A search through protein shapes found a regulator of the bridges between cells

The protein had a name before it had a useful explanation. Filed as TMEM184C in human gene databases for roughly two decades, it offered researchers little obvious guidance about what it did. The breakthrough came when a team searched for a resemblance in its shape.[3]

The resulting study, published in Nature on 9 September, connects TM184C to physical bridges between cells. Vesicles—small membrane-bound packages—move through those connections. Researchers at the University of Miami have identified a molecular regulator of that exchange, opening a question about how neighbouring cells share their contents.[2]

Searching a library of shapes

Proteins are chains of amino acids that fold into three-dimensional structures. Comparing the chains can reveal relationships, but shape offers another way to search. AlphaFold’s contribution is scale: the public database maintained by EMBL-EBI and Google DeepMind contains more than 200 million predicted protein structures. A shape that once would have required its own experimental investigation can now serve as a starting hypothesis across an enormous catalogue.[4]

The team compared 214,528,851 AlphaFold2 predictions with rhodopsin, a well-studied receptor, using structural alignment and additional geometric filters. TM184C emerged with seven membrane-spanning helices, or coiled segments, arranged like those of G protein-coupled receptors. Its sequence had not supplied the resemblance that would ordinarily point researchers towards that family.[1]

That search supplied a lead. The biological question still had to be answered by changing the protein in cells and watching what happened. A useful analogy is finding an unfamiliar machine because its parts are arranged like those of a familiar one: the resemblance tells you where to investigate, while its operation reveals its job.

The connections changed when the protein changed

TM184C accumulated in internal vesicles that travelled along microtubules and into thin cellular projections. Those projections could establish connections with neighbouring cells. Disrupting the protein altered cell shape and vesicle organisation and reduced connectivity.[2]

In cultured HEK293A cells, overexpressing TM184C produced a connectivity score about three times the empty-vector control score. The score weights bridges, projections and protrusions by their extent and connections; it is not a simple bridge count. Comparisons used 28–60 image regions per condition across two biological replicates. In separate co-cultures, differently coloured fluorescent tags revealed vesicles passing in both directions between TM184C-expressing cells.[1]

The protein’s tail also mattered: altering its regulatory motifs weakened its ability to support connections. TM184C recruits β-arrestin and is modified by GPCR kinases, both familiar parts of receptor regulation. Yet the team did not establish coupling to the G proteins that give GPCRs their name. “GPCR-like” is therefore the precise description: the structural clue uncovered familiar regulatory machinery doing a newly assigned job.[3]

The work also links TM184C to autophagy, the recycling process inside cells. Reducing the protein increased markers associated with that process. Together, the results place it at an intriguing junction: organising material within a cell and helping that material move between cells.[2]

What does the neighbour gain?

The next question is whether exchange helps both participants. Miami researchers are investigating these connections in normal cells and aggressive cancers, including glioblastoma. A route for sharing resources could support a stressed neighbour or benefit a tumour. Those are biological possibilities to investigate; this cell-based study did not test a treatment or demonstrate benefit in patients.[2]

TM184C gives that question a concrete handle. Researchers can now examine what changes when this regulator is present, altered or absent. The important move was from a protein recorded in a catalogue to a specific process that can be investigated experimentally.

Report an error or suggest a correction