Text from the interactive explanation

No room at the table

Lauren C. Radlinski, Andreas J. Bäumler

Gut microbiology · A Science Perspective, 2023

Trillions of microbes share the food that reaches your gut. When every dish a newcomer could eat is already taken, a dangerous invader has nothing to live on.

Invader blocked: every dish it needs is taken.

Invader population, last 40 s (log scale; dashed line = a foothold)

Live, illustrative simulation of the principle described in the paper. The menus and rates are invented for illustration.

“…improved predictive tools … could allow clinicians to administer tailored probiotic communities that prevent difficult or impossible-to-treat infections from occurring in the first place.”

Imagine leaving hospital after a course of antibiotics with a prescription for exactly the microbes your gut needs to keep a dangerous strain out. To design such a community, you would first need to know what makes a gut community protective at all.

That turns out to be a question about dinner.

1

Bouncers made of bacteria

The bacteria living in your gut, known together as the gut microbiota , do more than help digest food. One of their most important jobs is to keep disease-causing microbes from moving in. Microbiologists call this colonization resistance.

Germ-free mice

Mice raised without any microbes can live in sterile isolators, but they quickly succumb to infection once exposed to the outside world.

After antibiotics

Antibiotics disrupt the gut microbiota, which brings a risk of life-threatening opportunistic infection: an infection by a microbe that exploits a weakened defence, such as Klebsiella pneumoniae.

A missing explanation

Gene sequencing has catalogued gut communities in great detail, and more diverse communities have long been linked with better health. Yet no one could clearly explain why, or say what a “healthy” microbiota really is.

Spragge and colleagues went looking for the ecological rule behind these observations, rather than a list of lucky species, which differ from person to person anyway. The Perspective by Radlinski and Bäumler explains their answer. It is about food.

If you pressed Give antibiotics above, you have already seen the rule at work: the moment the residents were knocked back, food piled up on the plates and the gold invaders moved in.

2

What's on the invader's menu?

Bacteria live on carbon sources: sugars and similar molecules that they break down for energy and building material. Each species can use only some of them. Think of that as its menu. Our table holds 14 dishes, and Klebsiella can eat 8 of them, marked in gold.

The gut is also a slowly flowing river. Everything in it is gradually swept onward, so a newcomer can only settle if it multiplies faster than it is washed out . How fast it multiplies depends on how much food is left on its dishes.

Try it: the taste test

Tap a dish to clear it, as if a resident microbe had eaten it down. How many of Klebsiella's eight dishes must you clear before it can no longer keep up with the flow?

Klebsiella multiplies 11× faster than it is washed out, so it settles.

In a real gut, nobody clears dishes by hand. They are eaten by other microbes. So which microbes should sit at the table?

3

Choose the bodyguards

Two kinds of guests are waiting to be seated.

E. coli: the close relative

A cousin of Klebsiella with a very similar appetite. It eats 6 of the invader's 8 dishes.

A diverse crowd

Six unrelated gut bacteria, named here by colour. Each eats two dishes, mostly ones Klebsiella ignores. Between them they cover 4 of its 8.

Predict first: which table will keep Klebsiella out?

Pick one, then test the four tables below.

Choose a table to test.

Wheel: Klebsiella's eight dishes. Coloured bands show which seated guest eats each dish; glowing gold wedges are open seats. The centre compares its growth with the washout.

Invader population in this run (log scale)

Or seat guests one by one:

Each table you test fills in one panel. Blue disc: the pathogen's niche, meaning everything it can eat. Purple: E. coli's niche. Small circles: the diverse community. Redrawn after the figure “Resisting colonization by pathogens” in Radlinski & Bäumler (graphic: K. Holoski/Science).

4

Why it takes both

E. coli shares Klebsiella's core dishes, 1 to 6, but leaves dishes 7 and 8 untouched. As the taste test showed, one open dish is enough for the invader to settle. The crowd eats dishes 7 and 8, but most of the core stays full. Only together do they leave the invader nothing.

That is what Spragge and colleagues found with real bacteria. They analysed which carbon sources each member of defined bacterial communities could use, both with computer models and with laboratory tests. Complex communities protected against Klebsiella pneumoniae and Salmonella Typhimurium by using up all the nutrients the pathogen needed to grow . Nutritional overlap was greatest between these pathogens and their relatives, such as E. coli, which made E. coli a keystone species for colonization resistance.

But it was a keystone only in the context of a diverse community. Radlinski and Bäumler call this a “higher-order” effect : how two species interact depends on the wider community around them. It also suggests why diversity and health have so long gone together. A diverse community is better placed to empty every dish an invader might need.

The scenes on this page run a standard consumer–resource model, a textbook way to describe species competing for food in a flowing system.

Food on dish j arrives and is washed away at rate D, and is eaten by every species i on whose menu it appears: dRj/dt = D·(Sj − Rj) − Σi Ni · Vi · Rj/(K + Rj)

Each population grows on what it eats and is washed out at the same rate D: dNi/dt = Ni · ( Σj ∈ menu Vi · Rj/(K + Rj) − D )

Invasion rule. A newcomer can settle only if its growth on the residents' leftovers beats the washout: Σj ∈ menu V · Rj/(K + Rj) > D. The number in the centre of each wheel is that growth divided by D.

