s a great question. We use, cells, organoids, various animal models ranging from zebra fish to mice to rats and also human subjects. They all have pros and, and cons. The cells are the easiest to work with, of course. They allow us to do certain types of experiments that aren’t feasible in human subjects. Data interpretation can be a little simpler in cells compared to some of the more complicated systems.
Later we might talk a bit about isotope tracing. Isotope tracing is certainly much more complicated in human subjects. Not impossible; but much harder, more expensive for reasons we’ll discuss – So the cells have real advantage. The way I like to think of it is that by using cells we can define the highways in some sense. The analogy that I always think about is that metabolism is much like a street map.

What we’re trying to do when we think about metabolism in the context of a particular application is understand what the density of traffic is on any given street and how a particular cell gets from destination A to destination B. How we approach it in the lab is that cells provide a really good opportunity to define the roadmap, to define what streets there and how traffic is activated on the streets. What we find is that the ways in which cells transform nutrients in cell culture can look very different than the ways that cells might transform nutrients in animals or in humans. I do think using a combination of all of those different models is important. because, you really want to validate everything in human subjects ultimately, I would say is the endpoint.

Alice: And we’re going to have also a really great example of this added value of having a whole organism in the paper discussed, because we’re going to look at interactions between different organs or different tissues within the animals that you studied. Even though I’m a big defender of in vitro experiments, I think there is really a lot of great information you can get from cell culture. There are certain things that you cannot do or that you need to know to start to model them. If you want to do co-cultures, for example, you would have to have the idea in your, in the case we’re talking about, or we will talk about to grow the tumor cells with liver cells than see if they interact or not in the way that you found out. But you use the animal model really well for this in the paper we’ll discuss.

Garry: Thank you. Yeah, we try to go back and forth. I see it as kind of a ladder. It’s not unidirectional where you have to start themselves. I mean, sometimes you can start in human subjects and go backwards. Sometimes you start themselves, generate a question and validate. It’s really bidirectional.

Alice: Yeah. I think when you have the variety then you can use each one of them for what it is good for, and then combine them in the best way that you can think of.

Garry: Yeah, I totally agree.

Alice: You said most of your work is focused on metabolomics or you do a lot of your work with metabolomics. How did you discover metabolomics? Do you remember your first encounter with this method and how that was for you?

Garry: My path to metabolomics was certainly non-linear. I’m, old enough to say that when I was doing my work, metabolomics was not the buzzword. Now, I think, metabolomics has almost become a synonym with a lot of the work in metabolism. Most people that do in-depth metabolic analyses in some way or another are using. That wasn’t the case when I was starting in metabolism, my initial areas of interest were actually related to bacterial resistance to antibiotics. I was trying to understand how certain bacteria particularly noscomial pathogens develop resistance to different kinds of antibiotics that are used in the clinic, such as penicillin. And as it turns out, perhaps not surprising to you, a lot of these drugs have a strong metabolic component. They have a big impact on the metabolism and it’s specifically the metabolism of a particular structure in bacteria called the cell wall. So these antibiotics that we were looking at interfered with the production of the cell wall by the bacteria. I was using at the time nuclear magnetic resonance (NMR) to study bacterial metabolism. Bacteria are a little different from the other types of systems that we were talking about moments ago with cancer because the physiology is less important. You don’t have the type of interactions with a tumor. Of course; you have a bunch of cells that constitute tumor. Tumors aren’t just one type of cells, but bacteria are a little simpler in the sense that you can grow them in culture and it’s a little easier to understand whether it is relevant.

Alice: One could argue that now, especially if you talk about common bacteria, that you have this rather complex network now with the host metabolites and all different kinds of signals that are going to influence. So similar to what you would do with antibiotics, but in this kind of exchange. So we found a way to make bacteria complicated, haven’t we?

Garry: Yeah, that’s a great, point. Certainly in the gut microbiome there’s a lot of important interactions there and the context of Penicillin resistance or antibiotic resistance, least in what we were studying. – It was a little simpler. We weren’t trying to probe those types of questions, but absolutely. The simple thing is that NMR isn’t as sensitive as mass spectrometry, which is another major analytical tool that one can apply to study metabolism. But because we were looking at bacteria, we could grow them up in huge flasks of many liters. I was doing experiments sometimes with 10 liters of bacteria. That´s a lot of bacteria and analyzing them by NMR provided a lot of insight, as we matured in those studies, I became increasingly interested in intermediates and pathways that NMR couldn’t resolve. It seems natural to me that one would try to turn to a different technology to look at those intermediates to better characterize those pathways and start thinking about flux. We were using isotope tracing at the time, not really to trace pathways, but to increase sensitivity in NMR. Not all nuclei or NMR active. We added C 13 to bacteria samples in those days because C 13 is NMR active; C 12 isotope is not. This increased our signals. The idea for me was moving towards other technologies that would have better sensitivities that would allow us to probe those metabolic pathways. So it made sense to start thinking about mass spectrometry. At the time this wasn’t really put under the word metabolomics. We were using mass spectrometry to start thinking about those pathways. It was just biochemistry in those days. Now, of course, we’d recognize that as being metabolomics.

Alice: From what you just said, we can see that you have quite some experience with the method and you have quite a few years of working with metabolomics behind you. So maybe for the younger audience, the people who are starting with metabolomics, is there anything you can think of that you wish you had known that maybe you could share with them to make something a bit less difficult for them in the near future?

Garry: I think in those early days when I first started doing mass spectrometry analysis of metabolic extracts, we would do much like people do today: You isolate, get rid of all the macromolecules and you isolate all the small molecules and analyze them by mass spectrometry. – And I was just initially astonished at how many signals there were in the datasets. As a trained biochemist, I was familiar with central carbon metabolism and the handful of pathways that surrounding them. But we were seeing a couple thousand metabolites. When we first started doing those experiments and we were seeing, 10,000, 20,000, 25,000 peaks it was really shocking. And I thought at the time – extremely exciting. The computational resources that we had at the time were much less developed than they are now. So most of the things we couldn’t identify, I guess you could argue that we still can’t identify a lot of those different signals, but I was under the impression, at least initially, that a lot of those represented new molecules that had never been reported before. That those were so-called unknowns or novel metabolites that maybe weren’t in biochemistry textbooks yet, but that were present in cells and that served an important biochemical function.
One thing that I vastly underappreciated at the time was how complex mass spectrometry metabolomics data are. You could take one standard. Glutamate or whatever your favorite metabolite is and put it in a beaker and analyze it by mass spectrometry. In a perfect world you might think you get one peak for glutamate, but can see a hundred or 200 or more peaks that are derived from that one metabolite. The metabolite can break into pieces. It can adduct, it can fragment; there’s contaminants in the samples. And at the time I didn’t appreciate all that because I had underestimated the complexity of the data and I also overestimated how many molecules we were actually measuring. I think that led to much more sophisticated and complicated interpretations of the data. That were probably needed. So I would say, for those starting the journey: Assuming that something that you can’t identify as an unknown is probably not a novel metabolite. I think that shouldn’t be necessarily the default hypothesis.

Alice: Thank you. So let’s move on to the main topic that will be related to the paper we’ll discuss today.  The paper about a model of melanoma. Let us discuss a bit cancer and metabolomics. When we think of cancer and metabolism, we think of energy metabolism. And this is a big part of the paper that we’ll discuss and carbon metabolism. But could you give us an overview also from the work that you’ve been doing with your group? So what are the key elements of this energy metabolism, but also of other parts of metabolism important for cancer that you would like to highlight?

Garry: Absolutely. Energy metabolism is particularly fascinating in cancer cells because cancer cells employ a particularly surprising metabolic program. What characterizes cancer cells is an uncontrolled capacity for proliferation. So that is to say that we have one cell that rapidly turns into two cells and four cells et cetera. And so if you look at what characterizes cell proliferation or cell division, for a cell to turn into another cell or multiple cells, it has to replicate its contents. And so that means that it has to remake all of its genetic material; it has to remake all of its plasma membranes; it has to remake all of its proteins. So that is actually a substantial anabolic load or synthetic burden that’s associated with cell proliferation. So when we think about that, it seems that that would require a lot of energy. Making things generally seems like it should be associated with an energy demand. But what’s fascinating about cancer cells is over a hundred years ago now, one of the pioneers of biochemistry, researcher by the name of Otto Warburg in the early 19 hundreds, discovered that cancer cells do in fact take up a lot of glucose, which is one of the most prominent nutrients that’s available in our circulation.
What is surprising is not that they take a lot of glucose but that most of the glucose is transformed into a metabolite called lactate. – Generally lactate is recognized as a waste product of cells. If you transform glucose into lactate it only yields two ATP through glycolysis.
If you take glucose and you oxidize it in mitochondria, in contrast, it yields on the order of 30 to 38 ATP. So it would seem intuitively that a cancer cell with all of these synthetic demands would be trying to maximize the amount of energy that they can produce. The reality is that most of the glucose that they metabolize gets metabolized in a very inefficient way so that it yields a relatively small amount of ATP per glucose. And so for over a hundred years, cancer biologists have been trying to understand why in the world cancer cells would do this.
It’s not just in vivo, by the way we were talking about that – the difference between in vivo settings and in vitro settings. You see the same thing if you take a cancer cell and you put it in an oxygenated cell culture. Do the same thing where it takes up a lot of glucose and transforms the majority of it into lactate. I think that’s one of the really interesting energetic paradoxes. When these cells seem to need so much energy, why would they be yielding so much? Why would they be getting such a low output of energy from glucose?

Alice: Also one way that we’ve been looking at cancer for the last decades is through the angle of genetics or transcriptomics or genetic changes. Because we always think of the mutations in the genes that are at the origin of the problem with the cancer cells. We see more and more that metabolomics is a very interesting tool to study cancer. So do you have any arguments as to why metabolomics is a good omics to study cancer compared to other ones like genomics or proteomics?

Garry: I would say that the other omics are equally important. I wouldn’t say that metabolomics is any more insightful than the others. You could, in fact, make arguments in any of the omics, depending on what kind of context using one is more important than the other. I would say that most cancer cells are rapidly dividing. That does require metabolic alterations. It is more of a universal hallmark of cancer. There are metabolic properties, this phenomena that I described where most of the glucose taken up is similar across most cancers. That makes it nice because what you’re describing is certainly true, that there are a lot of different genetic changes across different types of cancers, but many of the metabolic properties are conserved. That is attractive in the sense that it at least invites the possibility that there might be ways therapeutically to approach cancer that could be more universal. Now in practice, whether or not that’s possible is a totally different story. I think the other point that I would mention is that metabolomics provides a nice biochemical readout of what’s happening with the phenotype. Doing Transcriptomics, for example, is a powerful type of analysis that a lot of people do. It’s very standardized. There’s a lot of public data that’s available and you get better coverage than you do with technology such as metabolomics. – But at the end of the day it doesn’t necessarily tell you what’s happening phenotypically: Just because you have an increased or decreased expression of a particular transcript doesn’t necessarily mean that that pathway is more or less active. Furthermore, even if a transcript does correlate with a change in biochemical activity how much the metabolic pathway changes? You can’t infer that from transcripts, at least not reliably. And so metabolomics is really a necessary  extension to understand biochemistry if you want to understand and be very precise with some of those types of assessments.

Alice: I think in cancer, as in other diseases, we often see the thinking work going in both directions – from the genetics to the metabolomics or from the metabolism to the genetics. And it works really well both ways, but you can’t really predict. I just had this conversation with another guest about a study on asthma where you can’t really predict which direction is going to be – they’re both very valid ways of looking at the data.

Garry: I think, now, there’s a lot more genomics information available. So more often we tend to start there because particularly in the clinic, there’s more information readily accessible.

Alice: When we prepared the episode together we touched on the topic of nutritional intervention – How do you see that being a promising tool for either studying or finding therapeutic ways to address cancer or maybe other diseases as well?

Garry: I think nutritional intervention is a very intriguing area right now and we are finding that cancer cells require certain nutrients. It’s a very provocative idea to say “what would happen if we deprived cancer cells of these nutrients? Would that prevent them from proliferating and thereby mitigate disease?”

Rachel Kelly

Rachel Kelly is associate professor at the Brigham and Women’s Hospital in Boston MA (USA)

Favorite metabolite
sphingosine-1-phosphate

Papers we discussed in this episode
Metabo-Endotypes of Asthma Reveal Differences in Lung Function: Discovery and Validation in Two TOPMed Cohorts

Phenotypically driven subgroups of ASD display distinct metabolomic profiles

Activities in consortia
EnviroGenomarkers
COMETS – The consortium of metabolomics studies
COMETS analytics

On similarity network fusion

“Metabolomics is its own beast. You can’t treat it in the same way that you would genetics”

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1

Available on Amazon and the biocrates webshop

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Jennifer Kirwan

Jennifer Kirwan is Head of the Metabolomics Platform at the Berlin Institute of Health at Charité (Germany)

Favorite metabolite
Melatonin, a close relative of tryptophan, Jennifer Kirwan’s favorite metabolite at MetSoc last year and the metabolite chosen by our previous guest David Wishart!

Papers we discussed in this episode
Translating metabolomics into clinical practice
Don’t forget to send your constructive criticism on this opinion piece in a comment via LinkedIn

Identification and validation of small molecule analytes in mouse plasma by liquid chromatography-tandem mass spectrometry: A case study of misidentification of a short-chain fatty acid with a ketone body

Her past and present activities in the metabolomics community include

• mQACC consortium, promoting best QA/QC practices in untargeted metabolomics. Learn more in the consortium’s recommendations here.
• Precision medicine and pharmacometabolomics task group of the Metabolomics Society
• German society for metabolome research (DGMet)

Sign up for The Metabolomist mailing list to be the first to hear about the latest episodes and news around metabolomics at https://themetabolomist.com

Finally available – Alices first book
The STORY principle – A guide to the biological interpretation of metabolomics
Also featuring some of the Metabolomists from Season 1
Available on Amazon and the biocrates webshop

Episode Transcript

Alice: Welcome to this new episode of the Metabolomic podcast. Today my guest is Jennifer Kirwan.

Jennifer: Thanks for having me.

Alice: Thanks for being here. You are the head of the Metabolomics platform at the Berlin Institute for Health in Germany. Can you tell us a little bit more about your background and the work that you do there?

