Hydraulic Activation of the AsLOV2 Photoreceptor

Abstract

How proteins transduce light into mechanical energy remains a central question in biology. This study tests the hypothesis that blue light activation of the LOV2 (light, oxygen, voltage sensitive) domain of Avena sativa phototropin 1 (AsLOV2), gives rise to concerted water movement that induces protein conformational extensions. Using electron and nuclear magnetic resonance spectroscopy, along with molecular dynamics simulations at high pressure, we find AsLOV2 activation can be initiated by blue light or high pressure, followed by selective and concerted expulsion of low-entropy, tetrahedrally coordinated "wrap" water from the protein hydration shell. These findings suggest that interfacial water serves as constituents to reshape the protein's free energy landscape during activation. Our study highlights hydration water as an active hydraulic fluid that can drive long-range conformational changes underlying protein mechanics upon light activation and offers a new concept for engineering externally controllable protein actuators.

Figure 1.

Photoactivation of AsLOV2 induces long range conformational change and modulates interfacial hydration dynamics. (A) Crystal structure of AsLOV2 in the dark state (PDB: 2V1A(62)), highlighting sites labeled with MTSL for ODNP, TiGGER, and DEER measurements (left). Upon photoactivation, the A’α (orange) and Jα helices (red) unfold and move further apart, as illustrated on the right. (B) UV-Vis absorbance spectroscopy confirms light-induced activation of AsLOV2. The wild-type protein exhibits a decrease in absorbance at ~450 nm under blue light illumination (top). In the N414Q mutant, time-resolved measurements at 450 nm reveal a 2.5-fold slower return to the dark state, indicating a stabilized lit-state conformation (bottom). (C) The distance distributions obtained with a DeerLab two-Gaussian model (see SI for further details) from the DEER measurements between T406C and E537C show an increase in the population of extended conformations (3–5 nm) in the lit state (top). The N414Q mutant further enriches the extended state population under illumination, consistent with enhanced structural stabilization of the lit conformation (bottom). (D) ODNP measurements reveal that light activation leads to a strong decrease in fast-moving water (k_σ) across all sites, and a smaller increase in slow-moving water (k_low) across most sites, suggesting reduced hydration dynamics at the protein interface. This effect is accentuated in the N414Q mutant, particularly at site L514C, where further reduction in k_σ and increase in k_low indicate enhanced local dehydration and stronger stabilization of the photoactivated state.

Figure 2.

Light-Driven Redistribution of Distinct Water Populations in AsLOV2 Revealed by High-Field ¹⁷O MAS NMR. (A) Single-pulse 1D 17O MAS NMR spectrum of 40% 17O-enriched buffered AsLOV2 solution, acquired with bulk water suppression at 800 MHz and a spinning rate of 10 kHz ± 3 Hz at room temperature, reveals three spectrally distinct water populations. Bulk water exhibits the longest T₁ (7.95 ± 0.92 ms ), bound water shows the shortest T₁ (1.69 ± 0.09 ms) and wrap water with intermediate T₁ (2.21 ± 0.05 ms). (B) Experiments conducted with fixed τ_ZC =2.7 ms and τ_D = 0, while varying the flip angle θ, show that two water populations—assigned as “bound” and “wrap” water—are preferentially observed at low flip angles (<100º). (C) Due to its slower dark-state recovery, the N414Q mutant allows time-resolved tracking of hydration changes. Upon return to the dark state, the system exhibits gradual: recovery of wrap water, reduction in bulk water signal, and recovery of bound water, all occurring on a timescale of τ_1/2 ~20 minutes, consistent with photocycle kinetics measured by time-resolved UV-Vis spectroscopy at 277 K. (D) In the wild-type protein, light activation leads to rapid and reversible: expulsion of wrap water, an increase in bulk water signal, and minimal change in bound water levels, as observed across repeated photocycles.

Figure 3.

