A tiny molecular syringe with bizarre origins could overhaul one of the thorniest problems in medicine: getting drugs to their target destinations inside the body. The source? Bacteria living in the gut of insects.
The brain child of Dr. Feng Zhang at the Howard Hughes Medical Institute and Broad Institute, the spring-loaded nanomachine looks a bit like a rocket ship. Once docked, an injector shoots down to penetrate the cells and deliver precious payloads.
When further developed, the molecular injectors could shuttle cancer immunotherapies only to tumor cells, sparing healthy ones and limiting side effects. The system can also safely tunnel into the brain—a notoriously difficult organ for drugs to access—potentially shuttling in proteins that could help diagnose strokes, Alzheimer’s, and other neurological disorders.
Published in Nature, the injector was inspired by the bacterial kingdom. Zhang is no stranger to exploring the dark matter of the bioverse. Best known for his seminal work on CRISPR-Cas9 gene editing, which originated as a bacterial defense system against viruses, Zhang has long taken hints from evolution to craft next-generation biotechnological wonders.
This time, however, his team brought another collaborator into the mix: AlphaFold.
Developed by DeepMind, the AI made headlines for its uncanny ability to predict protein structures. Putting the tool to use, the team optimized a core part of the bacterial injector, making it switch from their preferred target—insect cells—to a variety of mice and human ones.
Several proof-of-concept studies in both cultured cells and mice showcased the new syringe’s prowess. One experiment delivered a toxin to cancer cells without harming others. Another injected Cas9—the protein “scissors” in the gene editing tool CRISPR—into cultured human cells and edited the target genes with high efficiency.
This ability to plug and play makes the system a delivery powerhouse. “We show that just by putting a tag onto the protein, we can load different types of proteins into these needles,” said Zhang.
“Having the ability to deliver particular proteins into specific cell types would offer tremendous potential for research in the life sciences, as well as for the treatment of disease,” said Charles Ericson and Dr. Martin Pilhofer at ETH Zürich, who were not involved in the work.
The system, when combined with others, sets the foundation for a powerful mix-and-match toolbox for both research and medicine. Although currently only capable of shuttling proteins, co-opting other natural molecular syringes could expand the system to DNA and other biomolecules.
“It’s still early days for this as a technology,” said Zhang.
Imagine drug delivery as DoorDash. You want your order to come only to you, not your neighbors, and with the food intact.
It sounds trivial, but it’s a task that’s hard to achieve with drugs and gene therapy. Medication in the form of pills, patches, or intravenous needles—think saline bags or chemotherapies—enter the bloodstream. The result is that they flood different organs and tissues and often cause side effects.
In stark contrast, another problem is that some drugs can’t burrow into their targets. Cells are fortresses surrounded by a double-layered fatty membrane, with mechanisms that sometimes actively spit out unwanted intruders. When those intruders are gene therapy elements or therapeutic proteins, the cells’ defense system becomes a massive headache.
Scientists have devised ways to bypass these defenses. One is using harmless viruses to smuggle in vaccine materials. Another is lipid nanospheres, which are made of little fatty bubbles. Once merged with the cell, the bubbles “burst” and release the payload. While foundational for genetic engineering, these systems aren’t as precise as we’d like. Going back to the DoorDash analogy, the dasher will give you some of your order—while bringing the rest to your unsuspecting and unwilling neighbors.
In the new study, Zhang threw away the playbook and went completely outside the box. He and his colleague Joseph Kreitz tapped into a molecular syringe crafted by evolution.
The unexpected resource is a bioluminescent bacteria called Photorhabdus asymbiotica, which lives in the gut of insects. They come heftily armed: each is equipped with tiny molecular syringes—roughly 100 nanometers long—with “feet” that grasp host cells. Once docked, a plunger drives through the cell’s membrane, shooting in a toxin that kills the host—and in turn allows the bacteria to escape and colonize other cells.
The dangerous-sounding mechanism—dubbed a contractile injection system, or CIS—hardly seems fit for a safe delivery system. But one quirk caught the team’s eye: bacteria injectors usually only work with other bacteria, not animal cells. So why not rejigger the Photorhabdus syringe to also inject human cells?
The team first honed in on a part of the injector called tail fibers. These “tentacle-like things” help the nanomachine latch onto cells, explained Zhang. The key is matching the receptors, or docking stations, on the surface of cells. Each cell type has a myriad of docks unique to their biological character—a neuron, for example, has several that are massively different from those of heart cells. Those from different living creatures are even more divergent.
So it’s no surprise that the syringes, designed to work in insect cells, failed in human ones. Knowing that tail fibers are the crux, the team brought in a new collaborator: AlphaFold. Using the AI, the team generated a 3D model of the tricky protein found in a region that guides the injector towards insect cells.
They then genetically modified this region, chopping off the tail fiber’s end and adding different protein chunks to guide the injector towards specific mouse and human cells.
“[AlphaFold] gave us the information we needed to make a new delivery strategy that can be changed to target different cells,” said Kreitz.
Mix and Match
The team tested their programmable injectors with several experiments.
In one, they loaded the syringe with a protein that, once injected, caused human cells in culture to glow a vibrant green in the dark. A similar syringe was reworked to track down cancer cells dotted with epidermal growth factor receptor (EGFR) on its surface. Loaded with toxins, the treatment killed nearly all the cells with the receptor but spared others. Similarly, the team easily delivered Cas9 into a variety of human cells, which when supplied with a guide RNA edited the genome at predicted points.
Finally, in the ultimate test, the team injected the system into the hippocampus of mice. Infused with a fluorescent protein, the cells glowed a bright green. Importantly, although derived from bacteria, the injectors didn’t trigger an immune response.
The system isn’t perfect. Although efficient in tested tissues, the team is hoping to expand its range to different types of tissues and disease models. Another goal is to hunt for other natural injectors and potentially grow them into a whole family of delivery tools—in a vein similar to CRISPR’s growth. For now, the system only carries proteins. But further engineering could allow specific delivery of DNA, RNA, and other biomolecules, and perhaps even control their dosage.
“It’s still early for this approach, but I think it’s really important to explore [the system’s ability] to be able to treat many different types of diseases that affect human health,” said Zhang.
Image Credit: Joseph Kreitz, Broad Institute of MIT and Harvard, McGovern Institute for Brain Research at MIT