Water is not merely a passive background for life—it is an active, structured participant that shapes the behavior of biomolecules. Around proteins, water exists in at least two distinct states: bulk water, which is relatively disordered and freely diffusive, and a bound, well-ordered hydration layer, where water molecules form structured hydrogen-bond networks tightly coupled to the protein surface. This interfacial water is neither fully liquid nor solid; it is a dynamic, adaptive medium that stabilizes protein structures, mediates interactions, and enables conformational flexibility. From a broader perspective, life itself may be understood as a balance between enthalpy-driven ordering (structured hydration, binding, and molecular recognition) and entropy-driven processes (fluctuation, diffusion, and disorder). Maintaining this delicate hydration shell could be essential for preserving biomolecular integrity—suggesting a provocative idea: that aging, protein misfolding, and even neurodegenerative diseases such as Alzheimer’s and Parkinson’s may partly arise from the progressive loss or disruption of structured water. In this light, “keeping water” in its organized, functional state may open new directions in studying liquid–liquid phase separation (LLPS), cellular aging, and the physicochemical origins of life.

Biological background

Annexin V (AnxA5) is a Ca²⁺-dependent membrane-binding protein widely used as a model system for investigating protein–membrane interactions and cellular repair mechanisms, and is also a well-established biomarker for apoptosis due to its high affinity for phosphatidylserine exposed on apoptotic cell surfaces. Structurally, it consists of a conserved core of four α-helical domains organized into two interacting modules, along with a flexible N-terminal region that facilitates dynamic interactions with the surrounding environment. Upon calcium influx—such as during membrane damage—Annexin V rapidly associates with negatively charged phospholipid membranes and self-assembles into highly ordered two-dimensional lattices, typically forming trimer-based hexagonal arrays. These supramolecular structures act as stabilizing scaffolds that support membrane repair. Notably, both the assembly and function of Annexin V are highly sensitive to the interfacial environment, particularly hydration and ionic conditions, making it an ideal platform for elucidating how water, ions, and protein structure cooperatively govern membrane-associated biological processes.

Technical Advances

This study introduces a breakthrough in directly visualizing water at biomolecular interfaces by combining three-dimensional atomic force microscopy (3D-AFM) with all-atom molecular dynamics (MD) simulations. Unlike traditional ensemble-averaging techniques, this approach resolves the three-dimensional organization of hydration layers at molecular resolution in real space and near-physiological conditions. Using the membrane-binding protein Annexin V (AnxA5) as a model, the researchers mapped, for the first time, a continuous, nonlayered hydration network extending 1.5–2 nm from the protein surface into the bulk solvent. The integration of experiment and simulation enables not only structural visualization but also mechanistic interpretation of hydration dynamics, bridging a long-standing methodological gap in structural biology.

Conceptual Advances

The findings redefine interfacial water as an active mediator of supramolecular organization, rather than a passive solvent. The hydration shell surrounding AnxA5 is shown to form dynamic hydrogen-bonded bridges that stabilize interdomain interfaces and modulate the local energy landscape. Crucially, this structured water network enables configurational flexibility, allowing reversible and stochastic rotations of protein trimers—motions that are likely essential for membrane repair function. This reveals a new principle: protein dynamics and function are co-regulated by hydration architecture, where water acts as a tunable “soft glue” that balances stability and flexibility. The study thus shifts the paradigm from a protein-centric view to a protein–water co-assembly framework, where biological function emerges from their coupled dynamics.

Testable Hypotheses and Future Directions

This work opens several experimentally testable hypotheses:

1. Hydration-controlled function: Perturbing the hydration layer (e.g., via osmotic stress, ion concentration, or temperature) will directly alter protein conformational dynamics and membrane repair efficiency.
2. Water-mediated allostery: Structured hydration networks transmit mechanical and energetic signals across protein domains, enabling long-range allosteric regulation.
3. Hydration in disease: Disruption of ordered interfacial water may precede or drive pathological aggregation in neurodegenerative diseases, linking hydration loss to protein misfolding.
4. LLPS regulation: The balance between bound and bulk water governs phase separation behavior, suggesting hydration as a key control parameter in biomolecular condensates.
5. Origin-of-life implications: Primitive hydration structures may have provided the necessary balance of order and fluctuation to enable early biomolecular assembly and function.

Looking forward, extending this framework to more complex systems—such as intrinsically disordered proteins, large molecular machines, and crowded cellular environments—could establish a new field of hydration-resolved structural biology. Ultimately, by integrating high-resolution imaging, simulation, and theory, this work suggests a unifying vision:
Life is not only encoded in molecules, but in the structured water that surrounds and animates them.

Supplementary V6: Rotational motion of the central non-crystalline p6 trimer in the 2D AnxA5 assembly.

Article

Funder

This work was primarily supported by grants-in-aid for scientific research (21H05251, 23K23222, 19K06581, 23K05713, 23H02452, 24K01308) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (MEXT), World Premier International Research Center Initiative (WPI), MEXT, Japan. JST-CREST Program (JPMJCR24A4) and JST PRESTO Grant Number JPMJPR23JC. Calculations were partly conducted on a supercomputer at the Research Center for Computational Science in Okazaki, Japan (project: 25-IMS-C097 to T.S.).