High-pressure molecular dynamics simulations reveal site-specific redistribution of water associated with pressure-induced unfolding in AsLOV2. (A) High-pressure molecular dynamics simulations reveal pressure-driven partial unfolding of AsLOV2, initiated by unraveling of the A’α helix (orange) and extension of the Jα helix (red), mirroring the structural changes observed upon photoactivation. (B) Snapshots from the 3 kbar unfolding trajectory reveal that wrap water remains stably associated with compact regions of the protein, such as at t₁ (5% of total simulation time). By t₂ (12% of total simulation time), a notable decrease in tetrahedral wrap water emerges around the Aβ, Hβ, Iβ, and Jα regions, accompanied by increased bound water in Aβ and Hβ and increased icosahedral water in Iβ and Jα. At t₃ (27% of total simulation time) tetrahedral wrap water further decreases in Iβ and Jα, with a corresponding rise in icosahedral water. Jα begins to unfold at t₄ (53% of total simulation time) and fully undocks by t₅ (100% of total simulation time), where the wrap water redistributes to resemble the 1 bar state, though with slightly reduced overall occupancy, as shown in (D). The vertical positioning of the three plots is arbitrary. The wrap water values plotted here are scaled by 2.5X for clarity. Pink dots mark residues with >50% reduction in wrap water relative to the mean across all residues, while dark-blue and light-blue dots indicate bound and icosahedral water, respectively, which increase at residues losing wrap water at t₂ ns and t₃ ns. (C) Time-resolved residue-wise structural deviations relative to the 1 bar state reveal that pronounced changes begin to occur around t₅ ns, particularly in the Iβ and Jα regions, suggesting that hydration loss precedes major conformational transitions. (D) Violin plot showing the overall pressure effect on local water populations: tetrahedral wrap water decreases, icosahedral water increases, while bound water remains nearly constant due to residue-specific behavior when pressurized from 1 bar to 3 kbar.

Figure 4.

DEER measurements show that pressure and light co-operatively enhance the population of extended state conformations in AsLOV2, with effects amplified by molecular crowding (see SI for details of the analysis of the DEER data). (A) Schematic of MTSL spin-labeled AsLOV2 at positions T406C and E537C, illustrating conformational extension and concurrent release of wrap water upon stimulation by light and/or pressure. (B) Left: DEER-derived distance distribution $P\left(r\right)$ shows that application of 3 kbar pressure alone did not result in a detectable structural extension from the dark-state equilibrium conformation in the WT. Right: Combined light and pressure further increase the extended population to ~21%, highlighting an additive effect. (C) In the presence of PEG_20kDa as a molecular crowder, pressure alone increases the WT extended population to ~4% (left), and light plus pressure drives it further to ~26% (right), suggesting enhanced conformational responsiveness due to internal osmotic pressure. (D) In the N414Q mutant (no crowder), pressure alone results in ~3% of the population in the extended conformation (left), while light plus pressure increases this to ~26% (right), indicating greater stabilization of the lit state. (E) In the presence of PEG_20kDa, the N414Q mutant shows even stronger pressure sensitivity: ~7% in the extended state under pressure alone (left), and ~79% with combined light and pressure (right). (F–G) Summary of population shifts under different conditions, indicating that both external (hydrostatic) and internal (osmotic) pressure cooperatively enhance conformational extension in AsLOV2, with the N414Q mutant exhibiting a greater propensity for adopting the excited-state conformation.

Figure 5:

Residue-resolved analysis of pressure-induced conformational changes in AsLOV2 using high-pressure 3D HNCO NMR. (A) Representative two-dimensional ¹⁵N/¹H projections from high-pressure HNCO spectra collected across a pressure range up to 2.5 kbar. Peak trajectories illustrate pressure-dependent chemical shift changes, including both linear and nonlinear behaviors across different residues. (B) Quadratic fitting of chemical shift changes for three backbone nuclei—¹H, ¹⁵N, and ¹³C—was used to extract nonlinear coefficients $\left(c_i\right)$ for each residue. Bar plots show the distribution of absolute nonlinear coefficients $\left|C_i\right|$ across the protein for each nucleus, with horizontal lines indicating the statistical cutoffs (mean, and mean + 1.645 SD). (C) A composite nonlinearity score was calculated by normalizing and summing the nonlinear coefficients across all three nuclei (¹H_i + ¹⁵N_i + ¹³C_i-1), yielding an integrated metric for pressure sensitivity per peptide group. Structural mapping of the top 10% of residues with the largest deviations from the mean onto the AsLOV2 crystal structure (PDB: 2V1A (62)) reveals that the most pressure-sensitive regions are localized to the A′α and Jα helices and their adjacent segments. These helices are also known to be functionally critical in the light activation mechanism of AsLOV2, suggesting a shared structural pathway for both light- and pressure-induced activation via hydration-mediated “hydraulic” coupling.