Next, they proved that the bots could swim. In test tubes containing urea, the microbots reached speeds of up to 4 micrometers per second—“one or two body lengths per second,” says Sánchez. (Humans also swim around one body-length per second.)
Then it was time to show that the bots could also kill. But the team agonized over how to prove that they could actually treat an animal’s infection better than by just using passive drops of antibiotics. “That took some time,” de la Fuente says.
In the end, they devised a setup to test two important criteria: that antimicrobial micro or nanobots can treat infected mice and that their active motion plays a central role in that. The team used a needle to carefully scratch the backs of lab mice and introduced a superbug called Acinetobacter baumannii to infect the length of each wound. The process formed dense, hard-to-treat abscesses. On some mice, they dripped a dose of one of the two antibiotics at a single end of the abscess. Those doses had no nanobots, so to clear the infection the drug would have to diffuse on its own from one end of the wound to the other.
Next, a separate set of mice received thousands of antimicrobial bots administered in a tiny droplet. Some mice got bots loaded with LL-37, some got bots with K7-Pol. The team covered each wound with some nontoxic urea, expecting the bots would gobble the fuel and cover more ground.
That’s exactly what happened. Wounds that received antibiotics without bots only improved locally. The number of bacteria fell by 100 to 1,000 times—but only at the extremity of the wound where the dose was delivered. The rest of the wound fared as it would have if it had received no treatment.
But the nanobots carrying either antimicrobial peptide treated the entire wound and reduced the number of bacteria inside the wound 100- to 1,000-fold throughout its length, to levels that an immune system could handle.
And to clinch it all, when the scientists withheld the urea fuel, they found that the antibiotic bots didn’t heal the entire infection. Without that fuel, they only worked locally, just as the drugs without bots had. The fuel was essential—meaning the motor’s motion was essential, the team concluded.
The result is one of the most conclusive examples of the practical uses of nanomotors, according to van Hest. “It’s always very difficult to establish if this is really an effect of the motility of the particle,” he says. “In this case, the proof is direct and clear.”
Douglas Dahl, chief of urologic oncology at Mass General Brigham, calls the nanobots “phenomenal technology.” Like van Hest, Dahl sees a lot of potential for nanobots to keep knee, hip, and even penile implants safe.
Another application would be for treating kidney stones, which often harbor bacterial biofilms along hard-to-reach crevices. “When you go to operate on them, the bacteria can shower inside the patient and make them very sick,” he says. Similarly, urothelial carcinomas that affect the lining of the bladder, ureter, and kidney also grow in tight spaces that complicate treatment. He thinks self-propelled drugs could help doctors attack these elusive tumors and germs. Plus, between the urinary tract, bladder, and kidneys, you’ve got “plenty of fuel,” Dahl notes—enough urea to power a nano army.
In 1966, the sci-fi film Fantastic Voyage imagined a shrunken submarine on a mission through the bloodstream. While Sánchez’s nanobots can’t work in blood that flows much faster than they can move, he still envisions fantastic voyages through the body’s slower-moving fluids, like mucus and the skin’s interstitial fluid. And nanobots still have a way of making people dream about ideas on the border of reality. “As scientists, we’re all inspired by science fiction,” says de la Fuente. “And I think our job sometimes is to try to get those two worlds closer together. What seems science fiction today, hopefully, in a number of years, becomes reality.”