top of page

How scientists study the microbiome

If you’ve had a read of this blog before, or done any reading about the microbiome, you would know that there are a LOT of scientists researching how the gut microbiome affects our health, what our microbiome is made up of and how it changes. But how do scientists actually do their experiments? With our gut microbiome being inside of us, it's not that easy to learn about it. So let’s look at all the different ways we can understand more about our gut microbiome.

Animal models of the microbiome

A key aspect of good scientific experiments is that they are well controlled, meaning we only change the variable we want to investigate.

Have a look at this link for more on controlling an experiment.

This means that studies on the human microbiome are difficult, as we all eat different foods and have different genetic backgrounds. This means we sometimes need to use animals (who eat the same food and have a known genetic background) to properly study the microbiome. Researchers have to be highly trained to be able to use animals, and all animal research is done under strict conditions and following a stringent application process that ensures the project is justified and animal welfare is maintained. There are a few different types of animals (known as animal models) we can use for studying the microbiome.

Germ free (GF): These mice have no bacteria in their gut, or in other places they are usually found such as the mouth and skin. To facilitate this, they are raised within special isolator housing which ensures bacteria and viruses from the environment are unable to enter. By using this model, we can compare how normal mice, and mice with no microbiome, react to different medications or interventions.

Specific pathogen free (SPF): These mice are free of a list of specific pathogens (these include viruses and some bacteria). An example of where SPF mice are useful is when researching lung function- making sure the lungs are free of pathogens that affect the lungs (e.g. influenza), ensures the experiment is well-controlled.

Antibiotic-induced microbiome depleted (AIMD): In this model, four or more broad-spectrum antibiotics are given to the mice, usually in their drinking water, to significantly reduce the amount of bacteria living in the intestine

Original image source: ScienceDaily

Germ-free mice need to be kept isolated to prevent contamination... it is a little bit like the boy in the bubble.

Are these models useful?

As you would know from this blog, changes in the microbiome can lead to other changes in the body. Recent research has shown that AIMD causes lots of metabolic changes, including decreasing baseline glucose in the blood and changing sensitivity to insulin. This could be important if we were using an AIMD model to study diabetes for example, as insulin and glucose levels would be carefully measured here.

Additionally, as may be expected, the microbiome of a mouse is generally quite different to that of a human. Some research has shown that 85% of bacterial sequences seen in a mouse represent genera not detected in humans. However, there is also significant similarity in the distal gut microbiome between human and mice at a divisional level, and both have the same two most abundant bacterial divisions (Firmicutes and Bacteroidetes).

Sampling the microbiome

Hopefully you are all aware by now that in order to study the microbiome, in humans or animals, we need to collect poo samples. It is an unglamorous, but crucial aspect of a microbiologist’s job. However, collecting poo presents with its own unique set of challenges. Firstly, people seem to be very protective over their poo! Despite not needing it, and really not even wanting it, people are often quite hesitant to part with their beloved poo. This creates big problems for scientists who want study the microbiome as it impacts their ability to recruit people for their studies. Lucky for us, animals are less fussy about donating their poo, and their pellets can be collected straight from their cage.

Logistically, collecting poo samples from large groups of people is also difficult as the sample needs to be keep cool between defecation and donation. The storage and preparation of stool samples is an area of intense debate, with studies showing vastly different microbial profiles depending on the types of buffers the poo sample is stored with. This is further confounded by the fact that many of the bacteria that live in our guts are anaerobic. This means that they do not use oxygen, like we do, to survive. In fact, many anaerobic species are very sensitive to oxygen meaning that the longer a poo sample stays in fresh air, the more bacteria begin to die.

By maintaining poo samples in an anaerobic chamber it allows scientists to culture bacteria that would otherwise die when exposed to oxygen.

Anaerobic chambers are exceptionally useful in the preparation of poo samples. These are specialised chambers in which the air inside can be manipulated to reflect the environment of the gut. Usually, this means replacing oxygen with combinations of hydrogen, carbon dioxide and nitrogen. This makes them excellent for growing all sorts of bacteria, but also makes them extremely susceptible to explosion!