Jennifer: Yes. My original background is as a veterinarian and I worked as a clinician for a few years before moving by accident into metabolomics where I’ve stayed ever since. I discovered I love statistics and really enjoyed the work. So I moved out to Berlin from the UK about six years ago. And the Berlin Institute of Health became the third pillar of the Charity teaching hospital a couple of years ago, where we’ve remained ever since.
We are working mainly on translational medicine. And our primary role is effectively to support other people to answer their research questions in metabolomics. We also do some of our own research. Biologically we are very interested in the heart-gut-brain triage and how different metabolites are communicating between these three organs and how they may be involved in health and disease.
We also do a lot of research on quality management, quality assurance, and quality control in metabolomics applications.

Alice: Thank you. You are also involved in certain working groups in the metabolomics community? I think you’re quite active here as well.

Jennifer: We are extremely active. I’m a central committee member for the metabolomics quality assurance and quality control consortium. This is an international consortium dedicated to furthering better quality management in metabolomics. I’m an active member of the Precision Medicine Working Group for the International Metabolomics Society, and I’m a founding member and former board member of the “Deutsche Gesellschaft fuer Metabolomforschung” which translates into English as the German Metabolome Research Society.

Alice: As you mentioned, you didn’t start out in metabolomics like many people because metabolomics is a relatively recent technique, but, you really had a kind of indirect path to it. Do you remember if there was something that you found particularly challenging when you met this method in your work?

Jennifer: I found all sorts of things challenging.  I’d been working as a clinician and I’d been used to equipment that just works. And when you move into a scientific sphere, you are asking yourself why you’ve spent half a million or a million euros on Massspec equipment and then you spend most of your time troubleshooting.

Alice: Troubleshooting. It’s a really interesting perspective.

Jennifer: Yes, and we had lots of problems with a new instrument when I first started a PhD and I spent a lot of time staring at chromatograms, trying to work out what the problem was. In the end, I think it really developed my scientific skills and my lateral thinking skills, and it’s made me into a better scientist.

Alice: Yeah, I guess that comes from the fact that metabolomics is a very multidisciplinary technique you can’t be an expert at everything, but you need a language that’s understood by many different types of experts that that kind of opens your horizons.

Jennifer: So this is a really interesting perspective that you’re offering here. When I started in metabolomics, we were expected to be experts in everything. And I did everything from study design right through to my own bioinformatics and data analysis. And as I’ve stayed in the field, I’ve seen how it has diversified more and more into very specific fields where you have one person responsible perhaps for the analytical side and another person is doing only bioinformatics and another person may only be doing statistical analysis. This has tremendous advantages, especially as we are getting more and more complex in what analytics we can do. But you are absolutely correct that unless we learn each other’s languages and unless we learn something about the background as to how this data was collected, how this data was processed and how it was analyzed, it’s very easy unwittingly for there to be communication problems. – And this leads to misinterpretation of data. Absolutely.

Alice: And this is exactly the main topic of our discussion today. Because if, let’s say the analytical chemist not aware of certain details about or important things about the measurements and this is not communicated later on to the people who process the data, either the bioinformaticians or the people who do the biological interpretation of the data – this can have huge repercussions. We will have a very good example of this in the second paper we’ll discuss today.

Jennifer: I think this is a time really to be celebrating different specialties in teams. It’s very easy to join a team and see what somebody else can without realizing what you are bringing to the team and I think we need everybody on board bringing their own specialities to get the best possible results. And that’s analytical chemists, biochemists, statisticians, bioinformaticians. We need everybody at the table.

Alice: That means, when we want to apply metabolomics to the clinics or to clinical research, we also need to get the clinicians on board, the regulatory people on board and so on. Right?

Jennifer: So this is absolutely a big theme that we need to be discussing with clinicians and the regulatory bodies more. And we shouldn’t leave out other important groups such as patients, ethical advisors and privacy experts. Because I think the more that we move into the clinic, the more we start stepping out of our own comfort zone and we are going to need to bring more experts on board to make sure that we do a good job of this.

Alice: Thanks for this. That’s really true. So let’s move on then to the opinion piece: The first paper I wanted to discuss with you is relatively short in size but I really enjoyed reading it – it’s titled “Translating Metabolomics into Clinical Practice”. The goal is clear. You’ve published this very recently in nature reviews and amongst several things that I really enjoyed in this paper was that you write that metabolomics is on the precipice of transforming from a research tool to powerful clinical platform to improve pre precision medicine. You begin with the application of metabolomics to precision medicine. Why did you choose to focus on precision medicine specifically?

Jennifer: I suppose because this is where I see the real power of metabolomics. We know that medical doctors are experts in diagnosing multiple diseases but we still have diseases that it’s difficult to predict outcome. It’s difficult to predict response to medication. And there are still a number of diseases which we label these as clinicians as diseases of exclusion. So we’ve done every other test and we can’t find anything else wrong. And I think that by having a metabolomics approach and possibly a multi omics approach, we may be able to start subcategorizing diseases and subcategorizing groups of patients in terms of who’s likely to respond best to treatment A and who’s likely to respond best to treatment B. I think this is going to be of benefit to everybody as long as we make sure that we think about the ethical implications of cost and accessibility from the very beginning.

Alice: Very true. And there’s also the aspect of how metabolomics is responsive to the peculiarities of each patient – to the intra-individual variability and the inter-individual variability as well. So how we differ from each other and how our own metabolome will change over time. This is something that even other omics usually don’t give us access to. So does that bring extra power to metabolomics for precision medicine?

Jennifer: It brings extra power and it brings extra frustration. One of the beauties of metabolomics is that you have this large inter-individual variability, and we can use metabolomics effectively as measure of phenotype of the individual. At the same time it’s a very sensitive method and we’ll see differences purely down to age, down to diet, down to ethnicity, down to social lifestyle. These all need to be teased out when we’re thinking about diagnostic tests. This brings us to an additional challenge of what needs to be considered as a confounding factor and what we could actually consider as a co risk factor. So is age a confounding factor or if a particular disease is associated with an aging process, is it actually an important part of the equation when you’re analyzing the data?

Alice: That’s a really interesting point. You discuss several diseases throughout this paper and I found it particularly interesting when you talked about psychiatric diseases that are not necessarily an example that people would go to because it’s maybe not the most classical diseases. -Of course you also talk about cancer, you talk about things that are usually in the foreground for many papers that discuss metabolomics. I really like that you chose psychiatric diseases like depression, like schizophrenia to demonstrate also where metabolomics can help and where maybe we need more help because other tools have less potential to give us information. Can you explain why you chose to discuss these diseases?

Jennifer: It comes from a very personal history and I should be clear that we have worked on one or two projects looking at psychiatric diseases, but it’s not our major focus. Several years ago I read an article in a newspaper about an individual who’d spent time in a psychiatric institution and she talked about how she’d been treated as scary as other by the staff. Then one day a psychiatrist thought to do a clock test on her and discovered that she’d lost half of her vision. And from there they actually did some physical tests. – I think MRI can’t remember what the eventual diagnosis was but the conclusion of this article was how staff started relating to her differently once her disease was perceived as physical rather than psychiatric. Now we all know people who have depression. Some of us know people who have schizophrenia and many of us also will know people personally who have Alzheimer’s. I think when we stop trying to divide diseases into psychiatric and physical it gives us a new perspective on how we can look for treatments and possibly even cures. We are finding out more and more about the microbiome and the gut microbiome in particular and it is clear evidence that there’s a direct brain-gut link there through the vagus nerve, but there also seems to be other roots. And I think as time goes on we’ll find that psychiatric illnesses will have a biochemical and possibly a microbiological component to them as much as a sociological and genetic component to them.

Alice: Absolutely. There are many places where metabolomics helps us for this type of diseases and others also. There is the understanding, the mechanisms and the etiology, there’s the diagnostics, but there’s also treatment for choosing which treatment to give to people. Knowing who will be responsive to what. I worked a lot on depression last year and it’s still amazing to me that most people won’t respond to the first line treatment. They won’t respond to the second line treatment. And so we keep giving drugs to people and we just test it directly on them without knowing what effect it will have. When we start looking at the signatures of the response of two treatments in patients in their blood or in other matrices, you start to see that we might be able to predict who will respond to which treatment. At the minimum, avoid giving something that has really serious side effects to someone that we know will not respond to it. And also to directly give people what will be effective for them. And this is important for any disease, but it’s also very interesting for psychiatric diseases.

Jennifer: And I think this knowledge will also start spreading into physical diseases. So people with heart disease, for instance, have a much higher risk of depression than the general population. I think many people have just assumed it’s because they’re sick but there’s some evidence to suggest that there may be microbiome signatures of that, which may be playing into the depressive symptoms, and then the depression is actually potentially part of the disease not as a result of the disease. I’m careful what I say here because this is not my my specialty.

Alice: It also reminds me of what you said earlier, also when we plan greater projects maybe this is also a place where discussing with patient groups and with other parts of society that helps us to identify what are the important points to look at is really important because the perspective of the patients, of the families of the patients, of all sorts of people and institutions can really help to address this better.
Following up on the opinion piece that we’re discussing, I noticed there were quite a few mentions of artificial intelligence or machine learning that you suggest they can be really useful for different uses and applications of metabolomics in the clinics. Especially, there is one I really liked where you talk about regulation validation, which is often an issue that people face where they’re like – I have a great signature, I have really good ideas for biomarkers for this disease I’m interested in, but either I don’t know how to get into the clinics or it’s too expensive for me or there are many kinds of barriers on the way to, the clinics. You made an interesting suggestion where you write that, there could be a way to have only disease specific biomarker algorithms that would have to be validated and not the chemistry itself.

Jennifer: I should say that the initial idea of having a non-validated test was not mine. It came from a couple of discussions. One of which was with Annie Evans who has been working with Baylor Children’s Hospital – this is a paper by Lou et al. – And what they did to overcome this regulatory hurdle was they started running tests on blood samples using non clinically validated tests to direct where they send the tests to do the clinically validated tests, if that makes sense. Basically, you run a non-validated test and you have an answer that would suggest that it’s disease X. Now you can take a blood sample and you can run a validated clinical test to test for disease X. I think it’s a great idea because it saves time and money diagnosing sick patients quickly. What you can then also do is you can start building up databases. The more samples you run to start categorizing your unknown samples into metabolic categories that you may then in the future be able to diagnose with specific clinical tests. It allows you effectively to identify new subgroups of rare diseases. Now, this doesn’t stop you needing to validate your metabolomics test. But it lowers the barrier because now you are doing this as an in-house validation rather than necessarily needing to do a full regulatory FDA approved validation. However, I have a greater vision, which I hope others will share. That if we could start being able to validate wide targeted metabolomics tests, then we could start using these in a similar way. Where now we’ve got one test for multiple diseases and we are now just having to validate the algorithm. – That’s what I was explaining in the paper.

Alice: So this is the dream – You have a set of „pre-biomarkers“ that would cover a large ground of diseases that you would look for in the patients and then it would point you towards the ones that might be happening in the patient. Did I get that right?

Jennifer: In newborn screening, which is what Baylor is using this for, they’ve often got fairly extreme metabolic changes but we’ve all been in the situation where we are finding the same metabolites popping up again and again in various different diseases.
I’m actually part of a larger consortium. It’s a EU Horizon Grant looking at inflammatory diseases. It’s called IMMEDIATE. And we’re looking at how inflammation affects the metabolome and may be attenuated by microbiome changes. If we can pass some of these regulatory hurdles to have an analytical test accepted as the data is valid. This gives a lower barrier to just needing the data to prove that your algorithm is now valid.

Alice: Thank you. One follow up question on what you said. – You mentioned that it’s often the same metabolites that are seen to be changing in different diseases:
Can’t machine learning also help us with this? So instead of looking for metabolites X, Y, Z, that we would look for different patterns of change in these metabolites that might be different from disease to disease. And so we can have the same set of metabolites but the algorithm is the thing that tells us what is really happening, even though the data seems to be the same for all the diseases.

Jennifer: This is exactly the message that I was trying to get across. You can do it via machine learning and I think we need to be embracing machine learning more and more. We’ve got data to show that you may only need 5, 6, 7 metabolites to diagnose certain diseases. Which then in theory makes it possible to do by a hand. But, I’d say certainly from the discovery stage that machine learning is a very powerful tool when used correctly. One of the challenges that we have as a community is to make sure that our data for our machine learning tools is not only robust but also sufficient in quantity to actually make good machine learning tools.

Alice: This is gonna be one of the challenges. I think everyone is getting interested in the topic but there is a trade off between the type of models you use and the amount of data you will need. We need to have enough data to feed most of those algorithms. Yes. And it’s also about how complex you need to make your machine learning tool. Principle components analysis is often described as machine learning and it’s not particularly challenging to use. There’s always a temptation to use much more complex tools like neural networks, for instance. The more complex your tool, potentially, the more you are able to find patterns that are more difficult to find by hand, but you are more at risk of not being able to explain the decision making by the tool. And this can lead to unforeseen biases in your data analytics. I know the machine learning community is working hard on making sure that tools are explainable.

Alice: And you discuss also in the paper these biases and I guess in the application of machine learning to clinical applications we have the same kind of bias that we would get in any other applications of machine learning, for example, based on sex differences or ethnicity or also age differences that will have a large impact on the metabolism.

Jennifer: Then we would also have to have training sets that represent a large enough proportion of the population, or at least that define which part of the population was addressed, that we know where to apply it afterwards. So we have have these biases – And you also have the unexpected biases. Yes. There’s a famous example of a machine learning tool for, I think it was pneumonia and they wanted to predict which patients should spend time in the ICU and which patients could just remain on the normal ward. They found to their horror when they started implementing it, that it was saying that the sickest patients could stay on the ward. I think particularly patients with asthma. When they started investigating why, it turned out that their machine learning was using data were effectively the doctors were making a decision early on to send patients with asthma to the ICU so they were getting better and the machine learning tool was misinterpreting that as they shouldn’t have been in the ICU in the first place. It’s a beautiful example of why you really have to think about the data and not just accept what the computer’s telling you.

Alice: Yes. It’s a great example. Are there other aspects of the paper you would like to discuss?

Jennifer: So one of the things that we as a metabolomics community are enthusiastic about is data sharing. For this vision of transferring metabolomics into the clinic, to be realized we’re going to need large amounts of data. That is both going to probably need to be a community effort so that we can try and overcome some of these biases and get enough data – but it gives us the additional challenge then – how do we assimilate and compare data collected in different labs on different instruments. And we have this challenge of transferability of data. As a community, I think that we need to be fighting for standards of measurement with mass spectrometry, in particular. It doesn’t necessarily have to be absolutely quantitative, but we need a standard by which we can give a real number to our measurements that is meaningful.
What do I mean by that? Well, absolutely quantitative is quoting something as, micrograms per microliter or something similar. What may be more realistic may be to have a known standard that we can give a relative quantification to – That is validated.

