Key Takeaways:
- Researchers found that rather than washing pathogens away, strong fluid currents act as "guide rails" that align bacteria and accelerate their upstream migration. They discovered that this creates a "two-way invasion" where pioneer cells reach the source within minutes, seeding colonies that spread threefold faster than in still water.
- While bacteria easily navigate the smooth, rounded surfaces common in the human body, the team also found that they are effectively stalled by sharp, angular designs. Implementing these "sharp corners" in thin medical devices like catheters can inhibit bacterial flux and prevent the colonization of sterile environments.
- Their findings reveal that lower-tract infections may signal rapid upstream colonization in organs like the kidneys. Beyond diagnostics, these bacterial strategies offer a blueprint for biomimicry, guiding the design of microrobots capable of navigating complex fluidic systems to deliver targeted therapeutics.
"The UN estimates that, by 2050, common bacterial infections could kill more people than cancer," says Arnold Mathijssen , a biophysicist at the University of Pennsylvania who studies how active particles like bacteria move in fluidic systems. "Bacteria are remarkably fast, adaptive swimmers, capable of moving hundreds of body lengths per second while being subjected to strong fluid flows."
Rather than simply going with the flow, he says, pathogens can actively swim upstream against those fluid currents, a behavior that can lead to severe respiratory, gastrointestinal, and urinary tract infections (UTIs) and the contamination of dental and medical equipment such as catheters. "But how these microorganisms 'go against the grain' through these confined, maze-like environments has remained a mystery."
Now, Mathijssen and colleagues have fabricated nanoscale, multichannel tubes that mimic those found inside the body to reveal how the bacterium Escherichia coli (E. coli) travels upstream to invade and colonize spaces. Their findings, published in Cell Newton , shed light on pathogen motility in complex and often unwelcoming fluid environments, and offer solutions that can be implemented directly in biomedical devices.
"Rather counterintuitively, we discovered that wider channels with faster counterflows are actually more prone to invasion," says Ran Tao , first author and graduate researcher in the Mathijssen Lab in Penn's School of Arts & Sciences . "But these incursions can be inhibited effectively with sharp corner designs."
The team broke down the invasion into four distinct stages, tracked thousands of cells, and combined these data with simulations and mathematical analysis to uncover how bacteria colonize in cavities. In addition, they determined the effect of different conditions—flow strength, channel confinement, corner smoothness—to gauge bacterial proliferation. The combined approach allowed them to predict bacterial flux—the total number of cells moving upstream over time—across different microtube shapes and configurations.
Comparing smooth, rounded corners with sharply angled ones in otherwise identical setups, the team saw how easily bacteria were able to swim against the flow along gentle curves—surfaces that resemble much of the human body—filling those channels quickly. Sharp corners, by contrast, disrupted their motion and stalled their spread, resulting in barely any contamination in angular designs. And in thinner channels, bacteria found it much harder to take the crucial first step of escaping from a colonized area to begin their journey against the flow.
This means that devices such as catheters can combine patient comfort—by being thin—with enhanced safety, by having twists and turns, the researchers note.
The most surprising results emerged when the team looked at fluid flow itself. Conventional wisdom holds that stronger flow should hinder bacteria trying to swim upstream, slowing them down or washing them away. Instead, the current acted like a guide rail. Bacteria aligned with the flow, rode its structure, and reached upstream locations faster than they could in still or less abrasive conditions.
"Within minutes, we see the first cells arrive all the way upstream," says coauthor Suya Que , an undergraduate researcher in the School of Engineering and Applied Science . "Once they're there, those early pioneering cells seed new colonies and create a 'two-way' invasion that advances from both ends."
Rather than colonizing cavity by cavity, like a careful burglar checking each room, bacteria rapidly swim all the way upstream and then help themselves to the whole building. After reaching the front, they form streamer-like bioaggregates that are transported downstream by the flow, rapidly seeding the full length of the device and producing a roughly threefold increase in colonization speed.
Tao says that this has profound clinical implications for infections like UTIs, because the presence of bacteria in a lower part of the urinary tract may signal a much wider problem, possibly higher up in the kidneys, before symptoms fully register. Flow does not protect the body by default, and under the wrong conditions, it can accelerate the problem.
These findings open doors beyond infection prevention, Mathijssen notes. The same principles could guide the design of microrobots that swim against currents to deliver drugs precisely where they are needed, borrowing strategies honed by bacteria over billions of years.
"The mechanisms that they use to reorient against the flow direction and to swim upstream are very similar to that of a microrobot," says Mathijssen. "I think this is a very exciting area in biomimicry—learning from biology—that could help us create better biomedical tools and potentially new therapeutics."
Arnold Mathijssen is an assistant professor in the Department of Physics & Astronomy in the School of Arts & Sciences at the University of Pennsylvania.
Ran Tao is a Ph.D. candidate at Penn Arts & Sciences and graduate researcher in the Mathijssen Lab.
Suya Que is an undergraduate researcher in the School of Engineering and Applied Science at Penn.
Albane Théry of Penn Arts & Sciences is an author on the paper.
This research was supported by the Simons Foundation (Grant Math+X), the Department of Agriculture (Grants 2020-67017-30776 and 2020-67015-32330), the Charles E. Kaufman Foundation (Awards KA2022-129523 and KA2024-144001), the National Science Foundation (Grant, DMR-2309043), and the Penn Undergraduate Research Mentorship program (PURM).
