Good morning, everybody. Thank you very much for the kind introduction and for the invitation. I will give you some information about this method. I have no disclosures. Of course, when we think about analyzing the gut microbiome, we have still to keep in mind that the composition is relatively stable. But most of those bacteria reside within, as you can see here, this rather thick mucus layer. Therefore, it raises immediately the question whether analyzing in stool makes sense. But it makes sense because it's very practical. But, of course, we always have to keep in mind that also the mucus layer, the thick mucus layer, is an important reservoir of all those bacteria. It is also important that we understand classification so we know what we are speaking about. As you can see here, it starts with life. It goes to domain. Then we have kingdom, which is bacteria. And then we have phylum. Phylum, the most important and well-named, are Firmicutes and Bacteroidetes. Then we have class. We have order. We have family. Then we have genus and species. Those are those we are usually analyzing. When we look at all those aspects, then it's still very clear that most of the stool is bacteria. But most cannot be cultured. I mentioned it in a few moments in two papers. We have at least 4,000 species. We have around 30 phyla. The six most important are mentioned here. Now there are, in my view, four definitions which are important to mention. You have to know them so we know what we are speaking about. Alpha diversity, what's that? It's the measure of diversity in a sample. We have two names usually to consider. It's K01 or Shannon. Beta diversity is the measure of diversity between samples. It's named commonly Bray-Curtis. When there is no dissimilarity, it's zero. One represents no species shared between the samples. Then we have 16S ribosomal RNA, which is the most commonly used method to detect bacteria. We speak about bacteria in this case. It's the bacterial small ribosomal subunit. The DNA coding for this contains highly conserved areas versus highly variable regions. Then we have another term, which is OTUs, which is operational taxonomic unit. Those are 16S sequences which are grouped together at the level of similarity. Usually it includes all sequences within 94% similarity. Which methods are we using? Mainly DNA-based approaches. Who is there? What can they do? Usually it's either 16S rRNA or metagenomics. I mentioned the differences in the next slide. Secondly, we have metatranscriptomics. How do they respond? Third, we have metabroteomics. Then we have metabolomics. Ideally, you can say we combine all those various techniques to gain most knowledge from your sample. When we look at the most commonly used technique, which is 16S rDNA, then you can see here we have those variable regions, which are represented to scale. Usually V1 to 9 can be used to identify a specific genus or species of bacteria through the sequence. Most commonly used is V3, V4, and V6. When we compare the 16S method with shotgun sequencing or metagenomics, then the question is, what are the main differences? When we use 16S, then we select or we assess bacteria. This is very important to understand because when we use metagenomics, it's not only more costly, but it is able to detect, let's say, everything, including the host, but also viruses and fungi and so on. We have here a technique, which is probably the future. At the moment, it's more expensive than 16S, but it allows to detect everything. That's the reason it's called metagenomics. What are we doing then with those samples? They are processed and analyzed. We assign sequences to various samples. Then you assign them to clustering. Then you do a so-called BC analysis, which is a principal component analysis, to look for clustering. For example, disease A is clustering here, or disease B is clustering there. But when we thought that there are only 1,000 species, then there are two important reports this year, one in cell and one in nature, telling us the world in there is even much more complex. This is a paper by Nicola Segata from Trento in Italy. They did an analysis of almost 10,000 metagenomes, as you can see, from 32 countries, many different body sites, including seven non-Westernized datasets. They were able to identify that there exist around almost 5,000 species, not 1,000, and also showing that 77% of species were never described before. But this also tells us that the techniques we are using in this type of research are still not mature and are still under development. This is also proven by a second paper, which was published in Nature, also this year, where they identified 2,000 new bacterial species when assessing almost 12,000 human gut microbiomes. It's very clear we are not at the end regarding number of species. Probably many more are still coming up. Now, let's briefly speak about sample storage, because this is very important, because we are always concerned we are doing it the wrong way or what is the right way to collect your samples. I will refer on to two papers published in Scientific Reports. This is the first paper. They compared a minus-80°C freezer, then a 4°C fridge, then dual stabilizers, omni-chain and RNA later, then a DRIS-EDTA buffer, and room temperature. They looked at species richness. What is richness? Richness is the number of species in there. Of course, we usually consider more species, then it's, of course, richer. But, of course, we can also say, and this is probably reflected here, when you use room temperature, then there is outgrowth of certain species, and that's the reason it goes up. But when we then look at Shannon diversity, which is alpha diversity, then you can see that room temperature is doing badly. But also something which is used commonly, namely RNA later, was not doing very well in this study. The other stabilizer was better omni-chain and was comparable to freezing immediately or bringing the sample into 