Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/multidisciplinary-digital-publishing-institute/photo-switching-of-protein-dynamical-collectivity-La78PL3NVW
- Publisher
- Multidisciplinary Digital Publishing Institute
- Copyright
- © 1996-2021 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer The statements, opinions and data contained in the journals are solely those of the individual authors and contributors and not of the publisher and the editor(s). MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Terms and Conditions Privacy Policy
- ISSN
- 2304-6732
- DOI
- 10.3390/photonics8080302
- Publisher site
- See Article on Publisher Site
Abstract
hv photonics Article 1 1 , 2 , 3 1 Mengyang Xu , Deepu George * , Ralph Jimenez and Andrea Markelz Department of Physics, University at Buffalo, Buffalo, NY 14260, USA; mxu3@buffalo.edu (M.X.); amarkelz@buffalo.edu (A.M.) JILA, 440 UCB, University of Colorado, Boulder, CO 80309, USA; rjimenez@jila.colorado.edu Department of Chemistry, 215 UCB, University of Colorado, Boulder, CO 80309, USA * Correspondence: dkgeorge@buffalo.edu Abstract: We examine changes in the picosecond structural dynamics with irreversible photobleach- ing of red fluorescent proteins (RFP) mCherry, mOrange2 and TagRFP-T. Measurements of the protein dynamical transition using terahertz time-domain spectroscopy show in all cases an increase in the turn-on temperature in the bleached state. The result is surprising given that there is little change in the protein surface, and thus, the solvent dynamics held responsible for the transition should not change. A spectral analysis of the measurements guided by quasiharmonic calculations of the protein absorbance reveals that indeed the solvent dynamical turn-on temperature is independent of the thermal stability/photostate however the protein dynamical turn-on temperature shifts to higher temperatures. This is the first demonstration of switching the protein dynamical turn-on temperature with protein functional state. The observed shift in protein dynamical turn-on temperature relative to the solvent indicates an increase in the required mobile waters necessary for the protein picosecond motions, that is, these motions are more collective. Melting-point measurements reveal that the photobleached state is more thermally stable, and structural analysis of related RFP’s shows that there is an increase in internal water channels as well as a more uniform atomic root mean squared displacement. These observations are consistent with previous suggestions that water channels form with extended light excitation providing O access to the chromophore and subsequent fluorescence loss. We report that these same channels increase internal coupling enhancing thermal stability and Citation: Xu, M.; George, D.; collectivity of the picosecond protein motions. The terahertz spectroscopic characterization of the Jimenez, R.; Markelz, A. protein and solvent dynamical onsets can be applied generally to measure changes in collectivity of Photo-Switching of Protein Dynamical Collectivity. Photonics protein motions. 2021, 8, 302. https://doi.org/ 10.3390/photonics8080302 Keywords: protein collectivity; terahertz; thermal stability; fluorescent proteins; photobleaching Received: 16 June 2021 Accepted: 25 July 2021 Published: 29 July 2021 1. Introduction Fluorescent proteins provide vital insight into biochemical processes. Since the dis- Publisher’s Note: MDPI stays neutral covery of green fluorescent protein (GFP) in Aequorea victoria, researchers have generated with regard to jurisdictional claims in a large number of mutants that enable access to a rainbow of excitation and emission published maps and institutional affil- wavelengths. The power of these proteins is the formation of their chromophore auto- iations. catalytically from three sequential residues [1], thus enabling the incorporation of the fluorescent tag into a targeted protein’s expression. All fluorescent proteins (FPs) share several structural features: an 11-stranded -barrel with an internal distorted -helix to which the chromophore is attached [2,3]. The chromophore removed from the protein has a Copyright: © 2021 by the authors. poor emission yield, four orders of magnitude lower compared to when the chromophore Licensee MDPI, Basel, Switzerland. is inside the -barrel [4]. The -barrel structure protects the chromophore from external This article is an open access article quenchers and inhibits the dark state conversion through a cis to trans isomerization or distributed under the terms and light-induced protonation/deprotonation [5,6]. To achieve higher transmission through conditions of the Creative Commons tissues, longer wavelength excitation red FPs (RFPs) have been developed. Unfortunately, Attribution (CC BY) license (https:// the application of RFPs to imaging is limited by their higher tendency to be irreversibly creativecommons.org/licenses/by/ photobleached. Though this photobleaching is useful for certain specialized microscopic 4.0/). Photonics 2021, 8, 302. https://doi.org/10.3390/photonics8080302 https://www.mdpi.com/journal/photonics Photonics 2021, 8, x FOR PEER REVIEW 2 of 14 Photonics 2021, 8, 302 2 of 14 techniques, such as the fluorescence recovery diffusion measurements, it is more often a techniques, such as the fluorescence recovery diffusion measurements, it is more often a bottleneck for imaging applications, especially for those demanding high irradiance illu- bottleneck for imaging applications, especially for those demanding high irradiance illumi- minations such as single molecule microscopy. The mechanisms that lead to this decrease nations such as single molecule microscopy. The mechanisms that lead to this decrease in in photostability are a current area of study. Here, we examine the changes in the picosec- photostability are a current area of study. Here, we examine the changes in the picosecond ond structural dynamics of several red fluorescent proteins to see if and/or how dynamical structural dynamics of several red fluorescent proteins to see if and/or how dynamical changes may be associated with photobleaching [7,8]. changes may be associated with photobleaching [7,8]. Three engineered monomeric RFPs were studied: mCherry [9], mOrange2 (deriva- Three engineered monomeric RFPs were studied: mCherry [9], mOrange2 (derivatives tives of DsRed from Discosoma sp.), and TagRFP-T (a derivative of EQFP611 from Entac- of DsRed from Discosoma sp.), and TagRFP-T (a derivative of EQFP611 from Entacmaea maea quadricolor) [5]. The structures are shown in Figure 1 [10,11], along with their relative quadricolor) [5]. The structures are shown in Figure 1 [10,11], along with their relative B-factors. The B-factor, or Debye Waller factor, from the X-ray structural measurements B-factors. The B-factor, or Debye Waller factor, from the X-ray structural measurements provides a measure of the root mean squared displacement, reflecting the relative flexi- provides a measure of the root mean squared displacement, reflecting the relative flexibility bility of different regions of the structure. The absolute B-factors can vary for the sample of different regions of the structure. The absolute B-factors can vary for the sample