Identifying hidden gut bacteria

This fairly recent realisation of the significant health implications of a well maintained microbiome is all well and good, but if we’re going to help everyone reap the benefits there’s a lot of work to be done identifying and experimenting with the range of bacterial species that have colonised us. Working with DNA sampling technology, the University of Cambridge’s Alex Almeida is helping to reveal the many mysteries that remain in this field, and he spoke to our own James Tytko from his laboratory…

Alex - We can look at the microbes as individual organisms, so the number of actually microbial cells we have in our body. Current estimates sit at around 40 trillion microbial cells. Most of them are in the intestinal tract, that is clear. Now, if we take it to another level, so when we break it up to the level of species, in my line of work, the way that I classify what is known and what is unknown is based on whether we can actually experimentally characterise these species. So the way we do this is we try to isolate them and culture them in the lab first so we can actually work with them and perform experiments. Of current estimates of essentially the whole repertoire of microbial species that have been identified in the gut, 70% are unknown, so have not been cultured. That's a big limitation in the field and something that I'm looking forward to in my research to clarifying a bit more.

James - And we'll often hear that there are helpful and harmful bacteria that live within us at the same time. Help me to understand that a bit more. Can you provide some examples of bacteria that are good for us and ones that are not so good for us?

Alex - So this definition of good and bad bacteria is actually something microbiologists don't really like to use when classifying because whether a bacteria is good or bad is all about context: it's all about the interaction of that bacteria with the host, what are the other types of bacteria that exist surrounding it? How they all interact together. What I can tell you is that there are consistent trends in linking certain species with health or with disease phenotypes. One species consistently associated with health which has been quite extensively discussed as a novel potential probiotic is a species called Faecalibacterium prausnitzii (all bacterial species have Latin names.) It actually helps digest certain fibres and produces a compound that is known as short-chain fatty acids. Short-chain fatty acids have important roles in host inflammation which has been strongly linked with maintaining gut equilibrium, gut homeostasis and keeping things in check.

Alex - On the other hand, potentially disease associated species have been consistently associated with colorectal cancer is a species known as Fusobacterium nucleatum. What studies have shown is that colorectal cancer patients are consistently enriched in bacteria belonging to these species, and there have even been some experimental studies suggesting that the species is involved in tumour growth development as well. Obviously, there are more complicated species that have a bit of a mix of both being good and bad depending on context. The classic example is E. coli. Actually, E. coli is a normal commensal of the gut and is found in most individuals at the very low abundance. If E. coli reaches a significant level of abundance, that can create problems: it can lead to high levels of inflammation and even if certain strains of E. coli are able to translocate the intestinal epithelial, that can create problems because the bacteria can transfer to the bloodstream, transfer to other organs and that will cause problems for the individual as well.

James - This is why it's becoming an increasing priority within medicine to identify as many of the hidden bacteria in the gut that you spoke to earlier.

Alex - Yeah, exactly. So obviously the ideal scenario is to really find species that are robust and strongly associated with health and can benefit different individuals in different contexts in different health states. But we really need to first take a fundamental approach to understand what the biology of these species really is. What are they doing? What functions can they encode? This, before being too ambitious and immediately trying to use these species as potential probiotics.

James - And so this is where we come to your research and the tools we have are our disposal to identify more of these species.

Alex - So my research group is primarily a computational genomics lab. Genomics is the area that studies the DNA sequences of different organisms. In our field, for the microbiome world, we use what is called metagenomics, so it's beyond even genomics. It's actually looking at the whole genetic material of the community. So we do metagenomics on a traditional microbiome sample - usually what we work with are faecal samples to explore the intestinal microbiome - and we piece together these DNA sequences to make inferences about which pieces are there. So the DNA acts as a fingerprint, but since the DNA also acts as a sort of blueprint for what the bacteria is able to do, so what proteins it's able to encode, we can predict the metabolism and the function of these bacteria by looking at the DNA. We use large scale sequence data to really make inferences about what is the role of the microbiome, which species are there, what are they doing in different health contexts.

James - Tell me about some of the examples of species you've been able to link with certain diseases.

Alex - We have been looking at a wide range of diseases, some that are a bit more obvious such as inflammatory bowel disease or colorectal cancer, but we have also been looking at diseases that are not directly associated with the gut, like Parkinson's disease or multiple sclerosis. We definitely see a stronger link with diseases that occur in the gut but, even in cases of things like Parkinson's disease, there is definitely a strong link with the microbiome. The challenge is actually pinpointing when we find specific species linked to these diseases, why are they there? Is this just a correlation? Is the species present there as a consequence of the disease? Or is the species actually leading to a higher incidence of that disease? So we do find that there are some uncultured bacteria, species that we know very little about, that are very strongly associated with health in different contexts. I can give you some names but, to be honest, since these are uncultured, unknown species, their names are simply codes at this point. They carry very little meaning to people that are not directly in the field, but we are starting to see some good trends.

James - Highlighting there the difficulty with delineating cause and effect. What's the scale of the task there? How do we translate the work you are doing into medical diagnosis or treatments?

Alex - I think there are two levels we can think about. So, in the short term, the more perhaps realistic goal is to use certain species as biomarkers of a healthy or diseased state. And I think this could help with certain diseases where diagnostics can be quite invasive. If we think of things like colorectal cancer, obviously the gold standard diagnosis is doing a colonoscopy of an individual, but that's very costly, very invasive and carries some risk as well. So if we can develop a diagnostic tool as a method where, by analysing a faecal sample, we can identify specific species or specific compositions of the microbiome that clearly distinguish health and disease in an individual and we can use that information to identify individuals that are at greater disease risk, they could then be selected for further, in-depth screening. For a diagnostic purpose, this question of correlation versus causation is actually not as important because, even if the microbe is a consequence of the disease, it can still act as a good biomarker of disease that could be used for a diagnostic purpose. Now, obviously the field is becoming a bit more ambitious and thinking of how we can use the microbiome as a form of therapy to actually improve or protect against disease, and this becomes much more challenging. So in this case, we do need to establish causation. We need to identify species that directly protect or directly cause certain diseases. And then after we have that information, we can think about how we can enrich that particular protective bacteria or how we can reduce the bacteria that potentially is causing disease.