Engineering the Microbiome

Home , Protein production

The gut microbiome outlook LEWIS HOUGHTON/SPL A researcher tests donor stool before it is used for faecal microbiota transplantation. Engineering the microbiome

Modified bacteria and carefully formulated microbial communities could form the basis of new living treatments. By Claire Ainsworth

lthough it happened almost a decade transplants were proving so successful that aspects of our physiology — from conditions ago, Willem de Vos still vividly remem- it was no longer ethical to continue to give such as obesity to how the immune system bers his colleagues being told to halt people in the control group the conventional functions and even mental health. The suc- the clinical trial they had been run- antibiotic treatment with which the trans- cess of FMT in treating C. difficile also shows ning. De Vos was part of the team plants were being compared. “That showed that, in principle, the ecology of the gut can be Aconducting the first randomized clinical trial us that it worked and why it worked,” says manipulated to treat disease. Now, scientists of faecal microbiota transplantation (FMT) de Vos, a microbiologist at the University of are attempting to engineer gut microbiota that — faeces from healthy donors were used as a Wageningen in the Netherlands and the Uni- will allow them to do just that. last-resort treatment for people with a dev- versity of Helsinki in Finland. The antibiot- Synthetic biologists are working at the level astating, recurrent gut infection caused by ic-treated patients who relapsed were given of individual species, engineering gut bacteria the bacterium Clostridium difficile. About a the transplant, which cured them. not only to deliver therapeutic payloads but year in, the data and safety monitoring board The C. difficile story is one of a growing list of also to monitor and respond to conditions overseeing the trial had seen enough: the examples of how the gut microbiome shapes inside the body. Meanwhile, synthetic ecol- trial needed to end. But it wasn’t because the our biology. The community of microbes that ogists are looking at the gut as an ecosystem therapy didn’t work — quite the opposite. The lives in the gut has been associated with many and assembling communities of microbes that