protein, protein, due to the specifics of the experimental conditions, so to compare the B-factors due to the specifics of the experimental conditions, so to compare the B-factors among the dif amfer onent g thpr e oteins, differen we t pnormalize roteins, wethe noB-factors rmalize th to e the B-faaverage ctors to for thethe aveentir ragee fstr or uctur the ee nti for re the struc specific ture fopr r th otein. e speThese cific pthr rote ee inpr . T oteins hese th have ree substantially proteins haveenhanced substantir aesistance lly enhantc o ed both re- sistance to both irreversible and reversible photobleaching under different illumination irreversible and reversible photobleaching under different illumination conditions [12]. T co echniques nditions [to 12characterize ]. Techniquestr s to uctural characflexibility terize struc have turaincr l fleeasingly xibility hfocused ave incron easthe inglimportant y focused picosecond on the impo time rtant scale picowhich secondcorr timesponds e scale w to hilibrational ch correspo motions nds to lof ibrthe atiosolvent nal moti and onsamino of the acid sidechains as well as the long-range intramolecular vibrations associated with confor- solvent and amino acid sidechains as well as the long-range intramolecular vibrations as- mational change [13–16]. Here we use solution phase temperature dependent terahertz sociated with conformational change [13–16]. Here we use solution phase temperature (THz) time-domain spectroscopy, which is a benchtop optical measurement with small dependent terahertz (THz) time-domain spectroscopy, which is a benchtop optical meas- sample size requirements that enables a systematic comparison between different mutants urement with small sample size requirements that enables a systematic comparison be- as a function of photoexcitation. tween different mutants as a function of photoexcitation. Figure Figure 1. 1. Str Str uctur uctur es esof of mCherry mCherr(2H5Q.pdb), y (2H5Q.pdb T ), agRFP-T TagRFP (5JV -T (5 A.pdb), JVA.pd and b), mOrange and mOra (2H5O.pdb). nge (2H5O.p Note db). Note that mOrange is shown as the crystal structure of mOrange2 has not been reported yet. The that mOrange is shown as the crystal structure of mOrange2 has not been reported yet. The width width and color of the ribbon in the figure show the relative atomic B-factor distribution normalized and color of the ribbon in the figure show the relative atomic B-factor distribution normalized to the to the overall averaged B-factor value. overall averaged B-factor value. Comparing the THz measurements to CD measurements of thermal denaturing, the Comparing the THz measurements to CD measurements of thermal denaturing, the turn on temperature in THz absorbance with temperature is found to be inversely pro- turn on temperature in THz absorbance with temperature is found to be inversely pro- portional to the thermal stability. Molecular modeling using quasiharmonic vibrational portional to the thermal stability. Molecular modeling using quasiharmonic vibrational analysis suggests the protein dielectric response can be modeled by a Lorentzian centered analysis suggests the protein dielectric response can be modeled by a Lorentzian centered −1 1 at ~0.5 THz, (20 cm , 2.5 meV). A spectral decomposition of the measured frequency de- at ~0.5 THz, (20 cm , 2.5 meV). A spectral decomposition of the measured frequency pendent absorbance into the protein dielectric response and the bulk water intermolecular dependent absorbance into the protein dielectric response and the bulk water intermolec- −1 ular stretc str hin etching g modemode at 5.3 at THz 5.3 (1THz 77 cm (177 , 22 cm meV) , 22 [17meV) ] show [s 17 th ] a shows t whilethat the p while roteinthe dyn pr am otein ical dynamical transition is transition dependen ist dependent on the specon ificthe pro specific tein and pr p ot ho ein tob and leacphotobleaching, hing, the water h the as a water tem- has peraa tur temperatur e dependeendependence ce that is indthat epenis deindependent nt of the specof ific the RFspecific P or its p RFP hoto or sta its te.photostate. This is the This first ris ep the ort first of thr eeport proteof in the dynpr am otein ical tur dynamical n-on chan turn-on ging ind changing ependentl independently y of the solvent of tur the n- solvent on. turn-on. Photonics 2021, 8, 302 3 of 14 2. Materials and Methods 2.1. Sample Preparation The RFPs were expressed and purified as discussed in Reference 12. The original concentration of the protein solution was 4 mg/mL in 15 mM MOPS buffer with 100 mM KCl at pH 7.0. It was concentrated to 80 mg/mL in the same dialysis buffer to increase the optical density at THz frequencies, using Eppendorf 5424 Microcentrifuge with Millipore Amicon 10 K centrifugal filter. A Dolan-Jenner Fiber-Lite Illuminator light source and an Ocean Optics High Resolu- tion Spectrometer were utilized to characterize the absorption and emission spectra. The spectral peak shifts as a function of concentration (Figure S6 in Supplementary Informa- tion). In addition, it has been verified that each sample completely loses fluorescence after 30 s under the 4 W/cm 532-nm illumination condition. No protein aggregation occurs in the highly concentrated solution, confirmed by the dynamical light scattering (DLS) measurements using the Zetasizer Nano ZS90 system. 2.2. THz Measurements A conventional home-built THz time-domain spectroscopy system was employed to measure the optical absorption of RFPs within the THz range (0.2–2.2 THz) [18,19] as described in the reference [20]. In short, a home-built THz TDS system based on an 80 MHz femto-second laser and photoconductive switch was used to generate THz pulses. An electro-optic coherent detection with ZnTe as the EO crystal was employed. Time domain signal recorded were Fourier transformed to obtain the frequency domain data. A solution cell with 100-micron thick sample was placed in a gas exchange cryostat and cooled with liquid nitrogen. A silicon diode thermal sensor monitored the temperature directly adjacent to the sample aperture. After the photoactive state was measured, each sample was completely photobleached by 4 W/cm 532-nm illumination and at room temperature, then cooled down to 80 K for photobleached state temperature dependent measurements. The photobleaching was confirmed by monitoring the fluorescence while illuminating. 2.3. CD Measurements Thermal stability was measured using a Jasco J-715 Spectropolarimeter for temperature dependent far-UV CD spectra. Samples were prepared at 5 M concentration in 10 mM sodium phosphate buffer at pH 7.5 and sealed in a 10-mm path-length quartz cell. The sample was equilibrated at each temperature for 2 min, and then, 10 scans at 0.1-nm resolution were accumulated and averaged. The scan speed was set to 50 nm/min with 4 s detector response time. The temperature was gradually increased from 20 C to 100 C, in steps of 2 C, using a Jasco PTC 348 WI temperature controller. The spectrum of a buffer blank was used as the reference. The melting temperatures were determined from the peak of the ellipticity temperature gradient, d/dT, measured at 218 nm. In the case of photobleached mCherry, the increase in the melting onset is so large that d/dT maximum is not yet attained by the highest temperature measured (373 K). We estimate the lower bound of the photobleached mCherry melting temperature from the midpoint between the low temperature ellipticity value and the value at the highest temperature measured, 373 K (see Figure S1). 