Researchers at the University of Miami Miller School of Medicine have used artificial intelligence to find a previously unknown protein that helps cells build physical bridges to share resources. The discovery, published in Nature, reveals a new layer of cellular communication that had been hiding in what scientists call the dark proteome - the vast collection of proteins whose functions remain unknown.
The team, led by Dr. Daniel Isom at Sylvester Comprehensive Cancer Center, searched more than 214 million predicted protein structures by three-dimensional shape rather than genetic sequence. This approach uncovered hidden members of the G protein-coupled receptor (GPCR) family, a group of proteins that help cells sense and respond to signals. One protein in particular, TM184C, stood out.
How TM184C builds intercellular bridges
TM184C looked like a GPCR but behaved differently. Most of it appeared inside the cell, within the membranes of intracellular vesicles. The research team watched these vesicles travel along microtubules - the cell's internal transport network - and gather in thin projections that connect neighboring cells.
Those projections functioned as bridges. Through them, cells exchanged metabolites, vesicles, and even mitochondria, the organelles that produce cellular energy. When the researchers disrupted TM184C, cells formed fewer connections. Their shapes changed and vesicle organization broke down, indicating that TM184C helps build and manage these conduits.
"When we saw TM184C-positive vesicles moving through connections between cells, we realized these structures could be routes for substantial material exchange," said Jennifer Arcuri, Ph.D., a Miller School senior scientist and the study's lead author. "That completely changed how we thought about TM184C and made us consider how cells might use these connections to cooperate and compete for resources."
Cooperation, competition, and cancer
The discovery raises a basic question about resource sharing. In healthy tissue, this exchange may help cells survive stress by moving fuel or damaged components where they are needed. But an unequal exchange could let one cell gain at another's expense.
"I think cells coordinate until they have to compete," said Shraddha Chandthakuri, a cancer biology graduate student in the Isom lab. "When the cells are stressed, they may coordinate to redistribute the proteins, organelles and metabolites to support the survival of the population as a whole."
In cancer, where tumor cells often grow with limited oxygen and nutrients, cell-to-cell bridges could provide an alternate survival route. "Understanding their role could give us new insight into how these tumors communicate and potentially reveal vulnerabilities we haven't recognized before," said Bruno Colon, a molecular and cellular pharmacology graduate student studying how these connections work in aggressive cancers like glioblastoma.
An ancient function preserved across evolution
TM184C also appears to regulate autophagy, the recycling process cells use to break down old or damaged parts. When researchers reduced TM184C, autophagy markers increased, suggesting the protein helps tune the system. The team then studied a similar yeast protein called Hfl1. Removing Hfl1 caused problems in yeast cells, but adding the human TM184C protein fixed those problems - a result that shows this function has been preserved across roughly a billion years of evolution.
"For decades, biomedical research has understandably concentrated on the proteins we could identify and understand," Dr. Isom said. "But there is another layer of biology that has remained largely invisible to us. AI gives us a way to start exploring it systematically." He stressed that experimental validation must accompany AI. "AI cannot be blindly trusted, but can lead to really big things in the hands of experts and prepared minds."
Why this matters for science and research professionals
The TM184C finding demonstrates a practical workflow that combines AI-driven structural prediction with wet-lab validation - a model that research teams can apply to other dark proteins. The study, published in Nature, provides a template for moving from computational prediction to functional discovery without relying solely on genetic sequence homology. For researchers in cell biology, oncology, or protein science, the intercellular bridge mechanism also opens a new set of experimental questions about how cells share resources during stress, how that sharing breaks down in disease, and whether those bridges can be targeted therapeutically.
Your membership also unlocks:
