Most of the human body's proteins remain a mystery. Miami researchers just used AI to find one hiding in plain sight. It may help cells cooperate, or compete, inside a tumor.
The known protein universe contains hundreds of millions of members. Most remain largely unexplored, a vast biological darkness scientists have only begun to search.
Searching by Shape Instead of Sequence
University of Miami researchers at Sylvester Comprehensive Cancer Center took an unusual approach to that search. Instead of scanning genetic sequence, they used AI to compare the three-dimensional shapes of more than 214 million predicted proteins.
"For decades, we have largely explored protein biology using sequence as our guide," said Dr. Daniel Isom, the study's senior author and an associate professor of molecular and cellular pharmacology. "We wanted to know what biology we might be missing if we searched by three-dimensional structure instead."
That structural search, published in Nature, turned up hidden members of the G protein-coupled receptor family. That's a group that helps cells sense and respond to signals outside their walls. One protein in particular stood out.
A Protein That Builds Bridges
TM184C looked like a typical GPCR at first glance. It behaved differently. It showed up mostly inside the cell, packed into vesicles traveling along microtubules. These gathered at thin projections connecting neighboring cells.
Those projections acted like bridges. Cells used them to exchange metabolites, vesicles and mitochondria, the structures that generate a cell's energy.
When researchers disrupted TM184C, cells formed fewer of these connections. Their shape changed, and vesicle organization shifted too.
"When we saw TM184C-positive vesicles moving through connections between cells, we realized these structures could be routes for substantial material exchange," said Dr. Jennifer Arcuri, a Miller School senior scientist and the study's lead author.
Cooperation, Competition and Cancer
The discovery raises a real question. When cells share resources this way, who actually benefits.
"I think cells coordinate until they have to compete," said Shraddha Chandthakuri, a Miller School cancer biology graduate student in the Isom lab. Under stress, she said, cells may redistribute proteins and metabolites to support the survival of the broader population.
Bruno Colon, another graduate student in the lab, is now studying how these same connections behave in aggressive cancers like glioblastoma.
"Understanding their role could give us new insight into how these tumors communicate," Colon said, "and potentially reveal vulnerabilities we haven't recognized before."
Why This Discovery Needed AI
Isom was direct about the limits and the promise of the tool that made this finding possible. "AI cannot be blindly trusted," he said, "but can lead to really big things in the hands of experts and prepared minds."
Searching 214 million proteins by structure isn't something a research team can do by hand. That scale is exactly where AI earns its place. This is a lab that still runs every hypothesis through traditional experiments before trusting it.
What This Means for Miami's Cancer Research
This isn't a Miami company adapting someone else's AI tool. It's Miami-based researchers using AI to generate a genuinely new biological finding. That finding has direct implications for how tumors like glioblastoma survive and spread.
For Sylvester's own cancer patients, that combination matters. AI-scale search paired with bench science, willing to test what it finds, is the work a top cancer center should produce.
A Miami Hub Just Launched an AI News Anchor Named LIA