2.4. MD Calculations The temperature dependent absorbance spectra were calculated using quasiharmonic mode analysis and dipole autocorrelation. The MD trajectory of the net dipole was calcu- lated with the CHARMM 39 [21] and CHARMM 36 empirical force field [22]. The special formation by the auto-oxidation and cyclization of three to four amino acid residues re- quires the chromophore to be additionally parameterized [23]. We applied the residue topology and parameter files for the chromophore used in Reference 23 to the calculations presented here. The initial molecular structures were obtained from the Protein Data Bank, mCherry (PDB: 2H5Q) and TagRFP-T (PDB: 3T6H). The mOrange2 structure was Photonics 2021, 8, 302 4 of 14 obtained by modifying mOrange (PDB: 2H5O) with Q64H/F99Y/E160K/G196D mutations implemented using the CHARMM-GUI [24]. The CHARMM/TIP3P parameter set was used to model the fully solvated protein system starting from the X-ray determined structure. Each structure was first minimized at zero temperature using steepest descent method followed by the adopted basis Newton– 7 1 Raphson method until the total energy gradient reached below 10 kcal Å . The energy- minimized systems were then neutralized by adding potassium or chloride ions, randomly distributed throughout the volume. MD simulations were performed with an integration time step of 1 fs. Each system was heated from 100 K to 300 K, with a linear gradient of 1 K/ps. The systems were equilibrated for 10 ns in the isobaric-isothermal (constant pressure-temperature, CPT) ensemble. 4-ns long production trajectories were further performed under the CPT condition at 300 K. The dipole moments about the center of geometry were obtained, and the absorption intensity was determined from the power spectra of the dipole-dipole autocorrelation function [25–27]. 3. Results In Figure 2, we show the molar absorption coefficient (!,T) for the three RFPs at several THz frequencies for both the functional and photobleached states. Both protein and water absorption contribute to the spectra with the water dominating. Previously it has been stated that the THz absorbance follows root-mean-square-deviation (RMSD) measured by neutron scattering in the same energy range dependent on temperature and hydration. We identify two temperature regimes for the molar absorptivity: a low temperature harmonic regime (80–175 K) and a high temperature anharmonic regime (250–270 K). Below 200 K, (!,T) increases linearly with temperature indicating that the protein vibrations in the THz range are harmonic, which is consistent with the q dependent boson peak and linearly increasing RMSD shown in inelastic neutron scattering (INS) [28]. At ~200 K, the absorbance rapidly increases as the dynamics move to the anharmonic regime. The low temperature slopes differ for the three proteins, with TagRFP-T having the largest slope and mCherry the smallest. In addition, the low temperature slope decreases with photobleaching for all three RFPs. This low temperature linear regime followed by an abrupt onset in the picosecond dynamics is identical to the INS measurements of RMSD [29]. Zaccai and coworkers introduced two global flexibility parameters to quantify the RMSD temperature dependence [30]: the resilience k*, defined as proportional to the inverse of the slope of the RMSD temperature dependence at low temperatures expressed as <k*> = 0.00138/(d <3x >/dT), where the angular brackets denote an ensemble average over atoms and time, and the dynamical transition temperature T , where the dynamics sharply change. Here we take a similar approach. Assuming that the THz absorption, which selectively measures the optically active picosecond protein vibrations, is proportional to the vibrational density of states (VDOS) denoted as g(!) which is in turn derived from the Fourier transform of the atomic velocity autocorrelation: g(!) = <v(0)*v(t)> [31], we can use the harmonic approximation to express the absorption coefficient (!,T) and g(!) for a particular frequency ! as p p i!t (!, T) ~ g(!) = < mv(0) mv( t )>e dt p p i!t = <! mx(0)! mx( t)>e dt R (1) 2 i!t = m! <x(0)x(t)>e dt k T 2 2 2 ~ m! <x > ~ m! m! k C(!) =) k ~ ~ ¶a(!,T)/¶T ¶a(!,T)/¶T Photonics 2021, 8, x FOR PEER REVIEW 5 of 14 k T 2 2 2 ~ mω <x > ~ mω Photonics 2021, 8, 302 5 of 14 mω k C(ω) ⟹ k ~ ~ ∂α(ω,T)/∂T ∂α(ω,T)/∂T where k* is an effective force constant, and C(ω) is a frequency dependent coefficient. We refer to k as the THz resilience. For a given frequency, we evaluate the resilience by the where k* is an effective force constant, and C(!) is a frequency dependent coefficient. We inverse of the slope of the molar absorptivity temperature dependence in the linear re- refer to k as the THz resilience. For a given frequency, we evaluate the resilience by the inverse gime. of the slope of the molar absorptivity temperature dependence in the linear regime. Figure 2. THz molar absorptivity for photoactive (left) and photobleached (right) RFP hydrated Figure 2. THz molar absorptivity for photoactive (left) and photobleached (right) RFP hydrated samples: TagRFP-T (A,D), mCherry (B,E), and mOrange2 (C,F), with interpolated lines as a guide samples: TagRFP-T (A,D), mCherry (B,E), and mOrange2 (C,F), with interpolated lines as a guide for the eye, as a function of temperature at different frequencies. Note that vertical axis scale varies for the eye, as a function of temperature at different frequencies. Note that vertical axis scale varies −1 −1 with different THz response: TagRFP-T 0–75, mCherry 0–35, mOrange2 0–55 mM cm . The low 1 1 with different THz response: TagRFP-T 0–75, mCherry 0–35, mOrange2 0–55 mM cm . The low temperature region is indicated by grey shading. temperature region is indicated by grey shading. In Table 1, we show the resilience extracted from the low temperature THz molar In Table 1, we show the resilience extracted from the low temperature THz molar absorptivity with the measured melting temperatures (TM) from CD measurements. The absorptivity with the measured melting temperatures (T ) from CD measurements. The data show an increase in TM for the photobleached state of each RFP. Strikingly, the THz data show an increase in T for the photobleached state of each RFP. Strikingly, the THz resilience follows the thermal stability for all three proteins in both the unbleached and resilience follows the thermal stability for all three proteins in both the unbleached and bleached states. This is the first demonstration that THz resilience correlates with thermal bleached states. This is the first demonstration that THz resilience correlates with thermal stability and is consistent with previous neutron scattering measurements for hemoglo- stability and is consistent with previous neutron scattering measurements for hemoglobins bins from different species [32], suggesting that stronger resilience leading to higher ther- from different species [32], suggesting that stronger resilience leading to