Alice: I think it’s a good point, but isn’t there another way to address this? You have quantitative measurements, and of course you have variability for a given individual and between individuals, but then you can compare that to reference ranges as well for metabolites that would be quantified in large populations. – Then you could have ideas if you’re within the usual range or not, it’s actually what is commonly done in the clinics for all the metabolites that we measure in the clinics at the moment.

Jennifer: Yeah. So where you are able to quantify metabolites then absolute quantification is definitely the way forward. We know for instance that in the clinical labs there are well designed protocols to make sure that clinical labs are measuring things within a certain era of other clinical labs and this is definitely the way forward. I’m also thinking about much larger quantity of metabolites and particularly lipids which we have no standards for and for which absolute quantification suddenly becomes much harder.

Alice: Yes. My last question about that paper was – Did you get any feedback from the metabolomics community or the medical community on this opinion piece? Have you had a lot of feedback yet?

Jennifer: I had a lot of feedback. I had a lot of people asking me for access to it. I had it advertised on my LinkedIn and I got a lot of positive comments. I don’t know whether people are just very nice. So far I haven’t had any constructive criticism. I’m actually going to open to the community that if there is constructive criticism on my opinion, then I’m very open to hearing it. Because I think that science is about always being open to improving.

Alice: Yes. And it’s the point of an opinion piece. You give your opinion and then you hear the different opinions, but exactly that’s how the conversation gets going. It’s good. I would move on to our next topic, which is Biobanking. And especially considerations for biobanking of different types of samples for precision medicine. I think you have a lot of advice for this especially for people who want to use samples from biobanks and should maybe think of what to look out for and of course, for people who will be involved in collecting and storing samples and maybe other activities.

What are points that are important; that we should all know about?

Jennifer: I think if you are planning a study, then talk to the experts early. And this sounds obvious, but I think most of our listeners today are probably from the metabolomics community and will be very used to people turning up with samples that are 10 years old and want them analyzed  – And they were not collected with metabolomics in mind. To get the best quality data you want the least technical variation. And if you want the least technical variation, you need to think about it in the planning – We all know that the challenge with metabolomics is that the people collecting the samples are often not the metabolomics experts and they’re often busy study nurses or sometimes general nurses who have been asked to collect this as part of their everyday job when they’ve got X number of other more urgent tasks to do. And so I think the first place to start is to engage with the people that are working on your team, sit down, discuss what you are after and why it’s important.
Have the discussion about what’s most important, what’s going to affect results, and hopefully foster a sense of community of engagement and of enthusiasm, so that everybody’s on board to follow protocols – And has a good understanding of what that protocol means. Because it’s one thing to write a protocol, it’s another to follow what someone else has written in the same way. Protocols are essential. They should be written – I think if you have videos, it’s even better because then people can follow them and understand exactly what you mean. And as an individual, it’s very important that you understand what your study design requires. By which I mean: If you want your perfect metabolomics research where you want something that is going to be as representative as possible of your sample, then you need very careful biobanking techniques:
You need to think about temperatures. You need to think about how long you keep blood as blood, how long you keep plasma out of the fridge, or preferably the liquid nitrogen for. But if your overall aim is to have something that’s really robust as a biomarker, then you may be having a conversation with the clinicians that they’re more interested in biomarkers that can survive real life clinical conditions. I, as the metabolomics person, would always rather have perfect conditions. That’s clear, but we have studies ongoing that for cost and practicality reasons we made a clear decision that we are looking at a different scientific question. This may surprise people, but there has to be an element of reality of clinical life in your decision making.

Alice: Absolutely. It’s really interesting also from that perspective to consider the planning as a very broad way of looking at it. It’s not just planning the details of the experiment, but really planning what you hope that your work will turn into maybe 15 years down the line. Maybe one day this will be used as a biomarker in the clinics. It’s a very different question indeed than to look for the perfect signature in perfectly preserved samples. That’s a really good point. I like it a lot.
Of course I can’t, not mention this cuz you talk about the importance of project planning and also in the story principle, that is my very first step. The first step is you sit down and you plan your experiment from beginning to end. From thinking, I want to perform this experiment, get this kind of samples, and do this kind of analysis because I want to answer that question. And then this whole process is gonna be very important to determine if you have the luxury to really work on your study from beginning to end including collecting the samples, which is not always the case, but if you can plan it as the person who does the metabolomics or the team who does the metabolomics, if you can plan it, also considering which samples you want, how old they should get, should you get them from a biobank, are you collecting yourself? – These kind of things. It’s really crucial because for the quality of the samples it’s gonna have an impact, but also for the results of your interpretation it’s gonna have an impact. If you end up not having exactly the samples that you wanted to have or that you would’ve needed to answer your question, then the whole project is a miss. Do you have experience ordering samples from a biobank yourself? – Is there something that people should be careful about you think, or that they should ask the biobank?

Jennifer: Definitely how the samples were collected and aliquoted. And the time points and temperature points across the sample collection chain. I think this is crucial. For all I’m saying that we’ve got projects ongoing where we are looking for more robust biomarkers that can survive clinical practice. The success with metabolomics is often about how much you can reduce technical variability and noise. And the more variability you have, the more you end up having to throw out metabolites because they’re too technically variable. I think there are some ISO standards for metabolomics collection now. To the best of my knowledge, they’re not yet widely adopted. I’m going to be interested to see in the future how widely adopted they are by Biobanks. I think having some standardization of sample collection across Biobank is certainly going to be useful because it comes back to this intercom comparability of samples.

Alice: For people who are interested in samples from Biobanks – My next guest will be someone who did a study using brain samples from a Biobank. He insisted to discuss what you know about your samples because he saw some interesting things on the brain samples that he was using.

Jennifer: We basically worked with our biobank to design a protocol that’s suitable for metabolomics and proteomics collection. I think this is now in use for most studies that go into the biobank.

Alice: So let’s go to the second paper we wanted to discuss. The one focused on quality assurance and quality control. We also will be looking into a very interesting case where one of your post docs found that one metabolite was not what the world thought it was. You can find the link on the shownotes for this episode:
It’s called „Identification, validation of Small Molecule Analytes in Mouse Plasma by LCMS, a case study of misidentification of a short-chain fatty acid acid with a keto body“. The first author is Marielle Garcia Rivera. You wanted to discuss this paper today. So can you begin maybe by telling me why this is the paper that came to mind?/ Why you wanted to discuss this one specifically.

Jennifer: I really like this paper because of what it says about thinking. Maryelle, she’s a former postdoc of ours and a very talented chemist. She was given the job of implementing short chain fatty acid method from another lab and getting it working within our lab. As she was working on it, she noticed some inconsistencies with some of her results. In particular she was looking at adding a couple of extra keto bodies and when she looked at the 3-hydroxy butyric acid, she found that there was an extra peak where there shouldn’t be. She’s somebody who is naturally curious and really thinks about data. And so she investigated this and this peak happened to match with the same transition peak as acetic acid, which was one of the metabolites that the on short-chain fatty acids that we were interested in. Sorry. HBA is actually a keto body. She investigated this and found that not only the seemed HBA to be suffering from in-source fragmentation in the mass spectrometer but when we were analyzing the plasma samples, there was a strong probability that our quantification of the acetic acid was now being affected by the this HBA transition – the [putative] in-source fragment. I really love what she did here because she found an interesting result, she followed it through and ended up just adjusting the method so that we could detect both the HBA keto body and the acetic acid much more accurately and with confidence. And it’s a great example of how quality management and good scientific observation, really pays dividends in terms of improving the quality of your data.

Alice: And also I think of taking ownership of the experiment that you make. There are some people who would say – this is the protocol. I’m just following the protocol. But it’s really important when you see something that doesn’t quite fit to look into it and to try to understand why it’s not what you expect.

Jennifer: I think we need to be celebrating analytical chemists more. They are central to metabolomics and it’s where it all begins. We are not always appreciating the hard work that they’re putting in.

Alice: I’m not an analytical chemist, so I rely on the data that is provided to me by the people who measure it. And I like how the paper begins because of course you start a scientific paper, you always want to demonstrate the relevance of what you’re discussing. So you always start with health and diseases. This paper begins by saying both of those metabolite classes are very important, metabolites for immune responses, diabetes and cardiovascular disease. And, of course, that’s true of both short fatty acid and keto bodies. And from the point of view of the work that I do, in majority, which is the biological interpretation of the results, if you give me a concentration and you tell me this is this short chin fat acid. I’m gonna make a story based on an increase in this metabolites in the in the sample. And if you tell me we have both a short chain fatty acid and a keto body, it’s going to be a different story. So the implications are huge when we think of the applications we’re going to make of the data at the end. It’s really important.

Jennifer: That’s one aspect of it. And the other aspect you can imagine if you’ve got a 10 or 20% noise level because there’s an additional compound there. You could end up missing that there’s a change in either or both compounds because there’s too much variation in the data.

Alice: And that’s why it’s really a good example of the importance of QA and QC in metabolomics. What are your main recommendations in that regards? You’re, involved in this – Working group or it’s a consortium.

Jennifer: It’s a consortium. International consortium of people who are really engaged in quality management and metabolomics. I’m going to be honest and say: I got into quality management by accident. It’s not something that the majority of people wake up one morning and say, „I’m going to be a quality manager“.

Alice: You did it by accident or by necessity?

Jennifer: Well, by necessity.  But I’d also say, and here I have to thank my ex-boss Mark Viant,  another really talented chemist. He infused me about thinking about data in another way. So not just the biological interpretation, but how the technical variability and the way that it was collected may be influencing the final results. Because of his guidance I started getting very deeply involved in this subject. And the more involved you get, the more excited you get by what it’s then possible to do with good quality management techniques. This is a huge task. We wrote a white paper on the subject and we restricted ourselves just to quality control samples because we decided that if we did any more then the paper would basically be too large.
It would be a book, probably. I think that every stage of the process benefits from a good quality management strategy – every stage. There’s this wonderful lecture on statistics by David Broadhurst where he talks about marginal gains and this is the idea that he uses the British Olympic Cycling Team as an example. This is a fantastic example to use because effectively what they did was they looked at everything. They looked at everything from the saddles that they were using to the pillows they were sleeping on so that they could get the best night’s sleep. And they tried so save 1% of our time here, and we save 0.2% of our time here, and we saved perhaps 3% by doing this and as an overall result of these different things and these tiny incremental time savings they ended up with the most gold medals that’d ever won. And we are doing the reverse with quality management. We’re saying, okay, if we have a very tightly time controlled collection procedure for our samples at biobanking; if we have this very rigorously timed and controlled sample preparation procedure for our analytical preparation; if we make sure that our batch lengths are less than X number of samples, then we may at each step be reducing the variability in any individual metabolite by an incremental amount. It’s obviously different from metabolite to metabolite. But over the course of the entire pipeline of metabolomics this is going to massively improve your technical variability and that improves your statistical power – easiest and cheapest way of doing it.

Alice: Wow. And so for someone who is interested in improving their quality management, do they go to your paper or do they have other resources that they can learn from?

Jennifer: So I would say for quality management – I would never rely on one single resource. This is partly because I think we are still in the process of learning ourselves. But if you’re interested in quality management, our paper’s obviously a good place to start. We’re writing more papers. The mQACC consortium and if you’re really interested then go onto the mQACC website and apply to become a member.

Alice: Okay. We can put a link to the website in the show notes – then people can find it easily. Good. I think that takes us to your favorite metabolite. So you’ve actually already contributed your favorite metabolite last year, because you were so kind to speak to me in Valencia at the last metabolomics society conference. So I’ve asked you to come up with a new favorite metabolite and tell us why it’s so great. So what have you chosen today? 

Jennifer: My second favorite metabolite is melatonin, also known as N-Acetyl-5-methoxy tryptamine. But life is too short – so most of you will already know that melatonin, is produced by the pineal gland in the brain and it’s one of the mechanisms that we use for circadian rhythm and sleep wake cycles. It tends to increase in most people’s as it gets dark so that we are ready to sleep. One of the interesting things is how much is produced as extra pineal melatonin? The retina produces some melatonin. That’s perhaps not so surprising given it’s link to circadian rhythms. But so does the gut, the reproductive system, and certain immune cells in particular macrophages and mast cells. This becomes particularly intriguing when you look at melatonin’s other actions in the body. It acts as an antioxidant and it’s a fantastic three radical scavenger. It’s anti-inflammatory. It has a really interesting effect on the immune system because it can act on cytokines when we need to ramp up the immune system. It acts via receptors when the defense is no longer needed and it can actually also calm down the immune system. It’s also been shown at pharmacological levels to potentially be oncostatic and it’s now being explored as an adjunct in cancer treatments. It’s apotrophic. – And it’s even been shown by its own metabolite, AMK to potentially be a memory booster. It’s got all of these different mechanisms in the body and it’s produced in various parts of the body. And I think that as we explore the tryptophan pathway more and as we explore the gut microbiome more, we are going to find that melatonin starts popping up in some very interesting metabolic processes.

Alice: Thank you for this. It was really interesting. We discussed this the last time about tryptophan as well. I always find it really interesting that tryptophan is an essential amino acid. So we get it through a food and so the microbiome gets the first pick. If we think of now all the things that melatonin does we’re still a bit reliant on our microbiome to leave enough of the tryptophan to us so that we can have all these actions as well.

Jennifer: Well, I didn’t mention that melatonin is also consumed through the diet as well.

Alice: Of course. That makes sense. Then you can also take it in directly. Then we have more chances.

Jennifer: It’s also really interesting chemically because it’s an indole alkaloid and it’s an amphiphilic molecule – So it can pass over plasma membranes very easily. I think, we’re going to find more and more roles for melatonin and physiological processes. But I also suspect that it will start popping up as adjunct or even full treatment for certain conditions.

Alice: It’s really interesting. I also find it really interesting how metabolites, as we learn more about their functions, get completely different kind of personalities. If you hear melatonin until recently, you just say – „ah, it’s about sleep“. Now we learned with you that there are many, many other roles of melatonin. I like this a lot about metabolomics and about the study of metabolites that we realize they can do so many different things. The single metabolites has so many roles. It’s wonderful. Various groups have discovered that it seems to inhibit viruses entering cells and it’s been studied as a potential treatment for Covid 19. Again – Another use; and it’s antidepressant. I’d like to bring us back to biobanking because one of the things with melatonin is, of course, it’s light sensitive. So if you’re interested in studying this, then you need to think about how you are collecting your samples. And it’s also phasic. So again, experimental design. If you’re collecting it in different people, then you need to make sure that it’s similar time of the day.

