About a new paper from my lab  on why gut bacteria swim, and whether their host cares.
Many bacteria swim. It’s a great way to explore one’s surroundings, run away from toxins, or move toward regions with more food. Over the past several years, as we’ve used 3D microscopy to peer inside zebrafish to uncover the principles that guide gut microbiomes, including our own, we’ve seen different swimming tendencies from different bacterial species (see this post for more discussion). One species in particular is an extremely vigorous swimmer, as well as an excellent gut colonizer. Here’s a movie that I’ve posted before:
Caption: A native bacterial species inside the intestine of a living, larval zebrafish. Each “speck” is a bacterium, in this case a member of a highly motile Vibrio species. The movie is from a single plane within the gut, captured using a home-built light sheet fluorescence microscope. Scale bar: 50 microns.
Nearly every bacterium of this species zooms around the gut like a microscopic wasp. Why? Why bother to swim? After all, the conventional wisdom is that successful gut bacteria adhere to the intestinal wall. (This picture, in our opinion, doesn’t have much data behind it, and is based largely on looking at dead tissue.) Is this species’ swimming just a pointless relic of its prior non-intestinal life, or does swimming help it persist in the gut? Does its animal host notice that it swims? These are the questions we set out to answer.
In addition to swimming, we also looked at chemotaxis, the ability to sense rising or falling concentrations of chemicals to guide the creature’s navigation, for example to search for food. Does knowing which way to go help the bacterium? This question is even less obvious. The larval zebrafish gut is so small that diffusion mixes its contents rapidly. Moreover, swimming is fast, so just zooming along with random turns should allow a bacterium to quickly sample the entire gut.
We reported the answers to these questions, summarized below, in a paper  that came out a few weeks ago in PLOS Biology. It’s a massive article, describing experiments that, in various forms, spanned five years! As with our recent paper on antibiotics, it was a collaboration between my lab and that of Karen Guillemin, led by the amazing duo of Travis Wiles (Biology postdoc) and Brandon Schlomann (physics graduate student), who share co-first authorship.
Before continuing to our methods and our results, a brief advertisement: I encourage you to read my post about a book I’m writing. The topic: Life. All of it!
And now back to our feature presentation…
To see the effects of swimming or chemotaxis on the success of gut microbes, one can knock out genes involved in these processes. The bacterial species we’re focused on swims by rotating a single corkscrew-like flagellum:
One could delete the genes encoding the flagellar proteins — a common approach — but this would alter the outward appearance of the bacterium. Since host proteins can bind to flagellar proteins, this might lead to effects that aren’t related to the actual physical activity of swimming. Therefore, we knocked out part of the motor that drives the flagellar rotation. (“We” being Travis Wiles, whose molecular manipulation skills are amazing — but you’ll see an even better example in a few paragraphs.) A snapshot of the mutant bacteria looks the same for the normal and the mutant strains. Here are electron microscope image of normal (“wt,” wild-type), swimming-defective (“Δmot”) and chemotaxis-defective (“Δche”) cells:
Similarly, we knocked out part of the chemotaxis pathway. The bacteria could swim, but couldn’t alter their swimming in response to sensory cues.
Swimming and sensing are useful behaviors
We found that both the non-motile and non-sensing strains struggled to stay in the intestine. Their populations were an order of magnitude smaller than the wild-type (i.e. normal) strain and they fared poorly in competition with another bacterial species. Imaging showed that the disabled bacteria couldn’t withstand the mechanical transport of the gut; they were easily pushed into aggregates and swept downstream.
The chemotaxis-defective bacteria fared similarly poorly as their non-swimming cousins, again suffering from low abundance susceptibility to expulsion from the intestine.
The host senses swimming bacteria
The bacteria care if they swim; does the animal? Communication between the two takes many forms, including responses mediated by the immune system. We used a transgenic zebrafish in which the “TNF-alpha” immune signaling pathway is linked to fluorescent protein expression — fish cells light up if stimulated. We captured beautiful images of the spatial patterns of immune stimulation — see especially the second part of this video:
We found 100x lower TNF-alpha response in fish colonized by the non-swimming bacteria compared to the swimmers. Keep in mind that the surfaces of the bacteria are unchanged from their normal form; the only difference is their activity. We suspect that the non-swimmers are unable to push themselves into close proximity to the gut lining, and so avoid detection. The chemotaxis-defective bacteria are intermediate between the wild-type and the stationary microbes, in terms of immune stimulation.
There’s more to this story, and more that we don’t understand and hope to explore in the future!
A switch for swimming and sensing
You might be thinking that our comparison of swimming and non-swimming bacteria is still not-quite fair. After all, perhaps the non-swimmers are disadvantaged by being worse at getting into the host in the first place. Perhaps the lower immune response is just a surprisingly sharp sensitivity to low abundance. It would be great to take swimmers already in the gut and hobble them, making them unable to move. Or take non-swimmers and suddenly make them motile.
So that’s what we did. “We” again is Travis, who engineerred remarkable genetic switches that allow genes for motility (or chemotaxis) to be turned “on” or “off” in response to an externally added chemical. The gain-of-function switch is similar to those covered, for example, in my graduate biophysics course. The loss-of-function switch uses a neat Crispr-based interference strategy.
Making the non-swimmers swim, their populations rise. Making the swimmers stop, their populations fall. Our expectations also held for the chemotaxis switches. The host response also tracks behavior: as the non-swimmers switch is flipped on, the TNF-alpha pathway lights up!
The behavior of gut bacteria is in general a mystery. Conventional ways of studying the gut microbiome, looking at DNA sequences or perhaps fixed tissue, are blind to dynamics. Our system, simple though it is, reveals not only vigorous physical activity but its importance: the bacteria benefit from swimming and sensing, and their host notices and responds accordingly.
There’s a lot still to look at. As hinted above, it’s not clear from a biophsical perspective why chemotaxis matters so much. Compared to in vitro studies of chemotaxis, studies in “real” environments are very rare. There’s also a lot left to uncover about the immune response, including its physical behavior — many of these lit up TNF-alpha-expressing cells are themselves motile, crawling around the gut and encountering bacteria and each other.
As described, some of our results were made possible by “switches” that allow us to turn bacterial behaviors on and off. As a physicist, I’m stunned by this. We’re not surprised that we can precisely turn on and off electrical circuits, or other “physical” devices. Now, however, this degree of control exists for the messy world of living things. (I discuss this in my book, by the way.)
I’ll end with a “striking image” that Travis put together, that adorned the top of the PLOS Biology web site for a week after our paper came out.
A quick, crude watercolor of part of a flagellum and its motor.
— Raghuveer Parthasarathy, April 10, 2020
 T. J. Wiles^, B. H. Schlomann^, E. S. Wall, R. Betancourt, R. Parthasarathy, K. Guillemin, Swimming motility of a gut bacterial symbiont promotes resistance to intestinal expulsion and enhances inflammation. PLOS Biology.18, e3000661 (2020). [^ co-first authors]