higher thermal mal stability may be a general phenomenon. stability may be a general phenomenon. Table 1. The melting temperature for all samples was determined by the midpoint in the drop-off of far-UV CD signal at 218 nm. The effective force constant k is calculated for low temperature regime (100–200 K) at 1.0 THz, while other frequencies show similar tendency. mCherry TagRFP-T mOrange2 T (K) 359 2 345 1 353 4 M, Unbleached T (K) 373 2 350 1 367 4 M, Bleached k 43 14 18 1 21 3 Unbleached k 45 4 22 3 31 9 Bleached Photonics 2021, 8, x FOR PEER REVIEW 6 of 14 Table 1. The melting temperature for all samples was determined by the midpoint in the drop-off of far-UV CD signal at 218 nm. The effective force constant k is calculated for low temperature regime (100–200 K) at 1.0 THz, while other frequencies show similar tendency. mCherry TagRFP-T mOrange2 TM, Unbleached (K) 359 ± 2 345 ± 1 353 ± 4 TM, Bleached (K) 373 ± 2 350 ± 1 367 ± 4 Photonics 2021, 8, 302 6 of 14 k Unbleached 43 ± 14 18 ± 1 21 ± 3 k Bleached 45 ± 4 22 ± 3 31 ± 9 We now turn to the dynamical transition measurements which show for the first time We now turn to the dynamical transition measurements which show for the first that the protein turn-on is distinct from the solvent turn-on. The TD seen in the THz meas- time that the protein turn-on is distinct from the solvent turn-on. The T seen in the THz urements are different for the different proteins but for all three the TD increases with measurements are different for the different proteins but for all three the T increases with photobleaching. There has been no previous report of TD changing with the functional photobleaching. There has been no previous report of T changing with the functional state of the protein and the result is unexpected. The transition has been ascribed to the state of the protein and the result is unexpected. The transition has been ascribed to the slaving of protein dynamics to the solvent [33,34]. At low temperatures, the immobile wa- slaving of protein dynamics to the solvent [33,34]. At low temperatures, the immobile ter prevents large amplitude protein motions. A rapid increase in the solvent mobility at water prevents large amplitude protein motions. A rapid increase in the solvent mobility at T ~200K enables the protein to access the large amplitude anharmonic motions necessary T ~200K enables the protein to access the large amplitude anharmonic motions necessary for physiological function. While the specific activation energies for the solvent motions for physiological function. While the specific activation energies for the solvent motions can vary with the specific protein surface causing TD variation protein to protein, the TD can vary with the specific protein surface causing T variation protein to protein, the T D D change with photobleaching is not expected, as there is little change in the surface solvent change with photobleaching is not expected, as there is little change in the surface solvent exposur exposure. e. The The incr incr ease ease in inboth boththe the T TD and and T TM suggests suggests th that at possibly possibly the the pr protein otein str structural uctural D M dynamics dynamics ar are e playing playing a a rro ole le in inthe the temperatur temperature e dependence. dependence. T To o further further investigate investigate the the apparent change in the protein dynamics with photobleaching, we examine the frequency apparent change in the protein dynamics with photobleaching, we examine the frequency dependence dependenceof of t he the THz THz absorbance absorbance for fthe or th two e tw states. o staW tes e. first We consider first confitting sider fthe ittin fr g equency the fre- dependent absorbance for a given temperature to a power law frequency dependence, quency dependent absorbance for a given temperature to a power law frequency depend- a(w, T) w , where n is the power law factor for the temperature T. This phenomeno- ence, 𝛼 (𝜔 , 𝑇 )~𝜔 , where n is the power law factor for the temperature T. This phenome- logical approach is common for characterizing the typical broadband THz absorbance nological approach is common for characterizing the typical broadband THz absorbance for proteins [35–38]. The results in Figure 3 show that the temperature dependence of for proteins [35–38]. The results in Figure 3 show that the temperature dependence of the the power law substantially changes with photobleaching. For the unbleached RFP’s, power law substantially changes with photobleaching. For the unbleached RFP’s, the the power law is relatively flat below 200 K, then rapidly increases with a peak at 220 K power law is relatively flat below 200 K, then rapidly increases with a peak at 220 K and and then drops rapidly at higher temperatures. On the other hand, the bleached RFP’s is then drops rapidly at higher temperatures. On the other hand, the bleached RFP’s is rela- relatively flat for temperatures below 150 K, drops in value between 150 K and 200 K and tively flat for temperatures below 150 K, drops in value between 150 K and 200 K and then then remains relatively constant above 200 K. It is clear that the picosecond dynamics have remains relatively constant above 200 K. It is clear that the picosecond dynamics have changed with bleaching. changed with bleaching. Figure 3. Power law factors from fits to THz absorption frequency dependence for unbleached (left) Figure 3. Power law factors from fits to THz absorption frequency dependence for unbleached (left) and bleached (right) RFP samples: mCherry (red), TagRFP-T (blue), and mOrange2 (green) as a and bleached (right) RFP samples: mCherry (red), TagRFP-T (blue), and mOrange2 (green) as a func- function of temperature. The high temperature region above dynamical transition has been empha- tion of temperature. The high temperature region above dynamical transition has been emphasized. sized. The temperature dependence of the power law fitting is obviously complex and The temperature dependence of the power law fitting is obviously complex and does does not provide insight into the nature of the dynamical changes, except to dramatically not provide insight into the nature of the dynamical changes, except to dramatically demonstrate these changes are present. To better resolve the source of the dynamical demonstrate these changes are present. To better resolve the source of the dynamical transition temperature shift, we isolate the protein and solvent contributions to the THz absorbance. In the case of INS measurements, the separation of the solvent dynamics from the protein dynamics is accomplished by successive measurements of protonated protein with deuterated solvent and deuterated protein with protonated solvent [39,40]. For optical measurements, ideally the isolation of different dynamics can be done using distinct spectral signatures for the protein versus solvent. At room temperature, this is not possible as protein solution THz measurements are dominated by water, which is generally modeled by the sum of three Debye terms with relaxation times of 8.3 ps, 1 ps, and 0.2 ps and an intermolecular water vibration at 5.3 THz [41–43]. The relaxational absorption is sufficiently large that intramolecular protein vibrations can be