[Condensates demystified]: Biomolecular Condensates and the Drugging of the Undruggable: A New Paradigm in Therapeutic Discovery

Biomolecular Condensates and Drugging of the Undruggable

Recent advances in the understanding and manipulation of biomolecular condensates—membraneless cellular compartments formed via liquid-liquid phase separation—are fueling a new wave of innovation in drug discovery. Traditionally, drug discovery has centered on targeting proteins with stable, well-defined structures. However, many proteins implicated in diseases, including cancer and neurodegeneration, are either intrinsically disordered (IDPs) or rely on regions that are intrinsically disordered (IDRs) for their function, and have historically been labeled “undruggable.” Advances in technologies that facilitate structural studies of disordered and dynamic proteins (e.g., NMR, cryo-EM, small angle scattering, proteomics, etc.), and digital tools that predict with increasing accuracy the structural ensemble of IDPs/IDRs, and enable large scale in silico screens, synergize with the progress in biomolecular condensates science, to erase the “un” from “undruggable”.

Disorder-Centric Drug Discovery Comes of Age

A recent review by Lazar et. al., published in Nature Reviews in Drug Discovery, articulates the broader framework needed to drug IDPs/IDRs, highlighting how condensate biology intersects with disorder-driven pathology, especially in cancers, neurodegenerative diseases, and viral infections. Condensates formed by IDPs/IDRs often serve as functional hubs—and disrupting their formation or composition using condensate modulating drugs (c-mods) can yield therapeutic benefits.

Crucially, the review catalogs small molecules that target IDPs/IDRs either directly or by interfering with their condensate dynamics, via a range of phenotypic mechanisms of action (e.g., condensate induction, condensate dissolution, etc.). It also details an expanding toolbox—including phenotypic screens, virtual ensemble docking, and proteomics—for identifying IDP/IDR-binding compounds. Importantly, this work consolidates evidence that supports condensate-centric drug discovery by showing its feasibility, tools, and therapeutic rationale. The review outlines a comprehensive strategy for leveraging condensate biology to drug traditionally “undruggable” targets that drive diseases of high unmet need such as neurodegeneration and cancer.

Inducing Condensates to Suppress β-Catenin Oncogenicity

In a landmark study published by Yan et al. in Nature Communications, researchers demonstrated that a small molecule, Rosmanol Quinone (RQ), can induce condensate formation by β-catenin, a canonical “undruggable” transcriptional coactivator in the Wnt signaling pathway. RQ functions as an inducer c-mod, which partially unfolds β-catenin to promote its condensation into cytoplasmic condensates and block its nuclear translocation. Functionally, the entrapment of β-catenin in pharmacologically induced cytoplasmic condensates silences oncogenic transcriptional programs.

A nanoparticle of RQ conjugated to human serum albumin (Abroquinone, AQ), was selectively uptaken by β-catenin-driven liver cancer cells via micropinocytosis. The nanoparticle demonstrates superior biosafety and tumor-specific uptake, while triggering β-catenin condensation and tumor growth suppression in vivo.

This work illustrates a pioneering application of modulating condensate formation as a pharmacologic strategy for inhibition of transcription factors and overactive signaling nodes with no well-defined pockets to target. A structure-based approach to designing and optimizing such c-mods requires a thorough understanding of the molecular features that drive condensation in the cellular milieu.

High-Throughput Profiling of Condensate-Promoting Sequences

CondenSeq is a high-throughput pooled imaging method that screens thousands of protein sequences in live cells to decode how protein sequences drive formation of nuclear condensates with specific phenotypic signatures in cells, recently published by Kappel et al. in Nature Methods. Sequence analysis of >14,000 sequence tested in ~7.8 million cells, correlated with phenotypic condensate signatures confirms known sequence-dependent condensate-driving features, like valence, net charge, aromaticity and amino acid patterning. Crucially, this study shows that such features have context-independent effects, reinforcing the idea that condensate-promoting motifs can be reliably engineered or targeted across diverse protein scaffolds.

Further, by mapping sequence features that promote colocalization with nucleoli or chromatin, and concentration-driven buffering capacity, the study categorizes distinct classes of condensates and their sequence-level drivers. The ability to predict and classify these behaviors greatly enhances the precision with which researchers can design c-mods or engineer synthetic proteins for therapeutic use.

Implications and Outlook

Together, these studies make a compelling case that the science of biomolecular condensates has reached a level of maturity that justifies greater integration into the drug discovery pipeline:

• Modulating Protein Localization and Function: The β-catenin example illustrates how enforcing condensation can sequester oncogenic factors away from the nucleus, yielding durable functional suppression.
• Engineering Selectivity: The AQ strategy shows how condensate biology can be paired with delivery systems (e.g., macropinocytosis) for tumor-specific uptake.

• High Throughput Target Discovery: CondenSeq dramatically scales up our ability to study how protein sequence drives phase behavior in disease-relevant cellular models, offering a powerful approach to identify genetically-linked disease-driving condensate aberrations (condensatopathies).

• Systematic Sequence-Level Targeting: CondenSeq opens the door to rational design and optimization of c-mods across multiple disease-relevant proteins.
• From Phenotypes to Mechanisms: The integration of biochemical, imaging, and computational approaches demystifies the “black box” of condensate behavior, enabling mechanistic drug design even for disordered targets.

These advances converge on a single message: the conceptual and technical barriers that once segregated IDPs/IDRs and condensates from drug discovery are falling. As we enter a new era of molecular medicine, the frontier of druggability is expanding—not by narrowing our focus to classical active sites, but by embracing the dynamic, multivalent world of protein disorder and phase separation.

Authors

Ann Boija, PhD – Head of Research, Dewpoint Therapeutics

Francis Carpenter, PhD – Head of Data Science & Engineering, Dewpoint Therapeutics

Diana M Mitrea, PhD – Head of Scientific & Corporate Communications, Dewpoint Therapeutics