In the division of labor inside our cells, there are the “offices” – well-defined organelles walled in membranes – and “street corners,” in which proteins come together in wall-less “droplets” to carry out basic cellular functions. These droplets have been the focus of great interest in various scientific fields, as they play a role in everything from “policing” the cells’ activities to organizing their response to stresses, but understanding the basic molecular forces involved in forming these gatherings has been difficult. A research group at the Weizmann Institute of Science has now found a way to unveil some of their secrets by turning yeast cells into tiny test tubes and getting them to produce synthetic proteins that form simplified, well-lit versions of natural protein droplets.
The proteins were designed and created in the lab of Dr. Emmanuel Levy of the Structural Biology Department. Each yeast cell in the experiment produced two kinds of synthetic proteins. Attraction between the two kinds can cause them to assemble, ad-infinitum, into a gel-like network — something like “instant pudding” mixes that thicken milk by creating cross links between sugar or protein molecules.
These synthetic proteins coalesce spontaneously into droplets, just like oil in water. This so-called phase separation depends on the condition of the mixture: Even oil and water, when in certain conditions, will remain stably mixed without separating. A fine line – the phase boundary – stands between the conditions in which separation can occur and those in which it does not. Such boundaries exist in living cells, but their direct visualization has been elusive. Meta Heidenreich, a research student who led the study in Levy’s group, explains that chemists typically experiment with the temperatures and/or pressures of the mixtures to identify phase boundaries. When dealing with cells and proteins, however, fiddling with temperature and pressure were not options.
The synthetic-protein-in-a-living-cell setup enabled the researchers to track different variables, the first being the protein concentrations. Too much sugar, too much protein, or too little of both won’t create an instant pudding. Similarly, two much of one protein relative to the other or too little of both did not yield a droplet in the cell. Thus, the goal of the first part of the study was to identify the protein concentrations at which droplets form. By monitoring thousands of cells, each producing a different amount of the two proteins; some of them forming droplets and some remaining droplet-free, the researchers were able to see where the natural phase boundaries lie.
To this basic understanding, the researchers added a second variable: affinity – the strength of attachment with the opposite protein. In the experiments, the team tuned the proteins’ affinities by making tiny variations to their amino acid sequences. Again monitoring thousands of cells, the researchers were able to show how changes in affinity changed the phase-boundaries. Increasing the affinity between the two protein types moved the boundary curve downward, making droplet formation more likely even with low protein concentrations within the cells. This had been observed by work with proteins in test tubes, but no one had observed such changes in living cells.
Working with Prof. Sam Safran and his group in the Institute’s Chemical and Biological Physics Department and researchers in Oxford University, the group was able to theoretically relate the detailed properties of individual proteins to those of the droplet as a whole. Levy likens the proteins to birds in a giant swarm: What appears to be a single, coherent unit emerges from the individual actions of hundreds of birds. Many think that protein droplets and the flights of swarms may come down to simple rules each “individual” follows – one bird to its neighbor, single proteins following the rules of chemical attraction to another protein. If we look at these proteins as ‘social beings’,” Levy explains, “in our experiments we can chemically tune them to be more or less ‘friendly,’ and then observe how this changes the way the crowd forms.”
When dealing with cells and proteins, however, fiddling with temperature and pressure were not options.
Once they had shown that synthetic proteins could be a powerful tool to uncover the dynamics of the “street-corner” protein droplets, the team used their method ask a question about the way the proteins making up the droplet find one another. One theory held that as proteins are being manufactured off the messenger RNA, they can already be attracted to other proteins in an emerging droplet. Moving to the droplet would grant the added advantage of localizing further protein production “on site.” But the protein-making apparatus is bulky, with its long strand of messenger RNA, a number of protein-producing ribosomes and proteins strands in various stages of fabrication. The findings showed that this entire apparatus, did, indeed, drift to the site of droplet formation, so that the passive attraction between proteins ended up facilitating droplet formation. “Many research groups have been looking for explanations of RNA localization within the cell. Our findings hint that one should also look at affinity between proteins,” says Levy.
As imbalances in cellular droplets have already been noted in such neurodegenerative diseases as amyotrophic lateral sclerosis, frontotemporal dementia and Huntington, Levy and Heidenreich hope the method will help researchers gain new insights into the ways the malfunctions in these diseases occur.