Menus, rates and time scales are invented to illustrate the principle. They are not measurements from Spragge et al. Cell numbers on the tables follow the model's populations on a logarithmic scale.

5

Tailor a team

Different invaders have different menus. Choose a target, then seat the smallest team that leaves it no room. Gold rings on each card mark the target's dishes.

Seat some guests.

Team size: 0

The study did this at scale. Spragge and colleagues searched public repositories of genome sequences from human gut bacteria for communities whose genes overlapped with those of the relevant pathogens, and they accurately predicted which species curb a pathogen's growth. In principle, then, a probiotic community could be tailored to patients at high risk of a particular infection, for instance after a course of antibiotics in hospital.

And the two shelf probiotics? Many of today's probiotics are simplified mixtures of species with only a loose link to health, and whether they help after antibiotics is disputed, partly because they can get in the way of the gut community's own recovery. The study found that the nutritional needs of pathogens and probiotic species are often very different, much like the menus of our two shelf strains.

6

The sweetener loophole

A protected table has one weakness: food that no resident can eat. In the study, adding the sugar alcohol galactitol to mouse feed, a nutrient that only the pathogen could use, let it get past an otherwise protective microbiota. Try it on a fully protected table.

Blocked: the full community leaves Klebsiella nothing.

A dish no resident can eat is a seat no resident can take. This fits recent reports that sugar alcohols fuel blooms of pathogens even when the microbiota is intact. Sugar alcohols such as mannitol, isomalt and sorbitol are increasingly popular as low-calorie sweeteners. As the authors put it, they cut the calories in food but “may come at the cost of providing pathogens a fitness advantage to invade and disrupt an otherwise stable community.”

7

What the study actually showed

Reported Spragge et al., as described in the Perspective

Mapped the carbon sources used by individual members of defined bacterial communities, with computer (in silico) analysis and laboratory (in vitro) growth tests on Biolog plates.

Complex communities gave colonization resistance by depleting all the resources needed by an opportunistic pathogen (K. pneumoniae) or a pathogen (S. Typhimurium).

Nutritional overlap was greatest between these pathogens and related species such as E. coli, a keystone species.

The keystone mattered only within a diverse community, a higher-order effect.

Supplementing mouse feed with galactitol, usable only by the pathogen, let it get past a protective microbiota.

Mining human-derived genome repositories for sequence overlap with relevant pathogens accurately predicted species that curb pathogen growth.

Illustrative Built for this page

The table, its 14 dishes, every species' menu, and the colour-named crowd are invented to show the principle.

Populations, food levels and timings come from a simple consumer–resource model (see “Go deeper” in chapter 4), not from measured data.

The Salmonella menu and the shelf-probiotic menus are illustrative.

In the model, residents are better at eating any single dish than the generalist invader. That assumption is what lets a complete set of residents shut the invader out.

8

What comes next

The Perspective closes with questions that the next experiments will need to answer.

Invaders that change the menu

Salmonella actively reshapes its surroundings. The gut inflammation it triggers creates new carbon sources and raises oxygen and nitrate levels, letting it out-compete resident bacteria through aerobic respiration, the oxygen-using way of extracting energy. E. coli shares several of these oxygen- and nitrate-dependent pathways, which matter for resistance during inflammation.

A milder test so far

The mouse infections used an avirulent mutant of Salmonella, one unable to cause disease, so the extra oxygen and nitrate of an inflamed gut were not part of the test. How the findings carry over to pathogens that engineer their own environment remains to be worked out.

Beyond the gut

Do the same simple rules hold in other parts of the body that host a microbiome? And how do host responses such as inflammation shape these higher-order interactions?

9

Back to the hospital

Return to the patient from the start. Suppose a strong course of antibiotics has emptied their gut's table, and a hospital strain of Klebsiella is around. Here are three choices, each run on the same empty table.

Wait Shelf probiotic Tailored · invader population (log scale)

Tailored team: E. coli, Coral and Mint.

In this illustrative scene, only the tailored community leaves Klebsiella nowhere to eat. That is the possibility the Perspective points to. With antibiotic resistance widespread, predictive tools like the one Spragge and colleagues built could let clinicians give tailored probiotic communities that stop hard-to-treat infections before they start.

The principle underneath is simple ecology: leave no room at the table.

Sources & credits

The paper explained here: Lauren C. Radlinski and Andreas J. Bäumler, “No room at the table: The gut microbiota prevents infection by crowding out pathogens.” Science 382 (6676), 1244–1245 (2023). doi:10.1126/science.adl5891

The study it discusses: F. Spragge et al., Science 382, eadj3502 (2023).

Four-panel comparison: redrawn in this page's style after the figure “Resisting colonization by pathogens” in the Perspective (graphic: K. Holoski/Science). No figures or photographs from the article are reproduced.

Simulations: every animated table on this page is an illustrative consumer–resource model built to demonstrate the principle described in the paper. Species menus, rates and timings are invented, and the colour-named bacteria are fictional stand-ins for unrelated gut commensals.