When food poisoning hits, the idea of consuming anything, much less a concoction of bacteria, probably turns your stomach.
But for the microbiologists and genetic engineers in Harris Wang's Columbia University laboratory, a therapeutic bacterial brew could be the only way to disarm some of the most dangerous intestinal pathogens.
The idea to use good bacteria to thwart bad bacteria in the gut microbiome isn't a new one. But probiotics often fail to colonize the gut. "That's one of the challenges with live bacterial therapy. Because they have such a transient nature, their effects are also transient," says Wang, who is interim chair of the Department of Systems Biology at Columbia University's Vagelos College of Physicians and Surgeons.
To make a more long-lasting and effective treatment, Wang has taken a cue from the way bacteria naturally swap genetic material with members of their own species and even unrelated ones.
Horizontal gene swaps
This process, known as horizontal gene transfer, occurs frequently between microbes in the gut, often with negative consequences for human health. "This is how genes that confer multi-drug resistance and virulence spread," Wang says. "But it's also a driver for genetic innovation and allows for microbes to evolve much more quickly."
Wang's idea was to engineer bacteria to transfer genes into harmful bacteria that render them harmless. The transferred cargo includes a genetic editor that locates and disables toxic genes, essentially disarming harmful bacteria by editing their genomes.
Because the targeted bacteria don't die, the therapy could sidestep a major issue with today's antibiotics, which are designed to kill pathogens rather than disable them. "With antibiotics, pathogen elimination is never 100% effective, and variants inevitably emerge that are even less treatable," Wang says. "Leveraging natural gene transfer could be a more effective approach without generating resistance."
High on Wang's list of harmful bacteria to target are those that produce the Shiga toxin, which spread through contaminated water, food, and surfaces, causing severe food poisoning in nearly 1 million people each year around the world.
Shiga-producing bacteria represent a unique challenge for medicine. Antibiotics can kill the bacteria but also cause a massive outpouring of toxins from the dying cells. "Giving antibiotics to these patients only worsens their situation," Wang says. "There's really nothing other than keeping people hydrated and providing supportive care."
Because of the heightened risk, Wang's team took extra care to design therapeutic bacteria that didn't accidentally kill the pathogens while editing them. The CRISPR gene editor was deemed too risky, because DNA cleavage caused by the editor can result in bacterial death and toxin release. Instead, the Wang team deployed a gene editor recently developed by Columbia's Samuel Sternberg that works without cleaving DNA.
The therapeutic bacteria-dubbed BACTRINS-also transfer a gene encoding a synthetic antibody designed to interfere with the pathogen's attachment to the gut.
Defanging harmful gut bugs
In tests with mice, the results were striking.
As the team reports in Nature Biomedical Engineering, BACTRINS transferred its cargo to Shiga-producing bacteria with 100% efficiency, sharply reduced levels of the toxin in the mice, and increased survival. The treatment also acted like a vaccine, fully protecting mice from subsequent infections.
"Programmable microbiome editing gives us surgical precision, defanging pathogens like Shiga-toxin bacteria and turning harmful gut bugs into protective allies," says Carlotta Ronda, one of the study's authors, now a principal investigator at the University of California, Berkeley."Programmable microbiome editing gives us surgical precision, defanging pathogens like Shiga-toxin bacteria and turning harmful gut bugs into protective allies."
In addition to disabling "bad" bacteria, BACTRINS could also be used to deliver new capabilities to the microbiome's "good" bacteria, including the production of drugs that could boost the immune system or control metabolic diseases like diabetes.
"We see this work establishing a versatile platform for targeting a broad range of pathogens and advancing the delivery and expression of novel therapeutic payloads," says study author Tyler Perdue, who conducted much of the research as a graduate student in the Wang lab.
The system needs some fine-tuning before it can be deployed in people, but it "opens the door to next-generation living therapeutics," Ronda says.
References
Additional information
The results were published in Nature Biomedical Engineering on July 18 in a paper titled, "Precise virulence inactivation using a CRISPR-associated transposase for combating Enterobacteriaceae gut pathogens."
All authors (from Columbia unless noted): Carlotta Ronda (now at University California Berkeley), Tyler Perdue, Logan Schwanz, Diego Rivera Gelsinger, Leonie Brockmann, Andrew Kaufman, Yiming Huang, Samuel H. Sternberg, and Harris H. Wang.
The Columbia scientists were supported by the National Institutes of Health (grants DP2HG011650, R21AI68976, R01EB031935, R01EB031935, R01AI132403, R01DK118044, and R21AI146817); National Science Foundation (MCB-2025515); Department of Defense (S-168-4X5-001); Burroughs Wellcome Fund (1016691); Irma T. Hirschl Trust, a Schaefer Research Scholars award; a Pew Biomedical Scholarship; a Sloan Research Fellowship; an Irma T. Hirschl Career Scientist Award; the Simons Society of Fellows (#527896); the Columbia University Irving Medical Center Dean's Office; and the Vagelos Precision Medicine Fund.
Patent applications describing the CAST and gene delivery technologies have been filed by Columbia University. Samuel Sternberg is a cofounder and scientific adviser to Dahlia Biosciences, a scientific adviser to CrisprBits and Prime Medicine, and an equity holder in Dahlia Biosciences and CrisprBits. Harris Wangis a scientific advisor of SNIPR Biome, Kingdom Supercultures, Fitbiomics, VecX Biomedicines, Genus PLC, and a scientific co-founder of Aclid and Foli Bio, none of whom were involved in the study.