entirely neglected. At low Photonics 2021, 8, x FOR PEER REVIEW 7 of 14 transition temperature shift, we isolate the protein and solvent contributions to the THz absorbance. In the case of INS measurements, the separation of the solvent dynamics from the protein dynamics is accomplished by successive measurements of protonated protein with deuterated solvent and deuterated protein with protonated solvent [39,40]. For opti- cal measurements, ideally the isolation of different dynamics can be done using distinct spectral signatures for the protein versus solvent. At room temperature, this is not possi- ble as protein solution THz measurements are dominated by water, which is generally modeled by the sum of three Debye terms with relaxation times of 8.3 ps, 1 ps, and 0.2 ps Photonics 2021, 8, 302 7 of 14 and an intermolecular water vibration at 5.3 THz [41–43]. The relaxational absorption is sufficiently large that intramolecular protein vibrations can be entirely neglected. At low temperatures, however, the THz absorbance changes considerably. The THz frequency temperatures, however, the THz absorbance changes considerably. The THz frequency water absorption drops dramatically [44–51] as the water relaxation times increase with water absorption drops dramatically [44–51] as the water relaxation times increase with decreasing temperature [52] moving the relaxational water absorption loss into the MHz decreasing temperature [52] moving the relaxational water absorption loss into the MHz range, while the water intermolecular vibrational resonance [41,42], centered at 5.3 THz, range, while the water intermolecular vibrational resonance [41,42], centered at 5.3 THz, remains at these lower temperatures. Thus, we can use the 5.3 THz resonance to monitor remains at these lower temperatures. Thus, we can use the 5.3 THz resonance to monitor the temperature dependent solvent dynamics. the temperature dependent solvent dynamics. To guide our fitting of the protein contribution, we performed molecular dynamics To guide our fitting of the protein contribution, we performed molecular dynamics simulations using quasiharmonic mode analysis (QHA) and dipole-dipole autocorrela- simulations using quasiharmonic mode analysis (QHA) and dipole-dipole autocorrelation tion calculations. For QHA the solvent is minimized, and the harmonic approximation calculations. For QHA the solvent is minimized, and the harmonic approximation focuses focuses on the collective motions of the hydrated protein only, whereas the dipole–dipole on the collective motions of the hydrated protein only, whereas the dipole–dipole autocor- autocorrelation calculations include all motions contributing to the absorbance [25]. The relation calculations include all motions contributing to the absorbance [25]. The VDOS VDOS and autocorrelation results are shown in Figure S2 in the Supplementary Infor- and autocorrelation results are shown in Figure S2 in the Supplementary Information. mation. The QHA calculated isotropic absorptivity for the three RFPs below and above the The QHA calculated isotropic absorptivity for the three RFPs below and above the T are shown in Figure 4. There is negligible difference among the RFPs for these cal- TD are shown in Figure 4. There is negligible difference among the RFPs for these calcula- culations. The ordering in the net absorbance seen in Figure 2 is reproduced by the tions. The ordering in the net absorbance seen in Figure 2 is reproduced by the autocorre- autocorrelation calculations however (see Figure S2 in Supplementary Information). The lation calculations however (see Figure S2 in Supplementary Information). The calculated calculated QHA isotropic absorbance exhibits similar Lorentzian-like absorption peaks QHA isotropic absorbance exhibits similar Lorentzian-like absorption peaks centered centered near 0.5 THz which become narrow and slightly shift up in frequency with near 0.5 THz which become narrow and slightly shift up in frequency with incr easing increasing temperature. temperature. Figure 4. Isotropic THz absorption calculated by quasi-harmonic analysis at 200 K (A) and 300 K Figure 4. Isotropic THz absorption calculated by quasi-harmonic analysis at 200 K (A) and 300 K (B). (B). High temperature absorption shows a Lorentzian-like peak centered at ~0.5 THz for all three High temperature absorption shows a Lorentzian-like peak centered at ~0.5 THz for all three RFPs. RFPs. The same color scale is used for both plots. The same color scale is used for both plots. Based on the QHA results, we fit our measured absorbance with two Lorentzians Based on the QHA results, we fit our measured absorbance with two Lorentzians with with one frequency set at 5.3 THz for the intermolecular water vibration. Note that only one frequency set at 5.3 THz for the intermolecular water vibration. Note that only the the tail of the water resonance contributes to the spectral feature as background absorp- tail of the water resonance contributes to the spectral feature as background absorption tion and a slight change in central frequency will not alter other fitting parameters. The and a slight change in central frequency will not alter other fitting parameters. The results results for TagRFP-T in Figure 5A,B show that the model fits the data well for the un- for TagRFP-T in Figure 5A,B show that the model fits the data well for the unbleached bleached and bleached states, a substantial improvement over the simple power law fits and bleached states, a substantial improvement over the simple power law fits (see Sup- (see Supplemental Figure S3). The different colors represent temperatures from 80 K (red) plemental Figure S3). The different colors represent temperatures from 80 K (red) up to 275 K (purple). The full list of the temperatures is in the Supplemental Information. The jump in absorption at room temperature is due to liquid water forming due to the thawing of the solution. In Figure 6, we show the extracted resonant frequency and linewidths for the double resonant fits as a function of temperature. The temperature dependent behavior of the protein resonant band is distinctly different from the water resonance. Specifically, Figure 6A,C shows the amplitude and linewidth for the 5.3 THz water band for TagRFP-T unbleached and bleached, respectively. In both cases, the amplitude begins to rapidly decrease, and linewidth rapidly begins to increase at 200 K. The solvent transition is unaffected by the bleaching, as one might expect for the solvent dynamics. As the temperature increases, the water resonance broadens as the amplitude decreases with the net integrated intensity of the solvent intermolecular excitations remaining constant. The Photonics 2021, 8, x FOR PEER REVIEW 8 of 14 up to 275 K (purple). The full list of the temperatures is in the Supplemental Information. The jump in absorption at room temperature is due to liquid water forming due to the thawing of the solution. In Figure 6, we show the extracted resonant frequency and lin- ewidths for the double resonant fits as a function of temperature. The temperature de- pendent behavior of the protein resonant band is distinctly different from the water reso- nance. Specifically, Figure 6A,C shows the amplitude and linewidth