The other thing to consider when using a poo sample to study the microbiome is whether it accurately reflects the microbiome of the guts. Poo is primarily made of dead cells, debris and waste products… so are the bacteria in poo truly representative of the environment in your gut? To the surprise of many, it actually appears that they are quite similar. A study published in Cell Host and Microbe investigated the biogeography of the gut microbiome in monkeys. They found that the faecal microbiome was almost identical to the types of bacteria found in the large intestine (also called colon). The higher they went, the less similar it became, but even in the small intestine, the major groups of bacteria found were very similar.

The did however find a big difference in the types of bacteria that are found within the lumen of the gut (space inside the tube), compared to those that are stuck on the surface of the intestine. We call these adherent bacteria, and they are widely considered to be very different to the types of bacteria found in the lumen. They are also thought to be more important in regulating gut function, as they reside closer to the underlying tissue of the gut and have greater access to your immune system. So how do we sample these adherent bacteria?

Original image source: https://openi.nlm.nih.gov/detailedresult.php?img=PMC4397497_IJCB2015-481301.001&req=4

Well, one option is to perform a colonoscopy and collect biopsies. A biopsy is when a small piece of tissue is collected from the lining of your gut. With this piece of tissue comes all the adherent bacteria, and these can them be assessed using techniques explained below. Biopsies are a great way to look at adherent bacteria, however they are limited due to two important factors. Firstly, colonoscopies require significant preparation which involves fasting and the use of laxatives to empty the intestines. This makes the job of the gastroenterologist much easier, but it also wipes out all the luminal bacteria. Although proper investigation of the effect of colon preparation on adherent bacteria is lacking, one would also assume that adherent bacteria are also affected during this process.

LEFT: Large intestine in cross section. Adherent bacteria reside in close proximity to the colon wall, shown in blue/purple, whereas luminal bacteria are found in the centre of the colon.

Better solutions for donating poo

There are clear limitations to collecting poo to study the microbiome. Recently, a group from Brisbane has tried to overcome these limitations by developing a way in which a smear of poo on your toilet paper can be used to study the microbiome. This innovative approach eliminates the need for someone to have to poo into a bag, scrape off a small sample and keep it on ice. Instead, the toilet paper wipe is enriched with a special buffer than maintains the bacteria, preserving it for a number of assessments. However, in cases where the entire fresh stool sample is needed (e.g. faecal transplants), this technology is not useful. It also raises the question as to whether the bacteria present on the skin will contaminate the donated sample.

Alternatively, there are newer methods being developed in which capsules are swallowed, collecting luminal contents throughout the whole gut. This approach is exciting, however, it is unclear how to determine which region of the gut is being sampled (i.e. stomach vs colon).

Additionally, an new and exciting Australian invention is the ingestible gas sensor. These small electronic capsules can sense oxygen, hydrogen and carbon dioxide, and can then transmit the data in real-time to mobile phone or a tablet. From these capsules, we may be able to see what types of bacteria are producing certain types of gas, and work out what is normal and abnormal in the gut.

This approach is similar to that used for a group of English scientists who, instead of collecting and sequencing poo, are getting an ‘electronic nose’ to smell it instead! The e-nose was developed in order to mimic the olfactory processes that underpin smell to produce a unique olfactory signature - a little bit like a fingerprint of the smell. Poo with different bacteria will produce different olfactory signatures, based on the types of gas that each bacteria produce. This is potentially much cheaper than sequencing the microbiome, but has not yet been validated for its accuracy.

Take home message

As you can see, there are a lot of things to take into consideration when designing microbiome experiments. What we’ve covered here has really only scratched the surface, and we haven’t even gotten into how the samples are analysed! However, hopefully this makes it a little easier to understand next time you see something about an amazing new microbiome experiment. Let us know in the comments what else you’d like to know about how scientists research the gut or the microbiome.

bottom of page