Alice: Thank you very, very much. It was a lovely discussion and thank you for being on the podcast with us.

Jennifer: Thank you very much for inviting me.

Hannelore Daniel

Hannelore Daniel @TUM (German)

Favorite metabolite
lanthionine

Discussed paper by Krug et. al.
The dynamic range of the human metabolome revealed by challenges

Cited paper on sugar metabolomics by Mack et. al.
Exploring the Diversity of Sugar Compounds in Healthy, Prediabetic, and Diabetic Volunteers

Other resources
Talk on Gut Microbiome: Myths and Metabolites by Prof. Hannelore Daniel
(given at the 8th Munich Metabolomics Symposium – November 2021)

More about The Metabolomist podcast

Episode Transcript

Alice: Welcome to this podcast. And thank you for accepting this invitation. You studied biochemistry in the context of nutrition, from what I understand at the University of Giessen.

And in 1992, you became professor of biochemistry also in Giessen. After working in different places, you worked in the UK, you worked in the US in different research institutions.

And in 1998, you joined the Technical University of Munich as Professor for nutrition in physiology. And since 2018, you’re now retired. From what I understand, you’re a professor emeritus. Correct? But you’re still very active in the field and very interested in nutrition and biochemistry. – And what interests us today also metabolomics. So would you like to tell us a bit more. What are other points that you would like to point out in your career? And maybe tell us a bit about your research focus in the field of nutrition.

Hannelore: Thank you very much for this invitation to join this series of podcasts and this nice intro. I’m a nutritionist in background and yet I started early in my career with microbiome research in 1978 during my diploma project and treated rats with high concentration of antibiotics to bring the microbiome (at that time called gut flora) down and study the effect on nitrogen metabolism.

So that was my intro into the science world and I was fortunate to become a PhD student and could do a PostDoc et cetera. What has been guiding my entire career is to get a hand on very fundamental processes that relate to human nutrition.

So I was mainly interested in proteins that are important for uptake of nutrients in the intestine. Also for reabsorption of nutrients in kidney that you don’t lose essential amino acids. So I very early in my career started in cloning genes and trying to express these proteins in all kinds of weird organisms.

My peers all thought she is lost for nutritional science because I was doing such fundamental work. And, I also got interested in the genes and the genome and I was fortunate to have a microarray, which was very early. I could do my first microarray study in 2000, 22 years ago. I was fortunate enough to have money to buy equipment to do proteomics and buy equipment to do metabolomics.

That all was driven by very fundamental questions. What makes up human metabolism how do we get nutrients in and waste products out of cells. And that later became more holistic in that few of the entire human system. Not only in cell culture or in model organisms.

I have been working in yeast, I have been working in worms. I have been working in mice, in rats, in strange fish. And towards the end of my career, it was humans. And I did my first human studies pretty late in my career. And that was actually when I got interested in metabolomics and I saw what kind of fantastic tools do we get in hand now, to look in a more comprehensive manner into human metabolism.

Although that may be a little bit limited because we have mainly only access to plasma, urine, maybe some other body fluids.

Alice: There are lots of interesting points you’re making. One thing I noticed is, so is it right that then omics whether it’s genomics, transcriptomics, proteomics, metabolomics really changed the power of the types of studies you were doing? Or what did it change for your research?

Hannelore: It changed completely how nutrition science was perceived in the science community. It was such a privilege to have been in science when that happened. Because nutrition science was considered kitchen science. It didn’t worry about diets and body shape and things like that. So it was never considered to be real science.

Alice: And the omics helped with this?

Daniel: Talking omics – We were suddenly in the heart of all this life sciences talking omics.

Alice: Interesting.

Daniel: And it was fantastic. So that was really a game changer, for the entire science community that deals with food and nutrition and health.

Alice: And metabolomics specifically, what kind of relevance would it have then? I guess it’s particularly relevant for nutrition research. Right?

Daniel: Yeah, it is particularly relevant for nutrition research. And it has in essence two dimensions that made it so important. One is that you indeed can look into various metabolic states easily.

In the past we were limited to measure cholesterol, glucose, we measured with very specific means such as an amino acid analyzer – just 20 or 30 different amino acids. And here you look at hundreds of metabolites in a sample and what makes it even more attractive is that you can have tiny volumes and you can measure hundreds of metabolites in 10 or 20 micro liters. Fantastic. Absolutely fantastic. And even you can do that from a fingertip here. Put a drop of blood on a filter paper or little sponge and get hundreds of metabolites. It’s a dream.

Alice: It sounds like it.

Daniel: The second line is that we can measure metabolites that relate to the intake of particular foods. Where we can at least get a rough idea of what have volunteers, what have consumers really consumed?  Because when we do human studies and ask volunteers what did you eat in last week? They don’t tell us the truth. We know that. And yet here we have now a toolkit that allows by some of the characteristic components of individual food items that we can measure in blood or in urine of whether that food item has been consumed.

We’re not yet there that we can really give numbers in terms of how much was consumed. But at least if somebody tells us, that he doesn’t drink any alcohol at all and we found a ethyl glucuronite in urine. Then we know he didn’t tell us the truth. I call them not biomarkers. I just call them exposure markers.

Alice: That makes sense. But what would be your definition of a biomarker then?

Daniel: Biomarker has been defined by a clinical chemistry in terms of predictive markers in the health-disease trajectory. And if I find proline-betaine which is a marker metabolite for consumption of orange juice. It doesn’t have any bioactivity. It’s there. And it tells us orange juice has been consumed.

Alice: And it’s interesting the parallels you can make. Because of course you speak of exposure markers in the field of nutrition. But I come from the field of toxicology and we also have exposure markers. You see traces of the chemicals you’ve been exposed to – It’s the exact same topic. It’s there, it’s present. It doesn’t mean it’s having an effect or it’s a proof of the effect. But it’s a proof of it’s present. Yeah, absolutely. And you mentioned that you started your career with microbiome research. At which point did you reconnect with microbiome research?

Daniel: It came with this hype that came around the corner. I mean, yeah. Jeff Gordon’s paper changed almost the world. And it seems like you cannot get any paper published if you don’t include any microbiome analysis. I have seen trends in sciences for 40 years, coming and going. But I’ve never seen a trend like this microbiome hype like it is now present in all the areas of the life sciences and even other areas. 

Alice: I attended a talk that you gave a few months ago where you discussed some of the myths about the microbiome.

It was really interesting. For example, how we always hear that we have at least as many, if not more bacteria in and on ourselves than cells. It was really interesting to hear you speak about these myths and debunk this myths about the microbiome. Would you like to point to a few of them and maybe tell the audience. Which things they shouldn’t believe anymore?

Daniel: As I said, I have been working in gut functions throughout my entire career – From 1978 on. I would argue that I know the gut in the different regions from duodenum to anus quite well.

I could live with the knowledge that there are bacteria in the large intestine quite well for over 30 years. Because Nature, Science, Cell, New England Journal, Lancet – you name it. I have not read the word microbiome or gut flora for 30 years. And yet, science was explained to me at all levels. And now suddenly, the microbiome is brought into context of almost everything. That cannot be – frankly.

Alice: This contribute to some aspects, especially when you look at the metabolome, right?

Daniel: No doubt. If you ever have cut open a mouse or a rat and have taken out the contents of the intestine. You get a rough idea of how much is there. And I was even privileged to collect the samples from human tissues in ancient times. So I went to autopsy and you could get some samples. I was suspicious when I was reading that the human large intestine contains 1.5 to 2 kilo of, let’s call it biomass.

And it just can’t be. It was actually old literature that used antique technologies, a balance and it took out the intestine from sudden death victims, collected the content and put the organ per se and the tissue and the content separately on the balance. [The scientists] came to the conclusion that with large variability, young large intestine had a total volume of contents of roughly 230 milliliters. So large variability, no doubt. But the mean, or the median was 230 milliliters. They put it in a freeze dryer and put it again on the balance. And there were 36 grams of dry matter left, of the entire human colonic content.

And yet the literature was full of this 1.5 – 2 kilos, and, funny enough, there was a series of papers, by Milo and Sender. They traced this information on the volume or the weight of the biomass in large intestine. And at the same time also with respect to the numbers of bacteria because it was claimed that the human gut microbiome contained 10 times more cells, bacterial cells, than human body cells. So Milo and Sender traced it back to a study or a review published in 1972, by Luckey in American Journal of Clinical Nutrition, taking the number of bacteria in a colonic sample and protecting this number to the entire intestine – Not accepting at that time, that there is a huge gradient in terms of the bacterial density. So the numbers of 1.5 to 2 kilos, all originate from this one paper with a misconception about the density of bacteria and different regions of the intestine but it made it into Science and Nature and all top notch journals and was quoted and quoted and quoted. Then came Milo and Sender a couple of years ago and said that all is wrong. They did a very careful analysis of all the old literature and came to the conclusion that the total weight of bacterial biomass residing on and in humans is likely to be something like 200 grams.

Alice: Big difference. And without making a bad joke, it takes a lot of guts to write that paper to that goes against everything that everyone is saying. So it’s really brave.

Daniel: That’s why always emphasized the great work they did. And I strongly recommend to look up those papers.

Alice: Yeah. You find the link in the shownotes. This is really interesting resource for people who are interested in microbiome. Thank you. Are there other aspects of the microbiome you’d like to discuss maybe myths that you really want to fight?

Daniel: I know it’s not a myth that. I’m just very credible when it comes to mouse work. I have been studying mice. I’ve been studying rats. I don’t know how many animals went through my lab. I sometimes feel so sad about that. But you know, the mouse is not a little human. The mouse is still a mouse. And the black six mouse, whether J or N that is mainly studied is also only one mouse strain, of literally hundreds. And even if you measure some very fundamental things in these different mouse strains, you can see, they are so different.

It’s just spectacular. And we studied mainly one strain and we explain the world based on one strain. One mouse strain is like studying more or less one human.

And would you conclude that all humans are the same if you study just one?

Alice: No.

Daniel: A human is a human, two arms, usually, and two ears and things like that. So in this respect, yes, mouse is a mouse.

But for example if we look into the abdomen of a mouse you see a real fermentation chamber. That is a cecum. The cecum in humans is the appendix. It is a tiny thing. We have a colon that is structured in three regions. It’s completely different in terms of anatomical size and also in ultra-structure and the mouse cecum and large intestine relative to body mass, is almost twice the size of what the ratio would be in humans.

Alice: Are these things that can be compensated for with bioinformatics/ modeling? Or there’s no point in going there you think? (because it’s so different that it would be like comparing apples and oranges).

Daniel: I think it is even more complicated. Two things.

1. There’s a very nice paper that came out that struck me because I had been asking a number of colleagues of mine: Whenever you have a large fermenter in a biotechnology company. One of the biggest problems is get rid of the heat produced by bacteria. So I was asking myself of whether the gut microbiome in particular in this rodents contributes to body temperature management. Hard to measure.

Couple of months ago a nice paper was published. Doing exactly that and measuring the contribution of the bacterial heat that got produced in the cecum of a mouse and yet it is a substantial amount of total energy turnover. But that should not be used as a conclusion that in the human intestine heat production by the bacteria contributes in a substantial amount to overall energy expenditure. Again, organ mass in the mouse relative to total body mass is twice of what it would be in humans.

• And moreover, we feed all those rodents that are sitting in facilities diets that are prudent by any means because they are not cooked. The starch is raw. They are either irradiated with Cobald 16 (that is dry irradiation) or autoclaved, that is also dry heat. The starch is not swollen. Everything is raw.

Alice: Very different.

Daniel: Very different. And if you cut open the cecum of a mouse that just receive chow diet. You see the big chunks of the chow in the cecum. So I think we’re not taking into account many of these factors whenever we look at mice as a model for microbiome-host interaction. I’m not saying mice are useless.

Alice: You have to be aware of the limits of every model. Like every model has its limits and this is also true for mice.

And this is also true when we work with humans. As you said, you can make very interesting studies in humans. But if they lie about what they eat even without meaning it. If they give you false information that can really then make it difficult to use the data or to interpret it correctly. These are really important points yes. Were there other points you would like to discuss about the microbiome.

Daniel: Well, in terms of the reference, we are talking not about that gut microbiome and we talk about the stool microbiome. Only a very few studies, but this will change that’s for sure. Look at different sections.

Alice: So that means you have to take the animal and then…

Daniel: I’m talking about humans.

Alice: Are you talking about humans. Okay. How do you get…

Daniel: You can use tubings that you go in either from the rectal side (that’s what gastroenterologists do), and you can also use nasal gastro tubing. And you can go through these small intestine and collect at different sites. And little devices that can collect samples as they are traveling down the intestine. And you even can then identify where it is located and can activate it to collect a sample.

Alice: Wow. I didn’t know that this existed. This is really interesting.

Daniel: I mean, it’s not standard now yet. There are more studies now. Why is that relevant? There are some studies that have been doing that already showing that the microbiome is different in different regions of the small and large intestine.

And even in the large intestine, whether you go into the colon; transversum, there are differences. Some of the differences in comparison to stool are remarkable. So yeah, you have a spacial microbiome if you go from proximal to distal and the stool just does not represent the entire diversity that you find in small or large intestine. So that’s not only regional;  it’s also then if you go radial, which means from the lumen of the intestine towards the wall.

Because people think that the bacteria all the time in contact with the epithelium. That’s not correct. We have a mucus layer that in the large intestine, is between roughly 500 and 900 micrometers thickness. And it has an inner layer, it’s produced from the mucus proteins producing cells.

It’s a pretty sticky layer and the inner layer, is a mesh that is almost tight. So there are no bacteria, if you stain the inner mucus layer. You don’t see any bacteria, if you have a damage intestine for example in Crohn´s disease. It’s looking different. But in a healthy individual, this inner mucus layer is in essence sterile.

The outer mucus layer is a bit fluffy and bacteria can be found in there. Yet the density of bacteria is much lower than in the lumen. And at the same time you have, because a tissue is nicely oxygenated, you have oxygen gradient from the tissue surface into the lumen. Which means for all the anerobic bacteria, this mucus layer region is horrible (because it has a certain oxygen tension). So in essence, it would kill the bacteria. Nevertheless, bacteria that can tolerate certain pO2 levels may well sit there. So I’m just saying, even in terms of density, and in terms of the different species, if you look into the substructures (radial and longitudinal) microbiomes are not identical.

