Life in a bubble – droplet science

Phillip Sharp almost gave up. For practically his entire career, the Nobel-prizewinning biochemistry researcher at the Massachusetts Institute of Technology struggled to clarify one of the fundamental unsolved riddles of biology. How do the millions of tiny molecules within a cell come together at exactly the right time and the right place at the right concentration to perform all the necessary chemical reactions that are the beating heart of life itself? The transcription process, for instance, which is at the center of genetic information and its transmission? That’s Sharp’s specialty. “We’ve been thinking about gene expression for decades, and we were all still saying: ‘Transcription factors come to the DNA somehow, and then they perform the transcription and terminate it, somehow.’” But how exactly? “We didn’t know how to think about it,“ says the 75-year-old, who still works in his lab at MIT’s Koch Institute for Integrative Cancer Research. “I basically turned my back on the subject, because I couldn’t figure out how to do anything about it.”

But then, in 2012, a British cell biologist came to the rescue. While study­ing structures called P granules in stem cells, he stumbled on a phenomenon that fundamentally changed how molecular biologists think about how chemical reactions are processed in living beings. It would prove a breakthrough. Sharp thinks “it’s at the heart of basically every biological process.” Almost every task a cell has to perform involves dozens to thousands of different proteins or other chemical molecules, all interacting like workers and machines in a factory. But where production steps in a factory follow strict plans in terms of time and place, biologists had no clue how all the different molecules drifting through a cell’s cytoplasm are able to come together to form the complexes that perform basic living processes – and then dissolve again afterwards.
What signals call them to duty? Biology textbooks don’t hold the answer, although it’s an essential one. After all, if team play isn’t working at the molecular level, it results in diseases like Alzheimer’s, Parkinson’s and cancer. So knowing the answer to that question would potentially open up ways to tackle these devastating illnesses. “Tony Hyman made the essential discovery,“ Sharp says. A stroke of luck, as Hyman nearly didn’t become a scientist in the first place.

Unlike others, Hyman says he wasn’t drawn to the natural sciences from a young age. He didn’t perform extremely well in high school either, and was unsure afterwards which direction to pursue professionally. So he took a temporary job at University College in London paying £30 a week tending to stem cell culture media and preforming other background lab-tech work.

Densely packed with colloids

One of the scientists there convinced him that life under the microscope was fascinating, made him start an experiment, and encouraged his curiosity. One Friday evening, when the technicians and researchers had disappeared for the usual pub-crawl, he decided to spend the evening over his microscope instead. Since then, says Hyman, a fascination with what makes cells ‘alive’ never left him. He went on to study biology and pursue a career in the field that eventually landed him a job in Dresden as Director of the Max Planck Institute of Molecular Cell Biology and Genetics. That’s where he made the discovery that thrilled Sharp and many other scientists.

Just as water molecules distributed throughout the air condense as dewdrops on a leaf or cold windowpane in the early morning, protein molecules are able to come together to form tiny droplets. Like drops of vinegar in oil, they build separate fluid phases. This is essential for cell function, because in these droplets the proteins reach concentrations a hundred to a thousand times higher than in the cytoplasm circulating outside the droplets. Those concentrations are necessary for enabling chemical processes – like transcription – that are essential to life.

In an oft-cited paper published in Science in 2012,Hyman wrote that the question of exactly how biological macro­molecules form organised assemblies was first posed when the field of bio­chemistry was born in the early 20th century. At that time, he says, “biologists considered the cytoplasm to be densely packed with liquid colloid particles that constituted a separate phase, distinct from the surrounding aqueous environment.“ Some of these droplets – called ‘stress granules’ – appear when the cell is exposed to harsh conditions. Others, like the P granules found in germ cells, are associated with the development of oocytes and sperm. Yet other very tiny granules that are only revealed by modern imaging technologies are involved in transcribing, splicing and other processes in the cell nucleus. Although cell biologists already began describing some of these droplets a century ago, they didn’t know enough about the bio­logy of macro­molecules to make any sense of what they saw. Only today has it grown possible to watch proteins come together in living cells, forming and dissolving the condensate droplets. After endless hours of observations and experiments aimed at altering the structure of these proteins, Hyman and his colleagues in Dresden slowly began to understand the mechanism of action at work – how the proteins go through a phase separation process to form the condensate drops.

The key is a protein region with a name that highlights the misunderstandings the domain provokes. Molecular biologists call regions within proteins that are so fuzzy we have no clue what they do ‘Intrinsically Disorganised Regions’ (IDRs). In fact, most of the time, structural bio­logists trying to crystallise transcription factors in important proteins try to simply cut such regions off. “They were seeing them, but they just didn’t think about what they could mean,” Sharp says. It turns out that, like misnamed ‘junk’ DNA, or small RNA sequences – which turned out not to be chopped up rubbish from essential longer RNAs – the IDR tails of proteins are an important feature. “The function of low-complexity sequence domains, which are abundant in the protein universe, have long puzzled biologists, but these experiments support the idea that they may have evolved to mediate such liquid-liquid demixing,“ Hyman wrote in his 2012 study. His work helped show that IDR tails in proteins are able to form very low affinity bindings with one another. Under certain physical conditions, like high or low salt concentrations, hundreds or even thousands of IDRs can come together. Although the binding properties of individual IDRs are weak, the arranged mass of them enables a phase switch – a separation of the proteins into a fluid droplet suspended within the cytoplasm. Hyman calls them ‘condensates’.