interact to produce substances or behaviours signal can be detected in stool samples, rais- from organisms acting alone. for medical benefit. Both approaches are in ing the possibility of using the bacterium as a These emerging properties of the gut micro- their infancy, and there are challenges to get- living diagnostic test for inflammatory bowel biome have a profound effect on our biology, ting them into the clinic. Yet the technologies disease — which is often transient in nature such as by producing vitamins or molecules are already proving to be powerful tools, allow- and, therefore, hard to detect in the clinic. that modulate our immune responses. To ing scientists to explore the complex microbial The bacteria formed a stable colony in the understand these interactions and to devise interactions in our internal ecosystem. guts of mice for six months and responded to new therapies, researchers are building com- experimentally induced gut inflammation3. binations of different bacteria known as syn- Bespoke bacteria Importantly, engineered bacteria that can thetic ecosystems. For the most part, these Engineering individual microbes has an remember other kinds of environmental signal ecosystems are made up of naturally occurring impressive array of potential applications. would allow researchers to explore conditions bacterial strains, although some scientists are Gut bacteria have been altered to produce in different regions of the gut — something that experimenting with ecosystems containing therapeutic molecules to treat metabolic con- is hard to do with conventional stool samples. genetically engineered microbes. ditions, kill pathogens and trigger immune “What we really would like is the bacteria to be From a therapeutic point of view, synthetic responses to cancers. A strain of Escherichia like detectives and tell us what’s going on as ecosystems have a number of potential advan- coli engineered to produce the proteins they pass through,” says Silver. tages. FMT currently relies on faecal matter needed to correct rare metabolic deficien- Getting a genetic circuit to work in the lab provided by donors. Stool samples contain cies is now in clinical trials. And in 2018, a is hard enough. Translating that to the messy, team in Singapore revealed gut bacteria that competitive environment of the gut microbi- “What we really would it had engineered to stick to colon cancer cells ome presents an even greater challenge. Any like is the bacteria to be and secrete an enzyme that converts a sub- modification that imposes an extra burden stance naturally found in vegetables such as — say, extra protein production — on a bacte- like detectives and tell us broccoli into a molecule that inhibits tumour rium puts it at a disadvantage, resulting in that what’s going on.” growth. When given to mice with colon cancer, organism either being out-competed or ditch- the treatment shrank tumours and reduced ing its engineered function to survive. Partly recurrence1. Bacteria can even be engineered for this reason, researchers have struggled to highly complex mixtures of microbes that to sense signs of disease and respond by pro- get many engineered bacteria to make the leap vary from donor to donor, and each must be ducing therapeutic molecules. For example, in from test tube to animal models. Scientists screened for pathogenic microbes. If FMTs 2017, researchers took a gut bacterium com- are now working on ways around this; Silver, could be stripped down to just the key spe- monly used as a probiotic and gave it the ability for example, is using genetic elements that cies needed to treat people, simplified path- to detect communication signals produced naturally place a minimal burden on the cell. ogen-free mixtures of these selected microbes by pathogenic bacteria. The probiotic bacte- The final hurdle will be showing that engi- could be grown in the lab. Synthetic commu- rium then produces an antimicrobial molecule neered bacteria are effective and safe. What’s nities would offer a standardized therapeutic in response. The researchers showed that it more, unlike conventional drugs, engineered with a known composition, and would lift the helped clear infections in worms and mice2. bacteria could spread into the environment reliance on finding suitable donors. Studies such as this show the potential of and share their DNA with other bacteria. Research, including a few studies in people, live therapeutics, but so far the engineered Although the chances of them surviving in suggest that this approach could work. Mix- bacteria are comparatively straightforward the wild are thought to be low, the possibility tures of selected bacteria isolated from stool systems — they produce a therapeutic mole- of unforeseen consequences (not to mention samples have shown promise in treating peo- cule either at a constant rate or in response to the need to secure public acceptance and regu- ple with C. difficile. And it’s not just infections an environmental signal. Now, researchers are latory approval) has led researchers to explore that could be tackled, but also conditions such looking to broaden the scope of engineered a number of options to contain engineered as inflammatory bowel disease. In 2013, a team microbes and engineer bacteria with DNA con- bacteria, including kill switches that force led by scientists in Japan identified a commu- taining more complex elements designed to bacteria to kill themselves with a toxin if their nity of human gut microbes that could pro- work like electronic circuits. This is the realm engineered circuits turn faulty or if they leave mote the activity of inflammation-damping of synthetic biology, a discipline that aims to the body. immune cells called regulatory T cells, and apply engineering principles — such as stand- showed that this could ameliorate inflamma- ardized, modular components — to biological Constructing communities tory bowel disease in mice4. As well as devel- systems. While some researchers engineer individual oping therapies, stripping down conventional These complex feats of engineering are bacteria, others are turning their attention to FMTs is allowing scientists to work out which allowing bacteria to do simple computational groups of microbes. Just as a city functions bacteria in stool transplants are exerting a tasks, such as remembering a one-off stimulus as a result of many people doing different therapeutic effect — something that de Vos and long after it has passed. For example, a team jobs, the gut is a hive of interactions between his colleagues are exploring in conditions such of synthetic biologists led by Pamela Silver at myriad microbes carrying out different func- as inflammatory bowel disease and metabolic the Wyss Institute for Biologically Inspired tions. Some interactions are metabolic — one syndrome. Engineering at Harvard University in Bos- bacterium might produce something that One drawback of this stripping-down ton, Massachusetts, engineered a bacterium another consumes, for instance. Others approach is that it limits the applications of the to detect a chemical produced by inflamed are ecological, such as when one microbe synthetic community to functions that already gut cells. In response, the bacteria secrete a inhibits the growth of another. By working exist. There might be situations in which you molecular signal, and continue to secrete it together, communities of microbes produce would want to create a community with a even if the gut inflammation dies down. The molecules or behaviours that would not arise new function, such as producing a vitamin

The gut microbiome outlook

researchers hope that studies such as this will allow them to design minimal microbiomes with defined therapeutic properties, such as producing butyrate or vitamins. Perhaps the eventual application of microbiome engineering would be to combine synthetic biology and synthetic ecology. Scientists would create communities containing genetically-engineered microbes, the collective behaviour of which would deliver a therapeutic benefit. One advantage of this approach is that it would let engineers dis- tribute different metabolic tasks between different bacteria. This means all the physi- ological stress of making a drug or a vitamin would not be placed on just one bacterium. A number of teams have made progress in this area, including exploiting a system that bacteria use to detect the presence of other