Alice: So there are many dimensions that most studies don’t look into.

Daniel: Right. So, and then we take usually one stool sample. We sequence (shotgun or whatever you do) and then we start telling stories. We even report entire movies by taking one snapshot. Be careful. Because it’s well known, that the microbiome beyond all the technical problems in terms of standardization, there’s a large variability in how often people go to toilet. The total volume of stool, the color and consistency of stool, the frequency of stool emptying. And that all has been shown to affect the microbiome in terms of the diversity of the bacteria or relative abundance of bacteria.

Those information would be essential to give some meaning to the data is usually not collected.

Alice: And so to get this wealth of different dimensions in space and also in time would be one way to bring stories that are more meaningful about the microbiome

Daniel:  A single snapshot cannot deliver any information.

Alice: Yeah, of course it’s similar to metabolomics. It’s similar with metabolomics and this is the topic of the paper we wanted to discuss together. That’s in a study, like the one we’re going to talk about now: If you have a single snapshot or if you have 65 different time points with different challenges and changes over time.

You can tell different stories and you can go more in depth into what’s going and how the system is changing over time. So maybe we go now to the paper, if you don’t mind. Let’s discuss this. So for the audience the paper we are going to discuss today is a paper entitled:
The dynamic range of the human metabolome revealed by challenges.

It has five first authors. The first of them is Susanna Krug. It was published in 2012 in the FASEB journal. And it’s a really interesting study. You suggested this paper for us to discuss and I think people will learn a lot also about designing ambitious studies for metabolomics or for any omics. And how to interpret the results, how to organize and structure your work when you have such rich data. I think there’s a lot we can discuss. Before we look at the results, I just want to explain a bit the study design and also to ask you, how fun that was to design? Because it looks like an enormous amount of work.

Is 15 human participants you asked them under supervision to go through six different challenges in different series over four days. It was two times, two days and it started with fasting. And what I really like is that over the 36 hours of fasting there were eight or nine collection points already. So you didn’t just let them fast and then look before and after you really wanted to see the dynamics over time. Then you had different typesof challenges related to diet. So the type of liquid diet, glucose tolerance test and lipid tolerance test, and then an exercise challenge with cycling for 30 minutes and finally a stress challenge where people had to put their hand in cold water for a few minutes. Which can elicit metabolic responses as well.

And so besides the many time points where you collected samples you didn’t just limit yourself to plasma. You created plasma, urine and then breath gas or the breath of the participants and EBC, which is exhaled, breath, condensate I think. And you performed different types of metabolomics on these different materials and then did all sorts of analysis that slowly, progressively take us to what might be going on and also how it’s going on.

Before we look at the results and the interesting findings in this, can you comment on the organization of this study, the planning, how much of a challenge was it? Because it’s really ambitious.

Daniel: It was demanding  and it brought a number of emotional responses. We did it with a number of colleagues and said, you know, how can we at best get a picture of this?

Alice: Did you try to do everything in one study?

Daniels: Yeah. The idea was to use volunteers that are as homogeneous as possible. I mean, it would’ve been nice to have 15 clones.

Alice: It’s yeah. You can see like the sex is the same, the BMI similar, the age is similar.

Daniels: Any means we did super phenotyping. We put them into the sports facility to measure. How they perform? Then we housed them for four days in a cabinet. If you like.

Alice: You decide what they eat, you decide when they sleep, you decide everything.

Daniels: And they had catheters sitting in their arms while sleeping; and the exercise was really demanding as well. It was 75% vO2, which meant some of the youngsters that were not so well trained lost about the liter of sweat. Anyway, it was fun. But it also was a huge exercise to find an agreement amongst the experts of, you know, what can we do and what can we learn if we do that?

Alice: Then, so you perform this experiment, collected the samples, measured them, and then there’s a huge bioinformatic study that goes behind it. So you first you look at the data and then you do correlations to see what moves together and so on. And then you realize, okay, the acylcarnitines and carnitines are changing significantly. There’s probably something going on with beta oxidation.

And then you make a mathematical model of beta oxidation. If you can be totally honest about this, did you expect this to happen? Had you already planned to do this or did you really let the data tell you where to look?

Daniels: The latter. So let me, put it in a perspective. We thought we have the most beautiful metabolomic study done on the planet. We submitted to top notch journals. And they all didn’t like it. So what’s new here. It was shocking. It was really shocking. It was so frustrating.

Alice: Because it’s not just descriptive. Like you really show this dynamic range that you put in the title. Like you really show that something is happening.

Daniels: I mean it was only 15 people. And we know, we know, we know. We probably should have included microbiome that it probably would’ve flowed.

Alice: In retrospect that would’ve been simpler.

Daniels: So in the end, it ended up in FASEB. But we have a repository and we are still working, on the data. Because we have metabolites from, I don’t know, six different platforms. NMR; non-targeted, targeted, GC-MS/MS, LC-MS/MS; Two commercial platforms, two homemade platforms. And then PTR-MS for breath. A wealth of metabolites – roughly 800 for plasma in each sample. And our friends at the Helmholtz Center, particularly Gabi Kastenmüller (podcast episode with Gabi) are still working on the dataset. And make the entire dataset, which is already in the public domain more easily accessible, that people can study under which conditions for example, do we see an increase in metabolite X or Y or Z.

And I’m positive there will be two or three more papers now coming pretty late from the study. But in the end, I’m still excited in every time I look back into these profiles. There is something that catches my attention. I did this flat before outcomes.

Alice: Yeah. And this, I think, is a really interesting point to mention – I think it’s a bit of pressure you have even when you have a single data set of any omic (Metabolomics or other) you always have the pressure of saying, did I exploit this to the fullest? Did I find everything? And the answer is almost always “no”. And that’s okay I think. Because, of course, here you have an enormously rich dataset. But there are always things that you can only understand later or things that you can only focus your attention on now. And then you will look at the rest later.

And thing is an important thing to mention to people, to take off a bit of the pressure and you can always publish papers a couple years later or even 10 years later. It’s still relevant. As long as you have an up to date analysis of it. Why not?

Daniels: I can tell you, I got all the profiles for the individual metabolites (And the 60 samples over the four days of all the volunteers). Which means I had about 800 different traces. And I was sitting in planes and trains and just push the button from metabolite to metabolite, to metabolite, to metabolite. Just look at the profile.

I’ve seen this profile before. So I went back and I could recognize patterns easily. So I found four or five metabolites that show exactly the same profile. And how are they interconnected, linked to each other. So that was and still is an enjoyment.

Alice: That’s good to hear. And so could you tell us in a few sentences, what for you are the biggest messages from this paper? What did you find out? That was so interesting?

Daniels: Well, first I would argue that a single fasting blood sample is not enough to tell a movie.

Alice: And this is what we recommend also in the clinics, when you do tests and stuff. It’s all like, take you’d come fasted. We take your blood, then you can have a sandwich and go home.

Daniels: I mean, the question is, who decided that the morning blood sample after an overnight fast is the reference for everything on the planet. And that is one of the stories, because we were feeding the volunteers the night before a highly standardized real food. And what is really funny, because we were then putting them into this 36 hour of fasting. – We have now the complete washout, if you like of the food derived metabolites. And that is fantastic. And I’ve never seen that before. So you have individual metabolites that were ingested (it was a chicken meal with some veggies, garlic and onions and some other spices). And you can identify the metabolites derived from garlic and onion and you even see the entire kinetics.

And if you do a log transformation, you see they all follow first order kinetics. And I’ve never seen that before for so many metabolites. And I think that is a treasure. Because it also allows you to measure elimination half time. And you can ask why do different metabolites have different elimination, half lifes?

And I can give you a nice example: Methylhistidine derived from the chicken meat. Possibly first order kinetics and pretty fast eliminated. For other metabolites you have a completely flat elimination rate. Which means they’re probably protein bound and go around for much longer. Because you also see and in it, since we collected 25 urine samples. You see in essence the same pattern then with the appearance in urine. So I’m just saying that was the first learning curve. The second learning curve, 36 hours of fasting is catabolic state per excellence.

You see the push on lipolysis. You see the push on beta oxidation. You see the ketone bodies going up – So everything like in textbooks. There were two volunteers and particular one volunteer. It looked completely different. In essence you would argue: “he must be dead”. That is not possible that somebody can fast for 36 hours and not showing any substantial increase in ketone bodies.

But we could be sure that the guy was not eating anything ‘illegally’.

Alice: Because you are monitoring them closely!

Daniels: The other metabolites were down. So he must be in deep fasting.

Alice: But he didn’t have this capacity to produce ketone body.

Daniels: He is a mystery. And usually you would say, hey, we have 15 volunteers and that is a outlier. And you know, I don’t have to tell you, there are smart approaches to get rid of these outliers. Life sciences got rid of all the interesting phenotypes.

Alice: Here you want tosee this outlier. You want to ask her “why are you different? Tell me!”

Daniels: Exactly. That was discouraging. But he survived the four days, like all the others. But I cannot tell you how he managed to look so different. And what he relied in terms of his metabolism. I mean, he must have had tons of glycogen in liver.

Alice: This is a possibility.

Daniels: That is not easy to measure, of course, not too many people can measure glycogen non-invasively by MRS. I’m just saying that it was a learning curve as well. Coming back to acylcarnitines. I heard of the acylcarnitines because I was teaching biochemistry to students for 30 years of my career. So you have for the fatty acids, you have a CoA pool in the cytosol and you have a CoA pool in the mitochondria. And you have the inner mitochondrial membrane which does not allow CoAs to be shuttled there. – You need the carnitine to take the fatty acid, cleave off the CoA, put the fatty acid on the carnitine. Then you have a shuttle protein that translocates the acylcarnitine into the mitochondria. And then you reverse it and you hook up again CoA.

I never would have expected that in essence, every single fatty acid from the acetyl to the most long chain, appears as a carnitine derivative or conjugate in peripheral blood. They are all there?

Alice: This is a question always – Why do these things are supposed to be at the heart of the cells? What do we find them in the plasma? This is often a question.

Daniel: Because it’s a water soluble form of fatty acids. Fatty acids as fatty acids are nasty molecules. They go into each membrane you cannot control that they move in, put in their long hydrophobic tail and screw up the entire structure. So you better protect the cell from an overflow of free fatty acids. And that is possible, if you make them more water soluble by hooking up a carnitine and get rid of them. So we use the extracellular space – the plasma as a distribution space.

Because beta oxidation cannot be increased if you increase lipolysis. Because lipolysis goes on in catabolic state. But at the same token your production of ATP is limited by the reduction equivalents. And that is not very well coordinated. I don’t know why nature did not better connect the capacity of lipolysis to beta oxidation and citric acid cycle and respiration.

Alice: But there must be a reason cause usually there is.

Daniels: So if you have this overflow of fatty acids. You can expand the volume of distribution with the acylcarnitines. If you have a limited capacity for beta oxidation, even in mitochondria, you can get rid of the medium chain and or short chain derivatives. You put them out into plasma again. It is, in essence, a protection mechanism. It protect cells from an accumulation of fatty acids and of all the CoAs that cannot be on time be oxidized via beta oxidation. What I’m really interested in is which protein in the plasma membrane mediates this import-export business.

Alice: Back to the transporters.

Daniels: Yes. Back to the transporters. I mean, we know how it is done in the inner mitochondrial membrane which protein in the plasma membrane. And it must be then in adipose tissue. But also in muscle cells all over the place in essence. But it looked over the four days, we always saw this reciprocal change in acylcarnitine over free carnitine. So whenever the acylcarnitines increased in plasma, the free carnitine went down. It was a mirror like behavior.

Alice: It’s a beautiful figure. It’s also a really nice starting point for interpretation. It was really well made.

Daniel: That would argue in a naive manner that it is an exchange mechanism at the plasma cell membrane, as it is an exchange mechanism in a mitochondrial membrane: For each acylcarnitine put out, a free carnitine is taken up. And I would explain in a naive manner, at least, how this mirror like behavior can be seen? That was a nice piece of biochemistry, which I hadn´t seen before in this beautiful manner.

Alice: And then you show the variability between the participants increases as you put them under stress, right? This is another one of the findings of the paper.

Daniels: So we had this exercise built in – 30 minute on a bike at high force. As I mentioned, some of the guys really sweated. We were collecting blood sample before and after 15 minutes and after 30 minutes and then they stepped down from the bike. But we collected further blood samples. And there were of course metabolites that only show up in the exercise. First baseline, then -boom- going up, and -boom- going down back to baseline. That’s amazing.

That’s only for this two time points, amongst the 60 blood samples or time points. We had this huge increase and of course, if the first one that jumps to the eye is lactate. No surprise. But going from figure to figure what’s going on here! I found fumarate, I found oxalacetate, I found the entire citric acid cycle in plasma. It should not be in plasma.

Alice: So what is it doing there? Do you know?

Daniels: The exercise was so harsh. Just that they disrupted muscle cells. So they washed in essence  cellular contents into the plasma.

Alice: Yeah. Makes sense.

Daniels: And the strongest evidence was on one of the platforms – We had cAMP (cyclic AMP) as a metabolite and of course you wouldn’t expect cAMP to be in plasma. But here in the exercise, it suddenly popped up in plasma. Including some phosphorylated intermediates that should not be there either.

So exercise can be modest. Exercise can be harsh and that was harsh exercise. And obviously, the best explanation is that you have a real rupture of muscle cells.

Alice: Which is expected with intense sports. I mean, it really makes sense.

Daniels: So that was beautiful.

Alice: It is, there are lots of beautiful things in this and I’m happy to hear you still working with this data. So we probably get more beautiful papers out of it.

So it’s good to know.

Daniels: I hope so. 

Alice: So I think we’re already at the end of the hour. Do you have a favorite metabolite and why?

Daniels: I do have various favorite metabolites in particular those where you don’t find too much in literature. 

Alice: The interesting one.

Daniels: I give you an example. We were feeding the volunteers the standard liquid diet. Which was in essence a commercial product, used in clinical nutrition. In people that have impaired digestibility of nutrients.

So it’s an enteral nutrition solution. To bring in calories, to bring in all the essential nutrients. And of course it is defined – No other food item is so well defined as this kind of liquid diet and you can put it in a freezer. And you can use the same batch of a liquid diet a year later and challenge the volunteers again.

That was actually the dream that we do the whole thing. Then in a consecutive manner, yeah know, use the same volunteers five years later.

