Exploring Water Beyond the Solvent: Insights into Density Fluctuations and EGFP Unfolding via Luminescence Thermometry

Abstract

Water plays an active role in protein stability, but directly probing its density fluctuations at the protein interface remains challenging. Here, we use enhanced green fluorescent protein (EGFP) to investigate how low-density (LD) and high-density (HD) water motifs modulate unfolding in H₂O and D₂O. Fluorescence quenching during heating-cooling cycles indicates that unfolding begins at approximately 55 °C in H₂O and 64 °C in D₂O, consistent with the stabilizing effect of isotopic substitution. Circular dichroism corroborates this shift, with higher melting temperatures in D₂O (83 vs 79 °C in H₂O). EGFP Brownian velocity measurements, through luminescence thermometry, reveal bilinear temperature dependence with crossover temperatures of 55 °C in H₂O and 65 °C in D₂O, indicating that LD motifs persist longer in heavy water. Together, these results establish a fully optical strategy that directly links hydration-water structure to protein stability, providing a new route to study hydration-mediated dynamics in biomolecules.

1

(a) The 3D scaled representation of the EGFP structure based on the Protein Data Bank. The magnification presents the EGFP chromophore, downloaded from the RCSB PDB (RCSB.org) of 4EUL. (b) Emission spectrum of a 0.71 μM EGFP aqueous suspension at 25.0 °C with deconvolution into two Gaussian components (peak 1, green area, and peak 2, blue area). The solid lines represent the fitted envelopes (R ² > 0.996). The inset presents the anionic form of the EGFP chromophore with the corresponding amino acid residues (Gly67, Thr65, Tyr66, and Glu222). (c) Schematic illustration of EGFP thermal unfolding, aggregation, and partial refolding.

2

(a) Emission spectra of a 2.86 μM EGFP aqueous suspension at 30.0 and 77.5 °C deconvoluted into two Gaussian components (peak 1, green area, and peak 2, blue area). The solid lines represent the fitted envelopes (R ² > 0.997). Arrows indicate the redshift of the peak energies as the temperature increases. (b) Integrated emission intensity and (c) peak 1 and (d) peak 2 energies upon heating–cooling cycles. The repeatability of both peaks is 99.98% at 25.0 °C (Table S5, Supporting Information).

3

(a) Temperature-dependent emission spectra of a 1.43 μM EGFP suspension during heating and cooling. (b) Normalized emission integrated intensity (475–625 nm) over heating–cooling cycles; recovery decreases with increasing maximum temperature. The points correspond to the mean value of the emission integrated area across the 100 recorded spectra, while the error bars represent the corresponding standard deviation. (c) Fluorescence intensity recovery fraction after cycles up to 95.0 °C fitted with a Boltzmann function (r ² > 0.962, Table S3, Supporting Information).

4

(a) Temperature-dependent Brownian velocity of EGFP in aqueous suspensions (1.43–3.57 μM). The horizontal and vertical error bars were calculated as discussed in Sections 9 and 10, Supporting Information. (b) Brownian velocity at 30.0 °C across concentrations from 0.36 to 3.57 μM, including data (red circles) from our previous work for comparison. (c) Crossover temperature extracted from bilinear fits, independent of EGFP concentration (Table S8 and Section 10, Supporting Information).

5

(a) Fluorescence intensity recovery (2.86 μM) after heating–cooling cycles in H₂O and D₂O (Figures S13, S20, Supporting Information). The lines are the best fit to the data using Boltzmann functions (r ² > 0.988, Tables S3, S4, Supporting Information). The points mark the melting temperature T _m, while the arrows mark the onset of protein unfolding. Temperature-dependent Brownian velocity in H₂O and D₂O for (b) 2.14 μM and (c) 2.86 μM. Dashed lines are the best linear fits to the data (R ² > 0.944, Table S8, Supporting Information) used to determine T _c.