“We see this condensate biology all over the cell,“ says Sharp. “In signalling in the cell membrane (…) and we see condensate biology in synapses … in DNA replication and splicing.“ According to him, it’s a fundamental physical fact of biological systems that is quite widely utilised in human cells. It’s also one that “we haven’t been thinking about.”

As experts gradually accept that condensates are a necessary prerequisite in the complex protein machinery essential to processes like transcription, Sharp is already thinking about using this knowledge for drug discovery. The serial founder of successful biotech companies that include Biogen and
Alnylam believes that through understanding “what initiates condensates and dissolves condensates, we can think about how to use drugs to inhibit or stimulate that process.”

Is cancer a disease of condensate dysfunction?

If condensates are fundamental for cells, then it’s clear that condensate dysfunction could be involved in the onset of disease. One hypothesis several groups of scientists are currently pursuing is that some condensates might actually undergo a further phase shift – changing from a fluid to a solid state like that seen in the beta-amyloid-plaques that accumulate in the nerve cells of Alzheimer’s patients. “If those factors are concentrated and then undergo a morphological change, they can form amyloids – they can form pathogenic material,“ Sharp believes. “So we’re likely looking at a set of processes that also explain a lot of amyloid toxicity in different genetic diseases.“ That means cancer – by definition a genetic disease – might also be a target for condensate-led drug discovery. Recent research has shown that cancer cells, which depend to a great extent on the transcription of oncogenes, need to form large transcriptional condensates around the genes they rely on for their abnormal growth and cancerous division cycles. The search is therefore on for drug candidates able to change condensate biology, hence potentially making a difference in diseases like Alzheimer’s or cancer.

Sharp has joined the scientific advisory board of Dewpoint along with other well-known scientists like Bob Langer, Rudolf Jaenisch, Richard Young and Tony Hyman himself. It’s the first company seeking to harness the potential of condensate biology. Based in Dresden, Boston and now Berlin, the firm is currently screening small molecule libraries. And they seem confident they’ll find drugs able to influence condensates in ways relevant to the pathogenesis of neurodegenerative diseases and cancer – even though other library screens have been performed many times before by other companies and scientists.

RNA is an important component of condensates

“I would argue that few of those screens – maybe none of them – have been done in a situation where you’re monitoring the behavior of these droplets,” argues Whitehead Institute Director Richard Young. He’s convinced libraries will be “extraordinarily rich” in compounds that could alter droplet behavior. “Condensates are regulated in really fascinating and complex ways by many of the classical regulators of biological function like kinases, phosphatases and ubiquitases,” he says. “All of these enzymes that have for the last half century been described as playing important regulatory roles in diverse biology.” When they act, he thinks, they’re modifying proteins in a way that can change the partitioning of those proteins both in and out of the droplets. “In other words, they are actually controlling the component levels in the droplets where these functions are occurring.” If Dewpoint screens a library and finds a compound that changes the partitioning behaviour – for instance, the preference of a kinase substrate for a particular droplet – then “that’s a very interesting concept, because we already have drugs against kinases and against all these different enzymes.”

But even if you find a drug that influences a certain condensate in a profound and therapeutically relevant way, how do you manage to direct it to a specific condensate – and away from others? Can you be specific enough to not have a lot of side effects? “If you would have asked me this question say, 20 months ago – before I was aware of Tony’s work – I would have said: That’s going to be very hard,” Young admits. But he’s changed his opinion after seeing “empirical data that you can get very selective effects.” One approach is for instance to alter the partitioning property of a protein that’s drawn to the droplet by altering phosphorylation. Another way to increase specificity might be to use RNA molecules to alter condensate biology. “If you look hard at the components of condensates, they always have RNA components,” Young says. “These play pretty important roles in condensate behavior.” Any approach that modulates levels of these condensate-related RNA molecules could represent “a very reasonable path to a therapeutic,” Young believes. “There are many ways to modulate RNA levels, and to do it quite specifically.”

A field that’s still in its infancy

The goal now is to turn up or down 50% of given proteins that play roles in suspected Alzheimer’s or Parkinson’s pathways. “That’s an ideal scenario for a delivery of a small RNA,” says Young. “Particularly when you can deliver that small RNA once every six months or a year“. Sharp agrees that “RNA is heavily involved in these condensate processes.” He thinks biologists will use RNA-technologies to therapeutically modulate condensates, but also to better understand the function of these mysterious droplets within cells. “Condensate biology isn’t mature,” Sharp says. “Not even close.” 

Sascha Karberg

First published in European Biotechnology Magazine Spring 2020