Alice: But they all said, no?

Daniels: No, no. They’re all over the world now.  Got to know one last year. Made his career in industry. So coming back. I think it was a good idea. Because it is really specified in terms of all the fatty acids and the micros, the vitamins, et cetera. When we looked then at the appearance of all those metabolites. I found lanthionine and I’d never heard of lanthionine. And I don’t know where it comes from! I know where it comes from. But I still didn’t know whether it is, or it was contained in the liquid diet or whether it is produced.

On its way from the intestine into circulation and lanthionine is a strange molecule. It’s a non-proteinogenic amino acid.  It has two carboxyl groups. It has two amino groups and it’s connected from with a sulfur atom that comes from cystine. So it must be a cystine and serine fused.  

Alice: Interesting metabolite. You mentioned, you’re not sure it came from the liquid diet. Isn’t that? Isn’t the composition controlled of this liquid diet. You should know, shouldn´t you?

Daniel: It’s well controlled. But you don’t know whether it is produced by heat. Process, things like that. Because I can also, I think that is a learning curve as well.

We did an oral glucose tolerance test. Like millions of oral glucose tolerance tests have been done. And then I found metabolites popping up in plasma where I said that just cannot be. I mean, it’s pure glucose that is administered. And I learned that our people didn’t use pure glucose.

They went to a pharmacy, they bought a commercial OGTT solution. That is used by physicians for diet.

Alice: Which contains preservatives, flavorings and whatever.

Daniels: Yeah. Shocking enough. There was ferulic acid in there. There was hippuric acid coming in. Because hippuric acid we think microbiome, microbiome, microbiome. And there was benzoic acid used as preservative in this commercial OGTT solutions. So that was a learning curve as well.

Alice: That it also shows yeah. That you should really be careful.

Daniels: OGTT solutions. That you can buy either in the US or in Europe. Of course they contain glucose. But they also contain other things – be careful.

Alice: Especially when you use methods that are so sensitive to whatever is in it. At least you should know what’s in it. But if you can remove some of the noise it’s even better.

Daniels: Yeah, absolutely. So coming back. I don’t know whether the lanthionine is produced by heat treatment to preserve this liquid diet.  I was not familiar with the chemical structure. I’d never heard lanthionine before.

And let me come to one point that I would like to make or two points. I’m following now, metabolomics for 15 years. And I’m surprised it’s still almost all the studies report only fold changes. Metabolomics the technology allows to get real concentrations. I know that it is not trivial to do whenever I talked to young people and said, why did you not go for a quantitative analysis?

The answer was – it’s difficult. They said, yes, that’s your problem. You’re born too late. All the simple things I’ve done. Take on the challenge. So I would love to see more quantitative measurements. And however, if we go into clinical chemistry and you get your blood analyzed, you get such a leaflet with a hundred things measured in millimoles in milligrams per liter units.

If metabolomics wants to get more into diagnostics I think in the end it has to come with quantitative measurements. And my second point is the methods measure what they can measure. Each technology has limits and advantages.

It’s amazing how the number of metabolites has been increased over the years with combining platforms from NMR to GCMS et cetera, et cetera yet. I would love to see some more targeted platforms that go really into metabolic pathways and try to cover as many metabolites in a known pathway. We usually have a substratethat goes in. Maybe one intermediate and maybe one product or another intermediate, make ratio. And we draw conclusions and explain our data.

Alice: I guess it’s particularly difficult cause the structures are very similar I guess.

Daniel: We have been developing in a joint effort. Semiquantitative – It’s now on the way of being quantitative for sugar and sugar derivatives. Because I don’t have to tell you how you report the sugars!

Alice: You don’t.

Daniels: Yeah, because you have so many isomers and they’re hard to distinguish. Carina Mack has a developed a GC method and she ends up with about 50 sugars and sugar derivatives in urine and about 40 in plasma.

Alice: Nice.

Daniels: Sugar, sugar alcohols, sugar acids. And amazingly disaccharides in plasma and urine. If a student would tell me that sucrose is absorbed and appears in plasma, I would’ve said, no way. The sucrose is cleaved and the glucose and the fructose appears. No! The sucrose appears in plasma, who may choose it as well.

And sucrose appears in urine and lactose appears in plasma and lactose appears in urine. So many disaccharides, many other monosaccharides (trehalose, mannose, sucrose) – you name it. So, the world of carbohydrates in plasma or sugar and sugar derivatives is completely underreported.

We also have a paper where we demonstrate that there are some sugars that behave like glucose. You can see the same signature of a pre-diabetic state and a type-2 diabetes. It’s not only glucose. There are other sugars showing the same profiles. And they may be even more useful than glucose in predicting type-2 diabetes.

So that is an example where I say, you know, it would be good to go more into the study of these metabolites.

Alice: So if you were to push the development of metabolomics you would go in that direction at the moment. Thank you very much.

Karsten Suhre

Professor of Physiology and Biophysics

Director of Bioinformatics Core at Weill Cornell Medicine-Qatar

Suhre Lab @ Weill Cornell
Virtual metabolomics core facility

Discussed paper by Gieger et. al.
Genetics Meets Metabolomics: A Genome-Wide Association Study of Metabolite Profiles in Human Serum

Karstens blog about metabolomics, genomics, and where the two fields meet
http://www.metabolomix.com/

mGWAS server (collaboration project) 
http://metabolomics.gwas.eu

Other resources discussed in the podcast in github

Sign-up for The Metabolomist e-mail list

Episode Transcript

Alice: Welcome to this podcast. Thank you for discussing with me today.

I will start by introducing you with a short biography and then we can discuss your work with metabolomics in further detail. It was really interesting to look through what you did over the years. I was quite impressed. Looking at all the different things you’ve done was very eclectic.

So you studied mathematics and physics at the University of Osnabrück in Germany. Then suddenly you had the PhD in atmospheric chemistry and meteorology.

Karsten: Yes, but that wasn’t suddenly. During my studies I was half a year in England and studied fluid dynamics there. And then fluid dynamics led to metrology. Then the specialization in metrology was atmospheric chemistry.

And in the end, honestly, it’s not very different from what I do today. Atmospheric chemistry is like metabolome of just one organism, which is the globe.

Alice: That makes sense. So it’s all about the analytical methods. I get it now.

Karsten: Well, I could pretend that but in reality it’s a lot of personal reasons.

Alice: So did your PhD in Toulouse, France. And then a few years later, you habilitated in bioinformatics and structural biology in 2004, at the University of Aix-marseille.

You stayed in France for a few more years until in 2006 you were appointed professor for bioinformatics at the University of Munich and at the Helmholtz Center in Munich, Germany.

Karsten: Yeah

Alice: And in 2011, you joined to the department of physiology and biophysics at Weill Cornell as a full professor and became the director of the bioinformatics core at the Cornell campus in Qatar.

Karsten: That’s right.

As I was reading through your profile page on the Cornell website there’s one sentence that caught my attention. You wrote, “I identify as a bioinformatician and system biologist.” And if you’re remember writing this, I would like to ask you first. Why did you choose that verb?
Why do you identify as a bioinformatician? I mean, you’re a professor in bioinformatics. I found this really interesting.

Karsten: It’s because in a way I never really like to be boxed in somewhere. So when I was a physicist, I always said “no, I’m more, theoretic physicist and a mathematician”. I’m into computation stuff; always evading a the point I arrived somewhere. And so I said, because being a bioinformatician is also some kind of the definition of bioinformatics also changes a lot between countries.
For some people it’s just the computer guys for others it´s more the interpretation. I like is it’s basically the study of information content in biology. And I think that’s the definition that the NCBI, gives on their webpage. And that’s something that I identify with.

Alice: Something that fits: I love this choice of words. Because it’s also nice to see: You get the degrees, you get the job titles, but still you can still identify as whatever you like to identify. And then another thing I found interesting going through your bio is that it changed quite a lot.

Karsten: Driving this was the curiosity of how biology actually works. Because being a physicist initially when I studied biologists were frowned upon. And when there was the human genome sequenced. That changed everything. Suddenly biology was really mechanistic and much on front of it. I changed totally about this view.

Alice: So this is what pushed you towards biology then also. When the genome was….

Karsten: I mean it was a chance event. I was with CNRS. I could not go back. So I went to industry in my hometown for personal reasons. Did engineering for two years. And seeing engineering and research is not the same thing. The one guys know exactly what they want optimize things and the researchers don’t.
I didn’t fit in there. Luckily I was in CNRS and could go back. Theoretically, I would go back to atmospheric science. Just by chance I ran across this kind of paper of on in Le Monde on the discovery or the publication of the human genome. And when I was just applying for position to go back to the region of Marseille.
I said, well, maybe I just asked him. And it turns out that, in his lab all bioinformaticians were physicists, astrophysicists. He was a physicist. So at that time basically there were no bioinformatician. They were all converted biophysicist. So they were really the driving thing.

Alice: That developed later then. The job of the bioinformatician. That label didn’t even exist probably at the time or it was really a small niche thing. Or…

Karsten: Yeah I mean it was really just the discovery. The problem of really solving the sequencing problem. They had all the sequencing, but having the alignment was what bioinformatics initially was.

Alice: Hm. Genomics is still a big topic now. But metabolomics is a topic that’s grew a bit later. It seems like for the last 15 years you’ve been quite interested in metabolomics.
And that this still keeps you interested. Is there a reason for this?

Karsten: Yes. It’s the functional thing. I started biology or bioinformatics in 2002, when I went to Marseille and learnt it from scratch. And already then I had a colleague who was actually the first author on the first paper of the KEGG metabolomics pathway map.
And he was working in Marseille that point.
With him, I learned to analyze the metabolic capabilities of bacteria. So we sequenced different bacteria and the way of looking how intracellular bacteria live. So the kind of genes they lose, they become dependent on the host cell. That you could computationally analyze that he was really there.
You just took the million base pairs of whatever you had in the bacteria and broke that down in all the enzymes. – Then he could show which ones are missing.
And which bacteria does use what from his host. These were all obligated bacteria. That was already metabolomics in a certain way. Although we didn’t know it at that time.

And the other thing is the link between the function of what the genome really does? How does it really function how do you interpret that? And I think that these were the first steps to interpret function in the bacteria. And that later carried on when we went to Munich – Where the things got into humans.
Working on genome-wide association studies and then doing the GWAS with metabolomics. I think I was lucky to be at the right place, at the right time, with the right colleagues around me and everything.

Alice: Then this started a series of really nice papers. This is the main thing I wanted to discuss with you today. I’m not going to discuss the detailed statistics behind it.
I think it’s interesting for people to have a global understanding of how this is done. Because genomics and metabolomics are two very different data types. So it’s interesting to discuss this. But also to see what metabolomics brings in that environment.
Maybe I can give you my very superficial view of it.
And then you can give a bit more detail. Because one of the most interesting things for me was that there seems to be people who start from the genomics and then the SNPs. And then put their phenotypic trait, whether it’s metabolomics or something else on top of it. And look at where the associations are.
And there seems to be another technique where you start from the phenotype. Whether it’s metabolomics or something else. And then you find what’s interesting there.
And then you find what is associated with it in the genome. Did I understand this correctly firstly and can you comment on this?

Karsten: I think you should go maybe further back about the GWAS. I mentioned already the human genome project being at the beginning of everything.
And the human genome project was already showing that there are genetic variant between people and it wasn’t the genome of one person (but rather at least 10 different genomes were put together). The promise of the human genome project was – we just go and find all the variation and then we say that is the outcome.
And once we have that, we can treat everything. And of course, things turned out to be much more complex; some people say it didn’t work. (I would not agree with that. Absolutely not.) But some people say it wasn´t worth it? In any case, GWAS came and thought initially: “We just find the gene for diabetes and the gene for this and that”… but realized it’s not true.

You just can explain one, two, three percent of what you suspect to explain of the heritability. Nowadays, we know much more about variants and of lots of small effect sizes with many variants and some rare events with the larger effect sizes.
The one other thing that came as a question what do we do with GWAS? What’s the purpose? And there is the misunderstanding that people think you can predict something. You can`t predict what someone dies off this or gets this disease.
Today, it’s much more about understanding on the one in the pathways.
And, I think from a pharmaceutical point of view also target validation or target interpretation. For them, it is important if they want to work on a molecule inhibiting a certain protein, that there’s genetic evidence that if they tinker with this protein, something happens. – And that’s where the metabolomics comes in.
This concept of the intermediate phenotype.
That has been something that has been pretty early on in our first GWAS presented by Florian Kronenberg from Innsbruck brought it up. It was the hypothesis that these intermediate phenotypes are the link between the genome and the disease. And it’s not only metabolomics.
And it’s also, I think there’s no priority for metabolomics or any other Omics. The whole chain from the genetic variance, which is the starting point of the GWAS and the end point, whether someone gets a disease. – And in the meantime is all the other genetic variation that’s influenced by a particular genetic variant. You can measure that in metabolites and proteins, GWAS on glycosylation, on lipidomics, on micro-RNAs on anything you want.

Alice: Yeah. And this grows as the techniques grow as well. Like as the technology improves for glycosylation or things like this, then it’s more applicable to associate with other omics, I guess.

Karsten: Right. And that makes the value of metabolomics. Metabolomics is measuring the true end points of biological processes. Which has maybe also a little bit of an exaggeration whether they’re the true end points. But it was like you go from the gene to the RNA to the protein and then the metabolites where things happen. – Of course diabetes is glucose and gout is urate. So in the end, it’s true. But there’s so many other things as well.

Alice: Yeah, but this is the old paradigm as well. That’s more and more we see interconnections between the different levels and then it’s even difficult to see the levels anymore. Because everything is everything now. Is it correct that in your papers you start usually from the genome and then add on the other methods or at least the metabolomics. Or do you do both versions? Is there any advantage one or the other?

Karsten: No, it’s always together. Association study means one is the dependent and one is the independent variable. But that’s just artificial. Which you take as dependent or independent variable. It’s in the end it’s correlation.

Alice: Yeah. But let’s say if you’re looking at cohorts with diabetes or without diabetes. Then you might use one dataset to focus on the differences between those two groups and then add on the second layer. That would change some of the results. Wouldn’t it?

Karsten: Yeah. But that’s a different kind of study. When you do metabolomics, you should distinguish what you’re studying. Now we started out with GWAS and GWAS from metabolomics. Because that’s a bit what I was focusing on. But in general, you don’t do metabolomics for genetics.
There are two kinds of studies. The ones are this population studies. And that’s a bit where we were working on like the KORA or the Framingham’s, UK biobank. So where you take everybody who is more or less normal. And then you collect as much phenotype information and genotype information everything and then you can do all against all.
And on the other hand is more of this kind of clinical studies. E.g. “I want a cohort of diabetics and see what’s the case control” or “kidney rejection” and things like that.