for the 5.3 THz water band for TagRFP-T unbleached and bleached, respectively. In both cases, the amplitude begins to rapidly decrease, and linewidth rapidly begins to increase at 200 K. The solvent transition is unaffected by the bleaching, as one might expect for the solvent dynamics. As the temperature increases, the water resonance broadens as the amplitude decreases with the net integrated intensity of the solvent intermolecular excitations remaining con- stant. The broadening is consistent with additional excitations accessible with increasing Photonics 2021, 8, 302 8 of 14 thermally activated mobile waters. Figure 6B shows the frequency, amplitude, and lin- ewidth of the low frequency band for the unbleached protein. The central frequency is nominally at 0.6 THz and blue shifts at higher temperatures, in agreement with the QHA broadening is consistent with additional excitations accessible with increasing thermally peak in Figure 4. Both the amplitude and linewidth increase at a transition temperature of activated mobile waters. Figure 6B shows the frequency, amplitude, and linewidth of 210 K the fo low r th fr e equency unbleaband chedfor Tathe gRFP unbleached -T. That pr is, otein. the d The yna central micalfr o equency nset tempe is nominally rature for the low at 0.6 THz and blue shifts at higher temperatures, in agreement with the QHA peak in frequency protein motions is higher than that of the solvent. We will define two onset Figure 4. Both the amplitude and linewidth increase at a transition temperature of 210 K for temperatures, from the sharp turn-on points of each curve, TDS for the solvent and TDP for the unbleached TagRFP-T. That is, the dynamical onset temperature for the low frequency the protein. Figure 6D shows the results for the photobleached TagRFP-T. The photo- protein motions is higher than that of the solvent. We will define two onset temperatures, bleached state TDP shifts up to 230 K, while the solvent TDS remains the same. mCherry has from the sharp turn-on points of each curve, T for the solvent and T for the protein. DS DP Figure 6D shows the results for the photobleached TagRFP-T. The photobleached state T the same result (see Figure S4 in Supplementary Information). The picosecon DP d water dy- shifts up to 230 K, while the solvent T remains the same. mCherry has the same result DS namics turn-on at 200 K for both unbleached and bleached protein, whereas the un- (see Figure S4 in Supplementary Information). The picosecond water dynamics turn-on bleached TDP is 210 K and bleached is 230 K. Just as the melting temperature shifts up with at 200 K for both unbleached and bleached protein, whereas the unbleached T is 210 K DP bleaching, so does the turn-on for the picosecond dynamics. For mOrange2, the solvent and bleached is 230 K. Just as the melting temperature shifts up with bleaching, so does tranthe sititurn-on on is afor gaithe n apicosecond t 200 K fo dynamics. r both un For blea mOrange2, ched anthe d b solvent leached transition protein iss again . For at unbleached 200 K for both unbleached and bleached proteins. For unbleached mOrange2, the protein mOrange2, the protein dynamical turn- on is again at 210 K, but the bleached mOrange2 dynamical turn- on is again at 210 K, but the bleached mOrange2 is substantially different, is substantially different, with the protein dynamical transition nearly absent and a slight with the protein dynamical transition nearly absent and a slight inflection at 220 K. In all inflection at 220 K. In all three cases, the protein temperature dependence is not identical three cases, the protein temperature dependence is not identical to the solvent. to the solvent. 80 K (red) to 275 K (purple ) 80 K (red) to 275 K (purple ) Figure Figure 5. Co 5. Comparisons mparisons o of f t the he fr fr equency equenc dependent y depend THz ent m TH olar z m absorptivity olar absowith rptivit resonance y with fitting resonance fitting 1 1 −1 −1 lines for photoactive (A) and photobleached (B) TagRFP-T. The offset of 2 mM cm was applied to lines for photoactive (A) and photobleached (B) TagRFP-T. The offset of 2 mM cm was applied distinguish different temperatures, with the lowest temperature (80 K) in red and highest temperature to distinguish different temperatures, with the lowest temperature (80 K) in red and highest tem- (275 K) in purple. The top figure illustrates the decomposition of the absorption into two Lorentzian perature (275 K) in purple. The top figure illustrates the decomposition of the absorption into two resonances: protein long-range intramolecular resonance centered at ~0.6 THz and larger-amplitude Lorentzian resonances: protein long-range intramolecular resonance centered at ~0.6 THz and water intermolecular resonance at 5.3 THz. larger-amplitude water intermolecular resonance at 5.3 THz. Photonics 2021, 8, 302 9 of 14 Photonics 2021, 8, x FOR PEER REVIEW 9 of 14 Figure 6. Resonance fitting parameters for photoactive (top) and photobleached (bottom) TagRFP- Figure 6. Resonance fitting parameters for photoactive (top) and photobleached (bottom) TagRFP-T. T. (A,C) Amplitude (circle, red) and linewidth (triangle, blue) corresponding to water intermolecu- (A,C) Amplitude (circle, red) and linewidth (triangle, blue) corresponding to water intermolecular lar resonance centered at 5.3 THz. (B,D) Amplitude (circle, red), linewidth (triangle, blue), and cen- resonance centered at 5.3 THz. (B,D) Amplitude (circle, red), linewidth (triangle, blue), and central tral frequency (square, green) of the protein intramolecular resonant band; the photobleached state frequency (square, green) of the protein intramolecular resonant band; the photobleached state shows shows higher collectivity of the motion. The interpolated lines are used as a guide for the eye. higher collectivity of the motion. The interpolated lines are used as a guide for the eye. 4. Discussion 4. Discussion The correlation between structural resilience and thermal stability is perhaps intui- The correlation between structural resilience and thermal stability is perhaps intuitive tive and consistent with previous neutron studies; however, the correlation of the protein and consistent with previous neutron studies; however, the correlation of the protein dynamical transition temperature with thermal stability is somewhat surprising. The dynamical transition temperature with thermal stability is somewhat surprising. The rapid rapid onset in the protein dynamics with temperature has long been understood to be onset in the protein dynamics with temperature has long been understood to be associated associated with the thermally activated motions of the surrounding solvent [33], so any with the thermally activated motions of the surrounding solvent [33], so any trend with trend with protein structural stability is not expected. The protein dynamical transition protein structural stability is not expected. The protein dynamical transition arises from arises from the need to break hydrogen bonds within the solvent cage to accommodate the need to break hydrogen bonds within the solvent cage to accommodate motions. This motions. This phenomenon has been termed the slaving of the protein’s dynamics to the phenomenon has been termed the