Biomolecular Condensates (BMCs)

• Biomolecular condensates are dynamic, membraneless organelles formed via liquid-liquid phase separation that compartmentalize and regulate cellular biochemical processes.
• Phase separation, driven by multivalent interactions among proteins and nucleic acids, creates distinct dense and dilute phases with tunable viscoelastic properties.
• Active regulation through enzymatic modifications and physical forces controls condensate size, stability, and reaction kinetics in complex cellular environments.

Biomolecular condensates (BMCs) are dynamic, membraneless organelles formed by the collective demixing of proteins, nucleic acids, and other macromolecules through liquid–liquid phase separation (LLPS). Unlike traditional organelles bounded by lipid bilayers, BMCs emerge when the concentration of biomolecules exceeds a critical threshold, resulting in coexisting dense and dilute phases with distinct molecular compositions and physical properties. These condensates organize and concentrate specific biochemical pathways, regulate reaction kinetics, buffer cellular noise, and compartmentalize molecular processes without permanent barriers, playing central roles in gene regulation, signal transduction, stress response, and metabolism.

1. Physical Mechanisms of Biomolecular Condensate Formation

The dominant driving force for BMC formation is LLPS, wherein multivalent weak interactions among proteins and nucleic acids produce a demixing transition from a uniform solution to coexisting dense ("condensate") and dilute phases. The minimal thermodynamic framework is based on Flory–Huggins theory for associative polymers: $$\Delta G = RT\left[\phi\ln\phi + (1-\phi)\ln(1-\phi) + \chi\phi(1-\phi)\right]$$ where $\phi$ is the polymer (or sticker-rich macromolecule) volume fraction and $\chi$ is an effective interaction parameter that encapsulates the net sticker–sticker and sticker–solvent interaction energies. Phase separation occurs when $\partial^2\Delta G/\partial\phi^2<0$. The saturation concentration $C_{\rm sat}$ sets the threshold for condensate formation; above $C_{\rm sat}$, the size of the dense phase increases with total concentration, but the coexisting concentrations remain fixed (Peran et al., 2019).

Multivalency, often achieved by repeating modular domains or motifs in proteins, increases the effective interaction strength and lowers $C_{\rm sat}$, facilitating LLPS. In complex mixtures, the balance of homotypic and heterotypic interactions, encoded in structured or random pairwise interaction matrices, determines whether the system undergoes condensation (single dense phase) or demixing into multiple coexisting, compositionally distinct condensates (Jacobs, 2023).

2. Molecular Organization and Structural Models

Two canonical interaction paradigms dictate the mesoscale organization within condensates:

• Multivalent domain-motif interactions: Folded interaction domains (e.g., SH3, PB1, MATH) connected by disordered linkers allow formation of dynamic, reversible 3D network scaffolds. For example, networks of Nck SH3 domains and N-WASP proline-rich motifs, or SPOP-DAXX brush assemblies, generate condensates whose internal mesh size, topology, and rheological properties are tunable by domain valency, linker length, and binding affinity. These high-affinity contacts are largely preserved in the dense phase (Peran et al., 2019).
• Low-complexity domain (LCD) organization:
  • Sticker-and-spacer model: Intrinsically disordered regions (IDRs) with distributed "sticker" motifs (aromatics, charged/polar residues) separated by flexible spacers form globally disordered, highly dynamic networks. Stickers mediate reversible, liquid-like contacts, while spacers modulate network architecture and material properties. NMR evidence supports high backbone mobility and a lack of stable secondary/tertiary structure.
  • Cross-β polymer model (amyloid-like): LCDs containing local segments (e.g., LARKS, hnRACs) transiently adopt cross-β structures, forming labile, reversible fibrillar networks underlying droplet cohesion. Solid-state NMR, X-ray, and mutation studies confirm structured β-sheet cores that are reversible and finely regulated by post-translational modifications (e.g., phosphorylation) (Peran et al., 2019).