4°C. So, it's very clear that certain commercially available kits may provide an important alternative in case refrigerator or cold chain transportation is not available, which is once the case. And when you look at samples, and especially refer to the right side, you can see here room temperature and the DE buffer. You can see in this color proteobacteria coming up, which is not a good sign. So, probably using this as a storage component is, of course, not recommended. On the left side, you can see much better sampling with refrigeration and putting it at 4°C. And there is another study which is also very important. They compared different aspects, so immediate freezing or lyophilization or storage in ethanol or in RNA later. And very important message from this study is that the signature of individual identity consistently outweighed the storage effects. And I think this is very important. And it's shown here when you look at correlations, lyophilized versus ethanol and all the aspects they did. Always very nice correlations telling us that there is indeed individuality which is very important and probably more important than the storage method. I also want to draw your attention to other aspects which are coming up now. And, for example, one is circulating microbiome. There is a short paper in GUT published last year where they looked at circulating microbiome in different circulatory compartments. And in this case, they used full blood leukocytes from different areas, central venous blood, hepatic venous blood, peripheral venous blood, or portal venous blood, and showed that the microbial communities are not different between those compartments. So, obviously, you are able to detect parts of the microbiome in those circulating leukocytes. Of course, at the moment, we do not know what this means. But I think it's a fresh view on this topic and quite interesting. And they have shown in this short paper that there was a correlation. I think that's interesting with cytokine levels, with certain chemokines, and also with IL-1 receptor antagonists. So, theoretically, this could be of relevance in all those diseases. So, to summarize what I have said before now, best practice during sampling collection storage is not always an option, because you're probably not able to have the best strategy. But consistency is a must. So, when you're designing something, be very consistent in your collection. That's obviously the most important aspect. And still, bad samples will result in bad data. But so can good samples, on the lack especially also of proper controls. And you have heard in the first lecture how important it is to reflect on confounders. What means good controls in all those studies? Yes, age-matched, ideally BMI-matched, diet-matched, and no drugs as confounders. And we all know that's not so easy, but it's very important, because we have to consider that. So, a well-thought-through, detailed project plan, and most importantly, consistency during sample acquisition, storage, and processing are crucial for a successful microbiome study you're hopefully planning. And then we go further and ask our bioinformatician, because at the end, of course, it's perfect sampling, perfect storage. But then it's using the perfect or the right techniques. And then it's having the best bioinformatician. And, of course, those people commonly are poor, because they say, I'm getting some data because my PI said, blah, blah. Or we sequenced the metagenome. What to do with it now? Or can you look at the data and see if anything is interesting? It just tells us, before we start something like that, we need a very clear hypothesis and a very clear plan for what we are going for, and not just simply measuring something because we have the samples. So, what and how to collect? Well, stool, of course, yes. Biopsies, yes, interesting. But in practical medicine, usually we refer to stool, because that's practical and easy. Saliva, yes, as fresh as possible. Think about the proper medium. Think about omics and how to collect properly for doing the omics studies. Consider full blood sample for circulating microbiome. Could be something which is interesting tomorrow. And another, at the moment, evolving issue is tissue, studying liver tissue, something probably for the very near future. And at the end of my lecture, I give you an exciting sample how a metabolomic approach might succeed. This is a cell paper by Frederick Beckett's group. And what they did, they started with patient samples. They had access to portal vein blood from patients with type 2 diabetes. So they did metabolomics and were able to identify a metabolite with the name imidazole propionate. That's a histidine derivate, and it's a bacterial-derived product. And then in very nice studies, they showed that imidazole propionate affects glucose metabolism, as shown in several experiments. And here, for example, those are animal data. In germ-free mice, you cannot find this metabolite, but in conventionally raised mice, yes. And on the right side, you can see the data of the patients, indeed, that those patients, those type 2 diabetes patients, had increased concentrations of these metabolites. So that's a very nice example of how, from bedside to bench, you can do very nice studies revealing exciting effects of new identified metabolites. So, ladies and gentlemen, this is my key takeaway slide. It's obviously a far more complex world than previously believed. I just refer to those two new papers showing that there are around 5,000 species inside the human gut. Decide what you want to learn. Define a hypothesis before starting collecting, and it's all about consistency. This means consistent sampling is the most important thing. Either rapidly refrigerate or select a certain medium. So, at the end, have fun with your microbiome project. Thank you very much.

gut microbiome analysis

bacterial classification

metagenomics

microbiome diversity

metabolomics research