bacteria and to modify their gene activity in UNIVERSITY HARVARD AT INSTITUTE WYSS Synthetic biologists have engineered bacteria that remember the presence of a chemical and response. Researchers are using this feature, secrete a molecular signal that allows them to be identified. known as quorum sensing, to control the behaviour of mixed populations of bacteria, or degrading a toxin. Creating new functions only consuming these compounds, but were to, for example, allow bacteria that compete requires designing from the bottom up — also feeding molecules they had made back to with each other to co-exist and form a stable testing different combinations of microbes, A. muciniphila and, in the case of butyrate, a population. including those that don’t normally co-exist fatty acid needed by the cells of the gut lining, The potential paybacks of engineering the in nature, until one gives the desired outcome. to their host. gut microbiome are immense, but so are the Doing this by trial and error in lab experiments Researchers are also gaining new insights challenges to reaching this goal. Of all the soon becomes unwieldy, so instead research- into the relationships between microbes and human microbiomes to take on, the gut micro- ers have turned to computer modelling. between microbes and their host from the cre- biome is by far the largest and most complex. The aim here is to predict the emergent ation of minimal microbiomes — constructed Much remains to be learnt about its denizens, properties of a microbial community, based on microbial communities containing the small- their genes and their interactions. And that’s expected interactions between the microbes est number of species needed to create a stable before you get started on what the human host present. A team led by Elhanan Borenstein at ecosystem. A 2016 study showed how combin- brings to the party. Indeed, there is so much Tel Aviv University in Israel created computer ing a minimal microbiome with comparative variation between individuals that it’s still not models of the metabolic reactions inside clear what a ‘healthy’ gut microbiome looks individual microbes, and then modelled how “We’re making progress and like (see page S6). these would behave in the presence of another Even so, the potential payoffs are motivating microbe’s metabolism5. By simulating interac- learning a lot of interesting the scientists to aim high. Borenstein hopes tions between pairs of microbes, they showed biology on the way.” one day to take information about an individ- how new metabolic products emerged that ual — the microbes in their gut, their physiol- wouldn’t be seen if the microbes acted alone. ogy, their diet and their genome — and use it to Models can simulate ecological interactions genomics can lead to the design of a microbial build a full-scale computer model of their gut too, such as how the abundance of one microbe community with a desired property. Bärbel microbiome. Such an advance might make it affects the abundances of others. This can help Stecher at the Ludwig-Maximilians University possible to design personalized interventions scientists to design microbial communities of Munich in Germany and her team developed to treat or prevent disease. that are stable and therefore persist over time. the Oligo-MM12 minimal microbiome — a col- “This is not something we’ll get to in a lection of 12 gut microbes that helps to pre- year, or two or five,” Borenstein admits. “But It’s the ecology vent Salmonella enterica from colonizing the we’re making progress and learning a lot of Computer modelling and lab-grown guts of mice lacking any bacteria of their own7. interesting biology on the way.” communities allow researchers to gain a bet- The 12 bacterial species excluded Salmonella ter understanding of how microbes in natu- almost, but not quite, as well as a conventional Claire Ainsworth is a freelance science ral communities in the gut interact — both microbiome. By using genomics to compare journalist in Hampshire, UK. with each other, and with their human hosts. their minimal microbiome with a complex one, De Vos’s team grew four different bacteria the researchers singled out the ecosystem 1. Ho, C. L. et al. Nature Biomed. Eng. 2, 27–37 (2018). 2. Hwang, I. Y. et al. Nature Commun. 8, 15028 (2017). that usually live together in the mucus layer functions that were missing from their com- 3. Riglar, D. T. et al. Nature Biotech. 35, 653–658 (2017). 6 lining the gut . One species, Akkermansia munity, added three more bacterial species 4. Atarashi, K. et al. Nature 500, 232–236 (2013). muciniphila, breaks the mucus down into that could fill the gap, and produced a com- 5. Chiu, H.-C., Levy, R. & Borenstein, E. PLoS Comput. Biol. 10, e1003695 (2014). compounds that other bacteria consume. The munity that was as good as the conventional 6. Belzer, C. et al. mBio 8, e00770-17 (2017) team showed that the other bacteria were not one at keeping Salmonella out. Ultimately, 7. Brugiroux, S. et al. Nature Microbiol. 2, 16215 (2016).