Alice: That`s make sense.

Karsten: And in terms of data acquisition that also makes a huge difference. In the one case (population studies), it is generating this data once and for all in high throughput for thousands of samples. In the other case, it’s more like you spend less money because there’s less samples. But you enrich them in cases and you have more detailed phenotyping. You generate much more detail on the patient metadata. And in an ideal world, like with UK biobank, you have both. You have so much detail on them that you can do whatever you want. And then you just filter out like the diabetics and non-diabetics from UK biobank and do your case control study or similar kinds of things.

Alice: Okay. So let’s go back to the first mGWAS paper (by Gieger et al.). Could you tell us a bit about that paper and maybe your role also at the time? I’m interested in the people who do the interpretation and in that paper there is something really interesting that is the use of these ratios of metabolites rather than the association to the pure metabolites. Can you say, how you contributed to that paper or to that story and then tell us a bit especially about the ratios. I’m really interested in that.

Karsten: Yeah, I’m pretty fond of that paper. Because it goes a little bit with my move to Munich. – And there I was on a professorship on bioinformatics and it wasn’t really clear what I would be doing there other than teaching bioinformatics. I went to the Helmholtz Center at the time still called GSF. And that’s an interesting place because they host the KORA population study. There are different institutes involved. I was at the Institute of bioinformatics but there was an Institute of Epidemiology where Christian Gieger was working.
And also his colleague, Thomas Illig was involved in that. And then there was a core facility where they were actually running metabolomics samples.
We had them measured by biocrates. Because we were in the process of setting up the platform. I think the kit version wasn’t even officially on sales yet but we knew it was coming. Jurek Adamski was setting up the platform and with KORA, we said let’s generate some data, pay fee for service, measure 300 samples. – And that’s what we did: We got to ‘huge’ dataset. I mean huge 300 samples at the time. It’s not huge today.

Alice: At the time it was huge. Wasn’t it?

Karsten: It was. Especially the kind of data. I was very skeptical. For me it was like, how can you measure a drop of blood with these details how can that be precise? And then having another drop of blood and getting 500,000 gene variants out of that. Then do a correlation between them and find something meaningful. For me was a total surprise that that works.

Alice: Makes you feel very small. Doesn`t it?

Karsten: Yeah. And there was suddenly a lot of things in the data, which was probably also not correct. We found a lot of things later on miss annotations of metabolites which just totally normal. You have to know that. We ran the metabolomics against all other phenotypes of the KORA cohort, not only the genes.

I still remember, we had a project meeting at some point where we had like, I don’t know 15, 20 different phenotypes and we split them up between all the postdocs and the guys interested in one would be working on smoking and one on alcohol consumption and one on diabetes.
All the big topics that were in KORA. We spin them up and I computed the P-values and share them with them.

I think there’s a lot of papers that came out of that time. You can look them up. Where we for the first time we saw associations between phenotype and metabotype. And many of them made sense. But many of them were also complicated to analyze. I mean, we had papers with coffee consumption; we found that variant of nutrition style somewhere. – Sometimes a bit long shots but it was interesting to see that. My central project was the GWAS together with Christian Gieger.

The interesting thing for that project was that we needed more computer resources. The compute center for Munich had the 10th fastest supercomputer in the world at that time. And they were very proud of that. But they needed users. And normally physicists went there and used all the time, but they needed non physicist user. So at some point I got compute power from them and I wondered “how can I use that? How can I spend hundreds of thousands of compute hours on this machine?”

And that’s when this idea with the ratios came up. That initially we did the GWAS already and then we had so much compute time and we had already the idea with the ratios. We had it before, when we looked at the original data we got from biocrates of the mice. We just try it out because we just had the computer and it needed to be burned. And amazingly it worked! It generated tons of data.

Alice: So then with the computer there, you computed all the possible ratios between the metabolites. Okay. And then you checked which ones associated well with the data.

Karsten: Yes. And then you look at what we call the P gain.
So whether the P value really gets significantly stronger. Because you can of course always do ratios and if you do enough testing you will always get a little bit of a P increase. But what we were looking for and found was like we had a p-value of 10 to the minus eight and the GWAS.

And then we went to 10 to the minus 22. And then the very surprising thing was like the metabolites in the ratios. I tested all against all and some people came and ask why are you testing all against all? You should think before you do this? I really loved that because the biology actually tells you what’s right and what’s wrong. It’s not me coming up saying, oh, I know that this and this metabolites are linked together. I just tested really everything. – Also everything that didn’t make sense. And the ones that made sense actually stood out.
So whenever I look for something with a P gain in the end, also in subsequent paper, it really made sense. Even further on we had studies where we had unknown metabolites. Where we didn’t know what one of them was. With the P gain we could say: “This probably is linked to the other metabolite in the ratio”.

Alice: These are building alternative pathways then based on this with the few ratios that you did not expect. Did you look the parent to it to find? And you found the biology behind it?

Karsten: Yeah. There were the paper with Jan Krumsiek on mining the unknowns derived from this work. He was also one of the people involved at the time. And we could reconstruct pathways from partial correlations between metabolites and also from ratios and reconstruct underlying biology. I think one of our Gieger paper phrases was “if the function of FADS1 (one of our top hit genes) was known, we could actually have inferred it from the data alone”. That’s something I found pretty pleasing and something that very often replicated later in larger studies.

But it was already there in the first study. And I think that’s also why, I like the first paper most. Because you first described things for a first time. Then we had another paper in nature later. The first one, we submitted to science that they didn’t want it. So it just got to PLoS Genetics. The second time was easier. I would have said (as a reviewer): “Wait you already said that in your paper in PLoS Genetics”.

It went to Nature maybe also because the study was bigger, of course. And then GWAS became a bit like generating more and more findings which means there was not so much novelty anymore on the concept level. In terms of understanding biology, of course it was contribution more and more.
And even now – if one day we would have a GWAS on metabolomics in the UK biobank it would conceptually not be something new, but individually, on the function of each gene, the overlap with the diseases all that would be a good reason to do that.

Alice: And you think in the case of this type of studies you have really powerful statistics behind it, like holding the story together. But you find that it makes it more difficult to publish when you have new approaches to, like a new problem and new approaches to that problem. It’s not a good start. Even though that’s what everyone is trying to do. Isn’t it?

Karsten: Yeah. Sometimes it’s a bit too early for an ideas to get it right away into papers like nature and science. PLoS Genetics was good and I’m pretty happy about in the paper has almost as many citations as our following papers.

Alice: And the nice thing is that it’s available to everyone as well! You don’t need to have a license to read this which is nice.

Karsten: Yeah ok. If you want to put it like this – Yes!

Alice: Well, Open access is getting more but nature and science are not famous for allowing people to let them read their papers. So for us, the audience, it is nice.
You mentioned, that before computing all the ratios, you had already had the idea for the ratios. How did that come? Was it because you were interested in, for example, for this one and then you looked at the metabolites that were related to it. And then you saw a pattern there or how did you find out the first ratios? The ones that the computer didn’t find?

Karsten: Honestly, I don’t even remember in detail. But it was on this mouse data that we had from biocrates. And it was Elizabeth Altmeyer who was doing a PhD at the time with us. And I think our question was, “how can we make sense out of this data?” Just let’s play with it in every way a bioinformatician comes up with. Later on we had a paper where we thought about what ratios really are? what they mean?

But at the beginning it was just like poking in the dark and just say, oh, we could do ratios. – Let’s see what happens. And suddenly something came up and then you say, oh I can explain it. I think the rationalization came later to understand. I need to say a ratio could be a measure for the throughput of the reaction rate.
That’s one thing. But then it also could be a normalizing factor. Like if you normalize this creatine in urine. If there’s additional variation in your data sets like Fabian Theis from the Systems Biology group called it the French fries factor.

So if some people eat a lot of French fries and others eat little, then there’s a sudden lipid in your blood and if you normalize by that, you reduce the variance. And once you reduce the variance the P values get better. That’s the thing. Same way it is with creatine normalization. You have a signal in the urine. But if you have different dilutionn of the urine, you lose the signal.

Alice: Absolutely. Did you start out with a very defined workflow of how you wanted to analyze the data? For example for the Gieger paper you had a series of things you wanted to try out or did you have trial and error and then the kind of iteration process where you go – This could be improved with this and this. Maybe we could do slightly differently because we saw the results is not really what we expected. Is this something that happens a lot in your work or do you manage to go in a straight line because you know exactly what’s going to happen?

Karsten: I would like to say is the latter but that’s not true.

Alice: I don’t think that’s real. But I prefer to ask. You never know.

Karsten: The one thing which is really good with genetics is that you know that the genetic variance is causal. The genetic variant cannot be confounded. In other studies you can always be misled. You find an association for instance, these are just association. If you want to go for cancer, you do a case control study, you find something. And later on you find out that the metabolite you’re looking at is a metabolite derived from orange juice and you find out that cancer patients get orange juice at the clinic.

So that is not your marker for cancer. In genetics, there is almost no way that something would confound your genetics. So, if in genetics and GWAS, I tune parameters to improve the association like scaling the data or normalizing the data – Every time you repeat a fitting. I look at the P value, you done another test. So if you’re honest with yourself, you should not go for 0.05.
You should have a list and make a tick mark. And every time you tried, you should make a tick mark and lower the P value should be hitting it. Let us be realistic about that. But I think what I’m saying is if especially, you know an association is true, like for instance FADS1 and you later on you replicate that another data set and then you say, oh, should I lock scale my data? Should I filter this way, that way?

If I can tune the selection of a parameters to optimize the already known association without looking at everything else and if I would then go and use that as an adjustment to do the rest of the association. In my view that would be the correct way of doing it. Because then I wouldn’t be biased by the outcome of the test. The problem here is to be disciplined. You shouldn’t tune your thing to all the parameters until I find the most hits possible. In this respect you have to be honest with yourself.
That’s a dangerous thing because it is very frustrating to write a paper and say, I’ve found something spectacular and then it doesn’t replicate and the end comes back and bites you.

Alice: And this is something I also liked again in the Gieger paper. There was one part where – I don’t know if this was the true chronology or if it was just written nicely – the power of association was not good enough and the p-values were too high for this kind of studies. So you say, “this would have ended here. But we found the ratios”. It looks like, “maybe we would have been honest and we would have stopped, but, you know we found the ratios and suddenly it was super powerful”.
I liked this because it’s a nice picture of a kind of limitation that you have sometimes when you do the analysis and you find this sometimes in papers where people go, “We tried this and it didn’t work”, but there are all these experiments or all those tests that are done that we never hear about that are part of the work and that are part of the analysis and that don’t make it in the paper.
Do you have an idea of the time you spend trying things that lead nowhere? – As opposed to the time it took to do the actual thing that’s ended up in the paper?

Karsten: Yeah. I would make a difference here between genetic association study and clinical association study. The genetics is pretty straightforward. The way you do that is now pretty established. You can discuss how you scale your metabolites and things like that. You can maybe have more advanced things. You can do Bayesian or whatever association stuff. But that’s more routine now.

I think in metabolomes you have stronger signals. So in metabolomics GWAS we normally don’t go for the border line significant. Although they could still carry information. But there’s just no point in doing that because the next GWAS with a larger sample size will catch these. However for clinical studies of course it is quite different. Here you always start initially with the idea. I want to analyze my metabolome against this or that endpoint and then you do and that’s okay.
But then you start asking yourself, what is confounding? And you may find an association and you realize there is a weird metabolite coming up or this doesn’t really make sense. I have an example where we did a diabetes case control study. And we took the controls from a dermatology department at the same time as the cases, which was a good idea.

We eliminated the effects of batch separation taking samples of different places. But then, in the end, there were some metabolites coming where say, “are these really markers of the control?”. They were not really controls. They were people at the dermatology department.

Alice: The ones who all got the same lotion from the dermatologist or something like that.

Karsten: Yeah, probably not because it was very heterogeneous. There was one marker that was actually melanoma associated something. Although later on that it replicated in other studies that were not like that. So maybe I was wrong. Wrongly thinking it was wrong.

Alice: This is also the thing with naming of, especially of genes.
This can really be confusing. If genes are not studied much and there may be one or two papers that just led to the genes name. But it’s not so much based on it. And then you think everything is brain. You have to have this?

Karsten: Every name is derived from brain or cancer. That’s a problem with the genes. Well, the metabolites it’s a little bit less, although I think there are some cases. We jumped from one topic to the other. But you come also in the interpretation of metabolite associations. I mean this kind of pathway analyzes. That’s always something I always a bit reluctant as well. Although that’s also well-liked. But it is tricky because you have so many pathways and just saying, oh, this is a metabolite of oxidative stress.

There is not one metabolite of oxidative stress. It could be indicated from oxidative stress. But it could also be a lot of other things. And I think that is what makes maybe metabolomics more hard to interpret than for instance proteomics.
Proteins also have multiple functions, but not so many. Some metabolites could be at the basis of everything and to say, this is a metabolite for xy is very hard to pinpoint. Especially when you go to amino acids or more sensible nucleotides or lipids.

And then you also have the problem that metabolites come from everywhere. So they could come out of the liver. They could come out of the kidney, from the fat system.
The metabolites are in the blood for a purpose, at least most of them. The organism puts them into the bloodstream to carry them elsewhere. But of course, you also have others that are there because as cell died, they were leaked or are there because of other processes.

Alice: And they should be excreted later on.

Karsten: Exactly. Blood is a convoluted medium and you have other media, right? We have done things in saliva where we found a marker for diabetes that we also find in blood. So you don’t really know why does it appear there?

The urine studies people are doing CSF studies now. There’s a lot of interest in this but every time you have this problem of confounding with almost everything.

Alice: With metabolomics there’s this extra layer of what we eat. That’s especially if you look in the blood, but also in the other metrics. Like from when you start this, as you said, every organ contributes to the whole pool, but then there’s also what we eat and if the same person eats differently, then you might get different things. Which makes it extremely interesting but also very complex.
Do you ever use microbiome data or diet related data in your work with metabolomics?

Karsten: When microbiome personally less. Because we just don’t have the studies for it.