Sequence patterning of stickers and spacers directly controls not only condensate self-assembly but also nanoscale heterogeneity. Precise arrangement of sticker blocks can produce nano- to meso-scale clustering or globally uniform phases, tunable by design or mutation (Davis et al., 20 Feb 2025).

3. Functional Implications and Regulation by Active Processes

Biomolecular condensates act as programmable reaction platforms that can accelerate or selectively inhibit reaction kinetics through several interrelated mechanisms:

• Enhanced local concentration: By partitioning specific molecular clients, condensates increase the encounter frequency of reactants, e.g., facilitating viral capsid assembly, multistage transcriptional complex formation, or RNA maturation (Frechette et al., 15 May 2025). In simple models, the mean first-encounter time for reactants confined in droplets is shortened relative to the homogeneous phase, provided residence time permits encounter before escape (Fries et al., 9 May 2025).
• Kinetic control and active regulation: Cells exploit energy-consuming processes (post-translational modifications, enzyme localization) to modulate condensate formation, dissolution, and dynamics. Driven chemical reactions, maintained far from equilibrium by energy input (e.g., ATP hydrolysis), can arrest Ostwald ripening, stably maintaining multiple droplets of controlled size and number. Phase-field and stochastic models predict steady-state droplet radii set by the balance between reaction fluxes and molecular diffusion, modulated by enzyme partitioning and catalytic turnover (Kirschbaum et al., 2021, Fries et al., 9 May 2025).
• Surface and interfacial phenomena: Minor species acting as "shared" surfactants can finely tune interfacial tensions, triggering wetting/docking transitions between otherwise immiscible condensates. A low-concentration bridging protein with modest heterotypic affinity shifts condensate morphology from separated to nested or core–shell configurations with minimal changes in system parameters (Li et al., 2023).
• Non-equilibrium phenomena: Active processes (e.g., localized fuel-driven cycles, enzyme-driven conversion of scaffold material) not only control size and stability but also enable spatial positioning, arrested coarsening, division, or even self-propulsion of condensates in response to chemical gradients, as captured by sharp-interface control theory (Goychuk et al., 2024).

4. Material Properties and Dynamic Behavior

The viscoelastic properties of biomolecular condensates—diffusivity, viscosity, network relaxation—are intimately connected to their nanoscale dynamics:

• Friction and relaxation timescales: Rouse-type polymer models, extended to gel-like networks, analytically relate the single-chain reconfiguration time ($T_R$), center-of-mass diffusion constant ($D$), and condensate viscosity ($\eta$) via mean-field expressions:

$$D = \frac{k_B T}{N\zeta}; \qquad T_R = \frac{\zeta\langle R^2\rangle}{k_B T}; \qquad \eta = \eta_s + \frac{k_B T\, C_p\, \langle R^2\rangle}{36 D}$$

Here, the friction coefficient $\zeta$ is set by average sticker lifetime, determined in atomistic simulations to scale with inter-residue contact duration ($\sim 20$–200 ns), and controls the full chain and bulk timescales (Galvanetto et al., 2024).

• Sequence-tuning of properties: Computational and experimental results show that composition, valence, and patterning of sticker motifs (Arg/Lys, aromatics) tune the viscosity over orders of magnitude without fundamentally altering phase boundaries. Sequence design principles allow decoupling of droplet thermodynamics (e.g., $B_2$, phase stability) from dynamics ($D$, $\eta$), facilitating the engineering of responsive condensates (An et al., 2023).
• Aging and glass transition: Over time, liquid-like condensates may slowly solidify via sticker clustering and network percolation, controlled by sticker valence and spacer entropy. Minimal models predict glassy sub-diffusive aging, with relaxation times scaling exponentially with sticker cluster size—a mechanism underlying pathological hardening in neurodegenerative diseases (Roy et al., 2024).

