.png)
In our last article, we explored how modular design lays the foundation for building synthetic organelles. By separating structure from function, modularity unlocks the flexibility and scalability needed to tackle complex biological tasks (read more in Build Organelles: Modularity Is All You Need).
One of the most promising approaches to structuring organelles is liquid-liquid phase separation (LLPS), the natural process behind membraneless organelles like stress granules and nucleoli (read more in Organelles in Liquids). LLPS enables cells to create dynamic compartments by organizing molecules into distinct phases without a membrane, offering unparalleled versatility for biological organization.
Now, let’s design organelles from scratch! The first question is: how can we create biomimetic phase-separating molecules to serve as the structural modules for synthetic organelles? What makes these molecules special? What are their defining features? And how do they achieve multivalent interactions, the key to phase separation?
To answer these questions, we must dive into the characteristics and mechanisms behind these molecular architects.
Imagine hosting a party. At first, everyone mingles effortlessly. But as the room fills up, cliques start to form—guests cluster together based on shared interests.
That is exactly the same two faces of phase-separating molecules.
They can remain soluble and free-floating in water, like most biomolecules we know. But as concentration increases—more “guests” entering the room—they find comfort in clustering with their own kind, forming tightly packed groups. This is the basis of phase separation: molecules leaving the solution to form their own exclusive “phase.”
In reality, the boundary between soluble and phase-separating molecules isn’t black and white. I believe, nearly any biomolecule—proteins, RNAs, or DNAs—can exhibit phase separation. You just need to find the right condition to trigger it.
The real challenge is to achieve phase separation under physiological conditions. To do that, we need to understand the exact mechanisms and how to tune them.
The magic behind phase separation lies in multivalency—the ability of a molecule to interact with multiple components simultaneously. This multivalency arises through two main sources:
In 1973, Christian B. Anfinsen introduced the "thermodynamic hypothesis", also known as Anfinsen's dogma, claiming that a protein's structure is dictated by its sequence. Anfinsen later won a Nobel Prize for this discovery, and the concept became the foundation for protein science, and for AI tools like AlphaFold that predict protein structures. Most proteins we know conform to this rule, like hemoglobin, which are tightly packed into spherical, globular shapes, and display a water-friendly surface.
But some are more…free-spirited.
Intrinsically Disordered Regions (IDRs) lack a fixed 3D structure, making them flexible and dynamic. Think of IDRs as oily strings that can interact with each other nonspecifically, forming weak, transient bonds, through the following mechanism:
One fascinating subset of IDRs is Low-Complexity Domains (LCDs), repetitive regions like glycine-rich or serine-rich sequences. Proteins like FUS and hnRNPA1 use their LCDs to drive LLPS, forming stress granules and ribonucleoprotein condensates.
Phase-separating molecules thrive on connections, lots of them. If each component can interact with multiple other components, the crowd will form a network of bonds bringing together proteins, RNA, or other molecules.
Then will this network behave like liquids? In physics, we can analyze this structure by its range of order. If interactions are only specific locally but randomly organized on a larger scale, we say the material only have short-range orders, alnd will be a liquid. Otherwise, medium to long range orders leads to glassy or solid-like states.
Creating multivalency is surprisingly simple: just find or design molecules with multiple interaction sites. Examples in nature include:
So, you’re ready to design synthetic organelles and dive into the world of phase-separating molecules. But where do you find the building blocks, the data and tools to decode intrinsically disordered regions (IDRs) and protein-protein interactions? Thankfully, there’s a treasure trove of resources to help. Here are some of the best places to begin your exploration:
Of course, RNA-based phase separation is also a very interesting class of candidates. Yet, as it is still in its early days. You can find database like RPS 2.0 (https://rps.renlab.cn/) for LLPS-associated RNAs, but purely RNA driven separation is not fully curated until now.
Phase-separating molecules are nature’s master architects, transforming cellular chaos into organized functionality through the elegant interplay of multivalency and molecular dynamics. From intrinsically disordered regions that act like free-spirited connectors to multivalent modules that weave intricate networks, these molecules showcase the power of subtle yet coordinated interactions.
As we continue to decode the secrets of biocondensates, the possibilities grow exponentially. Whether you're engineering a novel metabolic pathway, debugging a stubborn biosynthetic process, or creating entirely new cellular architectures, phase-separating molecules are your ultimate collaborators.
Designing biomimetic organelles isn't just about mimicking nature. It's about understanding its principles and reimagining them for the challenges ahead. The journey to harness their potential is just beginning.
