Cellular droplets promote vital biochemistry–but may dangerously solidify

Ken Garber

Science, October 23, 2015

One day last year, a graduate student brought molecular geneticist Paul Taylor an ice bucket holding a test tube that contained a white solution. When she lifted the tube from the bucket, the solution became clear—and it went white again when she put it back in. Taylor was baffled at first, because the tube held a dissolved protein, and proteins are normally clear in solution. So he examined the white version under a microscope. The protein had apparently “demixed” out of water, forming tiny droplets in a process analogous to the separation of vinaigrette into oil and vinegar. What made the observation especially interesting was the identity of the protein: It was the normal product of a gene that, when mutated, Taylor’s team had found to cause amyotrophic lateral sclerosis (ALS), or Lou Gehrig’s disease.

Taylor is not the only biologist captivated by these liquid transformations .” Over the past year, several other groups have independently seen protein droplets form in test tubes and in cells, and four papers on the unusual phenomenon have just been published in Cell and Molecular Cell. Already, investigators are proposing that droplet formation is a key part of the machinery regulating gene expression in cells. When the process goes awry, they suggest, it can lead to the solid protein aggregates that are a hallmark of ALS and other neurodegenerative diseases.

“Together the papers are a tour de force and really a very large advance,” says James Shorter, a cell biologist at the University of Pennsylvania.

Steve McKnight, a biochemist at the University of Texas-Southwestern (UTSW), agrees that is the field is on the cusp of something profound  “We are seeing the first glimpses of… a whole new understanding of how cells are organized.”

The new work addresses a biological puzzle. Gene expression, and thus most of biology, depends on two large families of proteins: transcription factors, which turn on genes, and RNA binding proteins (RBPs), which transport RNA to the correct parts of the cell so that they’re in the right place at the right time to be translated into proteins. “The crown jewels of biology,” McKnight calls these families. Oddly, many of these proteins have tails that resemble the self-propagating infectious particles called prions, which cause mad cow disease and, in humans, Creutzfeldt-Jakob disease. But why?  “It’s been an enigma,” says McKnight.

The new papers provide an answer, at least for RBPs: the prion-like domains drive their assembly into the droplets that fascinated Taylor. These droplets, he and others hypothesize, serve to protect and concentrate the proteins and their attached RNAs, keeping them from dispersing into the cell fluid and ferrying them intact to their proper cellular destinations. Such structures, speculates Tony Hyman, a cell biologist at the Max Planck Institute in Dresden and co-author of one of the new papers, may have played a key role in the emergence of life more than 3 billion years ago by protecting its chemical reactions from a hostile environment, before cell membranes came about. “As soon as organized molecules would have formed in a primitive soup they would have started to phase separate to form reaction centers,” he says.

But the proteins’ ability to assemble into droplets has a potentially dangerous flip side—they can solidify in a manner that causes disease. In papers in the 27 August and 24 September Cell, Hyman’s and Taylor’s groups reported that given enough time and under the right conditions, RBP droplets can become fibrous structures resembling those that aggregate in the brain cells of people with ALS and Alzheimer’s disease. The teams also found that abnormal RBPs, encoded by mutant genes known to cause ALS, are quicker to aggregate. “This work is important because it provides a novel mechanism for formation of pathological aggregates,” says Ben Wolozin, an Alzheimer’s disease researcher at Boston University.

However, the aggregation has been seen only in purified RBPs in solution, not in living cells. “We’ve tried very hard,” says Hyman. “In vivo they don’t convert from liquid to solid.” He suspects that the molecular machinery that eliminates aberrant proteins normally prevents aggregation, and that disease aggregates appear when quality-control mechanisms are weakened by aging or disease.

The study of protein droplets has become a lively subfield. In the October 15 Molecular Cell, a UTSW-University of Colorado collaboration  provides independent confirmation that the droplets assemble from prion-like protein tails. And in the same issue, a Brown University team used nuclear magnetic resonance to show that RBPs and RNAs are not so much enclosed by the droplet but rather the RBPs interact with each other’s prion-like tails to assemble it.

Just how the prion-like domains initially drive droplet formation remains a mystery. Taylor thinks that weak bonding between the domains is enough to make it happen. McKnight’s lab three years ago was the first to suggest a role for these prion-like domains. In his hands, instead of droplets, they formed protein fibers, with similarities to disease aggregates but looser. He doesn’t dispute the droplet work: “They’re finding neat things, cool things, and if there’s a controversy it’ll be sorted out in short order.”

Drug companies are already paying close attention, because understanding how the droplets solidify in disease could lead to new ways to treat neurodegeneration. “I have been swamped with requests to teach pharmaceutical companies how to target this process,” Taylor says. “This is the biggest thing to happen in cell biology in my career.”