5. Multicomponent and Multiphase Organization

Biomolecular condensates routinely contain tens to thousands of different macromolecules. Mean-field and simulation studies demonstrate:

• Combinatorial capacity: The maximum number of distinct coexisting condensates scales superlinearly with component number provided condensates are allowed to "share" components. Pairwise interactions alone suffice to encode hundreds of distinct condensates, suggesting a combinatorial "condensate code" for cellular spatial organization (Jacobs, 2021).
• Engineering specificity: The minimum number of independent molecular features ("stickers") needed to program $\mathcal{K}$ distinct phases from $\mathcal{N}$ components is generally much less than $\mathcal{N}$, set by the rank of the interaction matrix as determined by a singular-value bound (Chen et al., 2023). Sequence mixing, block patterning, and feature allocation allow design of artificial or synthetic BMCs with prescribed composition and morphology.
• Internal and hierarchical structures: Sequence-encoded patterns and interaction networks can produce hierarchical, multiphase condensates (e.g., core–shell, layered assemblies), and the phase adjacency and morphology (docked/dispersed) can be systematically designed (Li et al., 2023, Jacobs, 2023). Hidden species or "computational intermediates" amplify the capacity of condensates for distributed information processing and adaptive classification, paralleling computational neural networks (Zentner et al., 9 Sep 2025).

6. Regulation of Condensate Size, Stability, and Kinetics

Physical mechanisms enabling robust size, number, and persistence of BMCs include:

• Electrostatic size control: In charged polymer mixtures, incomplete ion screening induces droplets with net charge, leading to long-range electrostatic repulsion that sets a finite equilibrium size:

$$R \sim \left(\frac{\gamma \epsilon}{(\Delta q)^2}\right)^{1/3}$$

where $\gamma$ is interfacial tension, $\epsilon$ the dielectric constant, and $\Delta q$ the net charge per droplet. This mechanism suppresses Ostwald ripening, stabilizing multiple droplets of equal size (Luo et al., 2024).

• Solvent/cosolvent effects and kinetic barriers: Protein-mediated cosolvent regulation can induce hard-wall repulsion between droplets or chains, stabilizing metastable finite-sized droplets with lifetimes set by the free-energy activation barrier to coalescence. The effective potential of mean force exhibits an anomalous hard-wall at a characteristic separation, set by the strength and concentration of the cosolvent (Liu et al., 26 Feb 2025).
• Active mechanical control: Viscoelastic condensates grown around active cores (e.g., centrosomes) can be tuned to rapidly assemble yet resist mechanical stress, with growth limited or shaped by the interplay of assembly rate, viscoelastic relaxation time, and strain-dependent incorporation (Paulin et al., 17 Jun 2025).

7. Outstanding Questions and Directions

Critical unresolved issues for the field include:

• Atomistic determinants: Which specific side-chain interactions (π–π, cation–π, H-bond, hydrophobic) and sequence patterns dominate LLPS and to what extent do they transiently adopt defined structure within droplets?
• Local order and mesh size: Do supramolecular or nanocluster assemblies within condensates form ordered or dynamically fluctuating mesostructures, and how do these correlate with functional output?
• Dynamics and molecular correlation: How are single-molecule and collective motions coupled, and how do emergent viscoelasticity and correlated dynamics scale with network topology and phase composition?
• Interface and boundary phenomena: Are interfacial regions chemically distinct from condensate interiors, and do they confer unique dynamic or mechanical properties?
• Integration of nonequilibrium processes: How do cells exploit active chemical fluxes, mechanical stresses, and spatial patterning to regulate condensate birth, growth, dissolution, and functional performance in fluctuating environments?

Addressing these will require integrated application of solution- and solid-state NMR, X-ray and neutron scattering, single-molecule fluorescence, advanced coarse-grained and atomistic simulations, and direct in vitro and in vivo reconstitution experiments, as well as systematic inverse design of synthetic condensates (Peran et al., 2019, Jacobs, 2023, Davis et al., 20 Feb 2025).