Mass spectrometry based proteomics, or MS proteomics, is a scientific method that can give us new and important information about NAFLD and other liver diseases. I would like to know how MS proteomics is different from conventional methods and what the perspectives are in relation to NAFLD. I met with Professor Matthias Mann who shed some light on the development and application of MS proteomic in patients with NAFLD. Thank you very much for joining me here today, Matthias. Thank you for having me. What I'd like to know first is what is MS-based proteomics? Yeah, it's a bit of a complicated technology. MS stands for mass spectrometry, and that's the method to basically weigh proteins and identify them. It's a high-tech method and what we in the end get is, what were the proteins in the sample, and how much was there of these proteins? Something about which ones are there and how many? What can the methods do that conventional methods can't? The biggest difference with conventional methods is that it can look at all the proteins all at once, and very precisely. Existing methods only typically look at one or a few proteins, and then sometimes because they're antibody-based, they may not even find the right protein if you're unlucky. That's the difference in mass spectrometry, it's very specific and it can give you information about thousands of proteins. Is it being used already in clinical medicine? Mass spectrometry is already used for small molecules, so when you have a drug of abuse or you metabolize a drug, then that is typically found in the plasma by mass spectrometry, so for small molecules. But for proteins it's not really established yet, but we certainly hope to change that. Have MS-based proteomics provided new insight into liver diseases such as NAFLD. Yes, it's certainly starting to do that, and our lab was involved in a study where we looked at NAFLD and we actually found new biomarkers. We also found drug targets that were already known, a new drug targets, and we then validated these results in mouse models. This is an example of that, it's already contributing. What has MS proteomic provided of insight in medicine today? MS based proteomics has already been used in the cancer contexts, for instance, and it's found some biomarkers there and some mechanisms of actions that was mainly in the research area. But now it's getting into routine use, and where it's actually already been used for decades is in small molecules, so when you want to look for metabolites in the blood, it's been established there for not proteomics, but mass spectrometer has been established there for decades actually. It is in that way in the clinic, but it hasn't been used for proteins yet, and that's a new thing and this is starting to happen. This is very interesting, so what you're saying to me is that you can analyze tissue as well as blood? Yes. It's in the nature of the technology that you can analyze everything that has proteins in it. It's also very robust. We can also go to archival tissues, for instance, that are decades old, and if just as long as we can extract the proteins from this, we can also analyze them. In the clinical context it's mainly body fluids that are interesting, so the constituents of blood, plasma and serum, but then also tissue like again in the case of cancer tissue. This can be analyzed by MS-based proteomics. This can tell you something about the diseased organ itself? Yes. For instance, in the plasma, we're looking for proteins that should be in the plasma. But then we're also looking for proteins that should not be in the plasma, so they are there because of organ damage. For instance, a prostate specific antigen is an example of that, so normally it should be not in the blood or very low levels, but if there's a prostate cancer, it could be elevated. But this goes for other organs too, so for instance, we could see proteins that indicate heart damage or proteins that indicate liver damage, for instance. You get a lot of information when you do your analysis. Does that give you any special ethical considerations? Yeah, indeed. Normally in diagnosis we have a specific question. We test for liver disease for instance. But it's in the nature of MS-based proteomics that it provides you all the information all the time. We may learn, yes, this person has liver disease or is in some risk group for liver disease. But we may also learn that this person is actually diabetes. Because uncontrolled diabetes, because we can see glycated peptides in the blood. Then the ethical question would be, should we not tell that person because they didn't ask for it, or would it even be unethical to not tell them because they could do something to prevent the damage from diabetes? The interests, they actually very interesting ethical questions associated with such a broad technique. You say there's a greater risk of accidental findings when you do something as far Yes. as this? Yeah, absolutely. Because the whole technology is almost independent of the disease and you say risk of incidental findings, but I will turn that around. It's a chance in some ways, its a chance for incidental findings. We can distinguish actionable incidental findings and non-actionable ones. If they're non-actionable, then it's probably better not to tell the participant or patient about it. But if it's actionable, maybe they can opt in before and can say, so if there's something else wrong with me, I want to know about that or if there's something else wrong with me, I don't want to know that. This a little bit resembles the area of genetic assessment. Yes. Do you feel that these two fields can help each other find a better solution for? Yeah, that's an interesting question, how genomics and proteomics are different or the same. Genomics tells you about the genes that you have, but you have inherited them, so it's too late to do something about it, so to say, where's the proteome tells you a combination of your genomic risks plus your environmental risks or your lifestyle risk, so and that is very dynamic, so you can still change it. Let's say we did a test and we saw that you had glycated peptides that will indicate uncontrolled diabetes, and now you could control it and then your proteome will be okay again regardless of your genome. That I would say is the difference between the two in a nutshell. Would you say that genetic analyses and proteomic analysis compliment each other? Yes, so I would exactly say that, so it's getting easier and easier and cheaper to get your genome, so I think we can see in the future that everybody would basically know their genome already, so that's a given, and that will modify your risks. But then that doesn't take account of everything that happened since you were born until now, and the proteome can reflect that, and it can tell you whether you are sort of in the red zone with your disease risk, and it can tell you probably what you should do to get back into a green zone so you don't have that risk. Maybe even regardless of some genetic risk factors that you had. Do you feel we can use proteomic in the future to diagnose patients with liver disease? Yes, we hope so, definitely. In fact, we just have a study that we hope to be publishing in the future, and we were looking at fibrosis, so that's difficult to diagnose in the early stages. Just taking the plasma proteome of a large cohort. We could actually see then if we apply machine learning on the results we actually outperformed the best available liver tests that there are now. But that said it would take a long time to get these tests robust enough, cheap enough, and approved, and also somebody would of course have to pay for them. This will certainly take a number of years. Thank you very much Matthias. My pleasure. Thank you.