People starting doing that. Especially in the twin study they have their separate papers with microbiome already. So I think microbiome I could go on forever as well. I think there’s a lot of catchy things there as well. Concerning nutrition – Yes, there’s a study we did also with the same team in Munich years back. It`s a human study, where we had 15 male healthy volunteers.
And they went for four days closed into the technical university study center and they were 36 hours fasting and they got controlled meals and everything. Gabi Kastenmüllers group (The Metabolomist Gabi Kastenmüller) is just preparing the web server that now has all this data integrated.
There’s also biocrates kit data, the Metabolon data, data from urine, from blood off course. It informs you also about which metabolites are stable over the course of the day (where you wouldn’t have to bother about fasting) and others where you probably would have to bother about fasting.
Like this there’s a lot of confounding there as well. For instance, a diabetes person is more likely not to be fasting than a non-diabetic person. Because a non-diabetic person could go without food. A diabetic person would probably be careful to equilibrate their nutritional intake and then you would have already a confounding factor.

Alice: Yeah, absolutely. But this principle of making databases of metabolomes or other ‘Omes’, is a really interesting thing as you’re working a lot on this as well to combine data about specific diseases or about specific topics to make it available to the community.

Karsten: I didn’t mention that in the introduction. We working in a virtual group together with Gabi Kastenmüllers group in Munich and she is doing all these web server things. They have an Atlas of Alzheimer´s that they presented bringing up where they connect that. And the group of Jan Krumsiek in New York; they are very much into the systems biology of it (like the gaussian graphical modeling and the networks; and how can we put P values not only on a symbol association but on a pathway or on the part of the network (The Metabolomist Jan Krumsiek)).

Alice: Would you like to discuss some specific tools are there, especially to help make sense of the metabolomics or to integrate them with other omics.

Karsten: I think in the end, it all comes down to people who are working on R and R studio. And there are so many packages out there. I think new generation is also doing python, which probably goes to the same thing. I’m not a python person. I always stick with R. But I think both are pretty close to each other.
It is certainly a good thing of looking what other people publish because there are so many new tools out there. A lot of stuff is how do you visualize your data and how do you nicely produce it?
So going away from these hairy balls that come out with network here, network there. – You just see it and say, okay, what do I make out of it?
I’m always thinking about my medical collaborator. Would he really pull something out of what I’m putting into the paper? Or is it just to show that I have big data and I can manage it and nobody else. I think, there’s a big gap between what the bioinformaticians and the systems biologists can do and what the clinicians (who actually are our clients in many cases) really do with it.

They say they want to do a metabolomics study and we run the whole thing and they have their data; you give it back to them and then they’re frustrated. They say, “ So, what next?” “What does PC AA 36.4 associates with smoking?” “What does it mean?” “Does it have implication?” – and many people walk away pretty frustrated to be honest. That’s something we still have to really work on to nail things down. What is really a takeaway message more than just saying this metabolites goes to that and this metabolites goes with that.

Alice: This is exactly the purpose of this project and also the purpose of this podcasts. To give people clues. So, the clinician or the research scientist who is not necessarily an expert in metabolomics and wants to understand these results. So, what are ways to help people to jump that gap?

Karsten: Let me just list the few things. That’s more like throwing buzzwords out there. I think one keyword is certainly Mendelian randomization which is using GWAS data to build causal relationships or confirm causal relationships.

So, if you have a lot of GWAS data, big GWAS data with much power, then you could in a way show that a certain metabolite is on a causative pathway to the disease. Which would then mean in terms of causality that if I thinker with that metabolite, I would change the outcome. And that’s what you want. Right?

Alice: When you work with this kind of associations like you build onto the kind of better annotation of the genome to try and understand where the metabolites fall?

Karsten: Imagine I would have an association with a disease like diabetes and an association with a metabolite and I would try to figure out is the metabolite something that causes the disease? So, if I bring down a certain value would that improve the disease or is it a consequence? – So it might be a good biomarker. But it doesn’t make a sense to target that to bring that into change that value.

So that’s instance I think something that is big at the moment and it really requires even larger GWASs on metabolomics. I think that that alone justifies the effort.
Second, there’s of course the ratios which I still support strongly
Third, gaussian graphical models (GGM), also called partial correlation networks.
That this is really interesting because if you just do correlation networks, you end up with a hairy ball of everything, connecting everything. So that’s something I would look at. You can predict variation in metabolites in people based on their genome (Only the genetic part of course). So, we could do that and link that to the disease.

What I haven’t mentioned yet, we have in the meantime also done EWAS (epigenome wide association studies). And that’s something which compliments in a way GWAS. Because GWAS is genetic, that’s from birth. You’re set to be a fast or slow metabolizer for this, at this gene. The EWAS as I see it is like there are certain parts in the in the genome that get methylated and that are switches to switch on or off genes.

And in some cases, they reflect what your body does.
So, there’s a very strong association of TXNIP with diabetes. And I think it’s a read out of how much of TXNIP the body actually needs to cope with the diabetes. It is not necessarily creating the diabetes. This could be the reverse, it could be showing you what the body actually does at the moment to cope with the disease. And that’s the tricky thing.

Is it a sign of disease or is it a sign of the body coping with the disease? But it can give you ideas on where you would go and try to treat the body. And then also a way of maybe early markers. Because if your body is adjusting and fighting diabetes (but you’re not diabetic yet), you might already have your transcription (epigenetic) profile changed.

Alice: You are already being challenged. Yeah.

Karsten: We did an EWAS paper later on in 2016 with Ann-Kristin Petersen where we did an EWAS on metabolomics with the same data just like before. We found 20 different genes where the CPG was associated with metabolites and independent of genetics. (There were others where the genetics was confounding.)

A large part of them were associated with smoking. I think smoking is a very good signal there. And then we had few isolated ones initially I didn’t know what they had in common. I think the message for people analyzing data was we were always looking at the top hit and that was an error. Because it’s not only the strongest association – It is all the metabolites what we call the metabotype you should look at.

And when we looked at it later again, we didn’t EWAS with BMI and diabetes in a Qatar cohort. And it turned out there were actually three of these genes (all with the same metabotype and marker of hypoglycemia). And they had been reported in the paper before to be the metabotype of diabetes.
All the three genes were diabetes-associated metabotypes. And only later people discovered that these genes were actually associated with diabetes and obesity. Our EWAS actually had the information already. We just didn’t see it.

Alice: No.

Karsten: It is also a good thing because there’s a lot of information in your data set. You could be looking for it. I think, especially for people working on metabolomics this should be very motivating because there’s a lot of stuff hidden in the data that is waiting to be digged out.

Alice: …and that you can’t really see yet if you don’t work with associations. But you work just based on the previous knowledge of biology. You just can’t right now because it hasn’t been discovered yet.

Karsten: There’s another thing which I always like to do. When you have data, normally you collect covariates such as age, gender etc. Always be looking at associations that you know of and use them as a positive control. This way you can be sure that your data is having the information you are looking for.
For example, from a previous study I know the strength of an association with ages. If suddenly my association is weaker or stronger I know that something going on. I can also use this effect in reverse if I know something does not associate. To do an association with a random number, it should always come out negative.
But if you say I do an association with a day off of data collection or this technical covariates. I mean, that’s very important to not just throw them in as covariates but to look at them and to make sense out of them.

Alice: Yeah, and this is a very strong point for annotating the data with as much information as possible and not removing the metadata that is given to us. Maybe it does matter if it was a Tuesday!
Karsten: There are interesting examples. We did a multi-center study on kidney rejection. And there were center specific metabolites. They were only measured in one of the clinics. And when I looked at the thing it was totally perfect because the different clinics did the initiation of the immunosuppression with different drugs.

So the metabolomes actually (only) measured the different drugs that the clinics use. I could have reverse engineered how different clinics initiated the immunosuppression. In another study, when we saw a metabolite appear (or disappear) sometime in the order of sample collection.

And when we looked at that, a clinician told us that they just changed the urine tube or the labels on the blood tubes. It doesn’t mean that your data is bad but metabolomics is very sensitive. For the urine tube case it turned out that they had some kind of preservative in there. And probably the other tube has a different preservative. You just have to be sure.

Alice: You need to know that.

Karsten: And you mustn´t use the one tube on the cases and the other tube on the controls, but you have to know about these changes.

Alice: As I mentioned before, I try to make a point for the place of creativity in interpretation of data. Do you see a place for creativity in your work? Do you think it’s something important for a scientist in general and for the type of work that you do in particular? Or what’s your, what’s your view on this?

Karsten: I think creativity, especially creativity in visualizing data is important. Because the most important thing is that you have to see things. I mean, you may have very strong P values but once you look at it it’s just driven by three data points.
That you can do with it with a simple plot. There’s a new generation of informaticians who really take data presentation tools from totally different fields and bring them over into science. Like this interactive Java visualization tools and things like that. And that’s another thing. The second point here is also to have your data interactive in a way that non-bioinformatician users could play around with the data in an easy way.

Alice: That can also help to communicate your results. If you have very dynamic study structures, like you should have different time points or different stages of a disease, that can really help as well. Do you spend a lot of time playing around with visualization once you have the data or right at the beginning maybe?
Karsten: I personally don’t find so much time for that anymore. But that’s really where the PhD student and the postdocs come in.

I think where they also can make their mark. Look here is really something that speaks to you and is convincing. It shouldn’t be a black box what you’ve done before – But in the end, the idea is how do you synthesize a message out of your data. Of course you can call that creativity if you want. It depends on how strict you are. Some statisticians would hate that word.

Alice: Yeah, but as you said in the visualization you can be very creative. For me it is creativity to go from a black and white table with little stars to a graph that maybe has only one or two colors. But is putting the data in the light where you can actually see the differences where you don’t have to compute it in your head to see it.

Karsten: You should create a hypothesis; it should teach you something, and then ideally you would either carry on with another experiment to prove that what your hypothesis generated is true or at least depending on what the study is do replication, especially in GWAS. It can be as creative as I want in the discovery but what I come up has to replicate in an independent study.

Alice: I have this in my own experience of analyzing data or looking at datasets where sometimes, I did this analysis and then I did a different analysis and then I looked at through different angles and sometimes I would forget to stop. I consider this as a symptom of perfectionism in a way.
That you want to get to this beautiful aha moment where it feels like you´ve finally explained biology. But it is unlikely that you’ll get to this ever, especially with one study. Are you familiar with this? Can you maybe comment on this?

Karsten: Yeah, I think that’s a very risky thing. Especially if you don’t have a fixed position. It could be that in the end you would never publish the thing and just get out of research. Because in the end it’s publish or perish in a way. I think it is a thought that I don’t see that dire but the thing is you need to be there and tell a story to someone. You don’t do it to file it away and when you retire you just say, oh, I have everything in my drawer but I never told anybody about that.

Alice: And do you think that statistics and significance alone can tell the story or you need something more?

Karsten: I think the most important is the statistics. P value always means what’s the likelihood that you see the single that you’re seeing by chance. And if whatever you have in front of you could just be in front of you by chance. It’s not worth it being considered further. So, in the end, everything you do should in one way or the other be supported by statistics.

Alice: Yes.

Karsten: That’s something I like to do very often is to just randomize the identifiers of my data. Because my biggest fear is that someone drops the box with the tubes and put them back into the wrong order and I don’t know about it. That’s why I test for age or gender associations to make sure that there’s not something generally wrong with my data.

Alice: That’s a good point. I didn’t know about this. Of course, I’m used to looking for outliers in groups and things like this. And we’ve had cases like this when I was working in academia where we just didn’t have so many replicates for a signature that were clear enough. But at least we had clear signatures. When you don’t have this, it can be really risky.

Karsten: Yeah. And that’s important to have this kind of markers of especially of sample integrity. It happens very quickly especially if your work as a bioinformatician and never see any sample tube. I make case of seeing the tubes. Normally, you just tell someone to ship the samples and you don’t even see them. We had the case in the past with genomics (not metabolomics).

We gave away a box of samples and they turned it by 90 degrees and took the tubes in that direction. The error was spotted easily and we did know what happened. And then if numerically we turned the thing back by 90 degrees suddenly everything mixed. We checked the sex match at that point. So it was totally unmatched before and after we turned them they matched perfectly.

Alice: Yeah that helps.

Karsten: And I think this kind of thing you would like to have for metabolomics, as well.
Of course, you would like to have somewhere replicates. Although if you go for this fee-for-service thing, it’s hard to just send three times the same tubes. Sometimes, initially we did that. We even sneaked in a few samples that didn’t tell the people that were duplicates in there.
It’s not really that conclusive because I think especially if you sneak them in, it’s hard to bring it up later. That’s the thing: If you have duplicate measures it has influence on your CV, the variance, technical things, everything like that.
When you work with core facilities and fee-for-service providers, you trust them to a certain point.

And once you create trust with them, you prefer working with their data rather than with others. Because suddenly they come to this other paper and say they used mass spec metabolomics to measure but it comes from a platform I’ve never heard about. And I don’t know what to think about it. I would like to see all this kind of standard things done before. Does the platform have the replicates and all standard associations so I can trust the data.

Alice: Yeah, it’s similar to comparing methods for other types of scientific research as well. For the omics, we always kind of expect any omic to be comparable to each other. And sometimes it is like this even though, of course, you have points where you can compare and you have the quality controls that it’s probably more robust than other kinds of biological research.

Karsten: And that’s also why we should not be too shy of replicating things. Everybody wants to be the first to something the for the second paper the Editor already says it’s not interesting. But I find most of this second paper is the one that really confirms it. It’s safe to say. And you discuss what is really replicated and not what the first paper just stretched to the end of how far you could go in your interpretation.

Also, with more data coming out and the community is growing.
It’s a good time now to start in the field. And of course, also in getting more standardization there’s also a gap in the community there. It’s not a gap, but I mean the people who are really the mass spec specialists. And then you have the people who analyze the data in between them. The mass spec specialist is often not too concerned about what people do with the data later. They want to be as precise as possible. Having a 10% coefficient of variance is not bad for something. But for a biochemists it might be already horrendous.

Alice: Yep. So very relative. Is there anything you would like to add on the topic maybe a message he wants to get out to the public?

Karsten: I end with my favorite quotes that I always put on my email. If you torture the data long enough it will confess to anything. And I think that’s really something I always put it there to remind myself. You can find everything if you just look at the data long enough.

In the end, we always are responsible to the people who read our papers are not metabolomics people. Who are critical to what we do while we have to be creative in our methods you have to be critical as well as what we claim out there.

Alice: It was delightful to talk with you. Thank you very much for your time.

Karsten: I wish you good luck with your book and everything.