slaving of the protein’s dynamics to the solvent. At the solvent. At the same time, it is understood that the solvent excitations are influenced by same time, it is understood that the solvent excitations are influenced by the specific protein the specific protein surface [34]. TD has been observed to vary for different measurements surface [34]. T has been observed to vary for different measurements [29,53,54] and for [29,53,54] and for different proteins for a single technique [32]. For example, INS meas- different proteins for a single technique [32]. For example, INS measurements using the urements using the same energy beamline for different proteins reveal onset temperatures same energy beamline for different proteins reveal onset temperatures as low as 200 K to as as low as 200 K to as high as 250 K. This variation has been attributed to differences in the high as 250 K. This variation has been attributed to differences in the specific protein-solvent specific protein-solvent surface interaction [40]. The direct dependence of the protein dy- surface interaction [40]. The direct dependence of the protein dynamical onset on the water mobility namical onset onset appear on the wa ed ter to be mo confirmed bility onset in appe twoaneutr red to on be studies confirmed wher in e tw both o nthe eutrsolvent on stud- transition ies where and both pr otein the so transition lvent trawer nsiti eoseparate n and pr ly omeasur tein traed nsi[ti 55 on ,56 were ]. For sepa example, rately while measured T DS for maltose binding protein (MBP) and hen egg white lysozyme (HEWL) are different; in [55,56]. For example, while TDS for maltose binding protein (MBP) and hen egg white ly- both cases, the T coincides with T . We note that both MBP and HEWL have two lobes sozyme (HEWL) are different; in both cases, the TDP coincides with TDS. We note that both DP DS surrounding a binding cleft. MBP is almost entirely -helical with a small -sheet region at MBP and HEWL have two lobes surrounding a binding cleft. MBP is almost entirely α- the binding site, whereas HEWL has one lobe that is mainly -helix and the other mainly - helical with a small β-sheet region at the binding site, whereas HEWL has one lobe that is sheet. The picture that emerged is that for a given protein, the T is dictated by the average mainly α-helix and the other mainly β-sheet. The picture thaDS t emerged is that for a given solvent binding energies to the specific protein surface, and the protein structural dynamics protein, the TDS is dictated by the average solvent binding energies to the specific protein follow the thermal activation of these surface-solvent excitations. In the results presented surface, and the protein structural dynamics follow the thermal activation of these sur- here, we see a somewhat extraordinary and different result where the protein dynamical face-solvent excitations. In the results presented here, we see a somewhat extraordinary turn-on clearly is different than the solvent turn-on. For each of the three proteins, the and different result where the protein dynamical turn-on clearly is different than the sol- solvent T remains essentially unchanged in the two photostates, indicating little change DS vent turn-on. For each of the three proteins, the solvent TDS remains essentially unchanged in the solvent–protein interactions. This is consistent with structural measurements of in the two photostates, indicating little change in the solvent–protein interactions. This is KillerRed and IrisFP using similar bleaching conditions [6,57]. For both KillerRed and consistent with structural measurements of KillerRed and IrisFP using similar bleaching IrisFP, there is little structural change with photobleaching. The slight decrease in THz conditions [6,57]. For both KillerRed and IrisFP, there is little structural change with pho- absorbance also is consistent with the structure remaining intact, as it has been found the tobleaching. The slight decrease in THz absorbance also is consistent with the structure low temperature THz absorbance increases substantially with structural loss. Photonics 2021, 8, 302 10 of 14 The protein dynamical onsets however are not the same as the solvent onsets and are dependent on photostate. Even in the unbleached state, T is shifted up relative to the DP solvent T for all three proteins. This difference with the previous comparisons of T DS DP and T for MBP and HEWL may in part arise from the more rigid -barrel structure of DS the RFPs. The shift increases dramatically by 20 K with photobleaching for mCherry and TagRFP-T. This shift has not been reported previously and requires a further examination of the solvent slaving idea. As previously discussed, at low temperatures, the water motions are limited, trapping the protein configuration. As the temperature increases, water cluster motions are thermally activated, lifting the constraints on the protein dynamics. If in the measurement frequency range, the motions are highly localized, then the number of thermally activated mobile water clusters needed for the motions to occur is small, and temperature dependence of the protein dynamics will closely follow the solvent dynamical transition. Larger populations of mobile water clusters are required to execute delocalized motions, thus leading to an increase in their turn-on temperature relative to T . In DS the picosecond range, it has been shown that solvent fluctuations have an Arrhenius temperature dependence [6,34,58]. We can simplistically relate the Arrhenius dependence to the thermal population of mobile waters, N(T): k T N(T) = N e (2) tot where N is the total number of waters, E is the activation energy, k is the Boltzmann tot A b constant, and T is the temperature. T is then the temperature at which there is a sufficient DS high number of mobile waters to detect by THz absorbance, whereas T indicates that DP the population of mobile waters necessary for the protein motions to be detectable by THz absorbance. We can define the fractional increase in the necessary mobile water population for the protein contribution as f : N(T ) DP f = (3) N(T ) DS Previous RMSD measurements have reported activation energies between 20–40 kJ/mol [58–64]. Using this range of activation energies and T and T we extract DS DP from the THz measurements, the number of mobile waters needed to detect the unbleached protein dynamics is 2–3 times the number to detect the solvent turn on, whereas for the bleached state, the number increases to 5–23 times. As there is no solvent accessible surface area (SASA) change with bleaching, this large increase in the mobile waters necessary for the picosecond protein motions contributing to the THz signal suggests that the bleached state motions are more spatially extended, with more distant regions moving in concert. The increased collectivity of picosecond-timescale motions in the photobleached state suggested by these THz measurements is consistent with enhanced internal coupling through water channels formed by photoinduced alteration of the internal protein structure. For each of the two RFP’s [6,57] whose structure has been solved in both unbleached and bleached states, under bleaching conditions similar to those used in our studies, the -barrel structure is nearly unchanged; however, a CAVER analysis shows that additional water channels appear in the photobleached state (see Figure S5 in Supplementary Information). The dissipation of the excess energy via strong structural fluctuations provides an avenue for the water channel formation. These additional water channels can provide H-bond coupling within the -barrel interior. The impact of the water channels on the collectivity is evident in a comparison between the CAVER water channel maps and the B-factor maps for KillerRed and IrisFP (see Figure S5 in Supplementary Information). In the photobleached state, the B-factor uniformity increases in the same regions as the water channels form. The enhanced coupling provided by the water channels is also consistent with the increase in thermal stability that we measure. Finally, we note that these same water channels likely are responsible for the loss in fluorescence in the photobleached state. All organic fluorophores, including RFPs, suffer from irreversible photobleaching after exposure to Photonics 2021, 8, 302 11 of 14 prolonged and excessive illumination [65]. For RFPs, candidate mechanisms leading to fluorescence loss are oxidation and/or cis-trans isomerization of the chromophore [6,12,23]. Under the lower intensity illumination conditions of our study, oxidation is thought to be the dominant mechanism whereas the cis-trans isomerization mechanism occurs for more extreme conditions [66–68]. While oxygen is required for initial chromophore maturation, it has been found that photobleaching for mOrange2 and TagRFP-T is oxygen sensitive, and oxygen-free conditions result in the improved photostability [5]. Photobleaching via oxygen diffusion through the water channel in 7– 10 region in mCherry has also been discussed [23]. The presence of water channels in the photobleached state can explain an increase in dynamical collectivity, an increase in thermal stability and a loss of fluorescence by increasing oxygen access leading to trapping of the chromophore in a protonated state [57,69–71]. 5. Conclusions We find an increase in structural stability and vibrational collectivity with RFP photo- bleaching, consistent with enhanced intramolecular coupling via internal water channel formation with prolonged photo excitation. Both the strength of THz absorption and THz low temperature resilience correlate with thermal stability. The temperature dependent THz absorbance spectra can be used to separate the solvent and protein dynamical onsets. We find the dynamical onset of the protein motions does not coincide with that of the solvent and that it increases in the photobleached state. We suggest that the shifting of the protein dynamical onset relative to the solvent arises from the threshold mobile water population needed for the protein motions to be accessible. Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/photonics8080302/s1. Figure S1: Melting Measurements of FP. Figure S2: Absorption of FP. Figure S3: Molar Absorptivity of FP. Figure S4: Temperature dependent parameter fit to THz absorption spectra. Figure S5: Debye–Waller B factor surface plots. Figure S6: Fluorescence peak of mOrange. Author Contributions: M.X. performed THz TDS, CD, light scattering, fluorescence measurements, calculations, and analyzed data; D.G. performed THz TDS, fluorescence measurements, and analyzed data; R.J. provided samples; A.M. designed, and conceived measurements. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by National Science Foundation MRIˆ2 grant DBI2959989, IDBR grant DBI1556359, MCB grant MCB1616529, the Department of Energy BES grant DE-SC0016317, and the NSF Physics Frontier Center at JILA, NSF PHY 1734006. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Acknowledgments: The authors thank Prem P. Chapagain at Florida International University for providing RFP topology and parameter files. RJ is a staff member in the Quantum Physics Division of NIST. Certain commercial equipment, instruments, or materials are identified in this paper in order to specify the experimental procedure adequately. Such identification is not intended to imply that the materials or equipment identified are necessarily the best available for the purpose. Conflicts of Interest: The authors declare no conflict of interest. References 1. Tolbert, L.M.; Baldridge, A.; Kowalik, J.; Solntsev, K.M. Collapse and Recovery of Green Fluorescent Protein Chromophore Emission through Topological Effects. Accounts Chem. Res. 2012, 45, 171–181. [CrossRef] [PubMed] 2. Zimmer, M. Green fluorescent protein (GFP): Applications, structure, and related photophysical behavior. Chem. Rev. 2002, 102, 759–781. [CrossRef] [PubMed] 3. Tsien, R.Y. The Green Fluorescent Protein. Annu. Rev. Biochem. 1998, 67, 509–544. [CrossRef] [PubMed] 4. Stavrov, S.S.; Solntsev, K.M.; Tolbert, L.M.; Huppert, D. Probing the decay coordinate of the green fluorescent protein: Arrest of cis-trans isomerization by the protein significantly narrows the fluorescence spectra. J. Am. Chem. Soc. 2006, 128, 1540–1546. [CrossRef] Photonics 2021, 8, 302 12 of 14 5. Shaner, N.C.; Lin, M.Z.; McKeown, M.R.; Steinbach, P.A.; Hazelwood, K.L.; Davidson, M.W.; Tsien, R.Y. Improving the photostability of bright monomeric orange and red fluorescent proteins. Nat. Methods 2008, 5, 545–551. [CrossRef] 6. Carpentier, P.; Violot, S.; Blanchoin, L.; Bourgeois, D. Structural basis for the phototoxicity of the fluorescent protein KillerRed. FEBS Lett. 2009, 583, 2839–2842. [CrossRef] [PubMed] 7. Kao, Y.-T.; Zhu, X.; Min, W. Protein-flexibility mediated coupling between photoswitching kinetics and surrounding viscosity of a photochromic fluorescent protein. Proc. Natl. Acad. Sci. USA 2012, 109, 3220–3225. [CrossRef] 8. Mizuno, H.; Mal, T.K.; Wälchli, M.; Kikuchi, A.; Fukano, T.; Ando, R.; Jeyakanthan, J.; Taka, J.; Shiro, Y.; Ikura, M.; et al. Light-dependent regulation of structural flexibility in a photochromic fluorescent protein. Proc. Natl. Acad. Sci. USA 2008, 105, 9227–9232. [CrossRef] 9. Shaner, N.C.; Campbell, R.E.; Steinbach, P.A.; Giepmans, B.N.G.; Palmer, A.E.; Tsien, R.Y. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 2004, 22, 1567–1572. [CrossRef] 10. Shu, X.K.; Shaner, N.C.; Yarbrough, C.A.; Tsien, R.Y.; Remington, S.J. Novel chromophores and buried charges control color in mFruits. Biochemistry 2006, 45, 9639–9647. [CrossRef] 11. Liu, R.; Liang, Q.N.; Du, S.Q.; Hu, X.J.; Ding, Y. The crystal structure of red fluorescent protein TagRFP-T reveals the mechanism of its superior photostability. Biochem. Biophys. Res. Commun. 2016, 477, 229–234. [CrossRef] 12. Dean, K.M.; Lubbeck, J.L.; Binder, J.K.; Schwall, L.R.; Jimenez, R.; Palmer, A.E. Analysis of red-fluorescent proteins provides insight into dark-state conversion and photodegradation. Biophys. J. 2011, 101, 961–969. [CrossRef] 13. Niessen, K.A.; Xu, M.; Markelz, A.G. Terahertz optical measurements of correlated motions with possible allosteric function. Biophys. Rev. 2015, 7, 201–216. [CrossRef] 14. George, D.K.; Knab, J.R.; He, Y.; Kumauchi, M.; Birge, R.R.; Hoff, W.D.; Markelz, A.G. Photoactive Yellow Protein Terahertz Response: Hydration, Heating and Intermediate States. IEEE Trans. Terahertz Sci. Technol. 2013, 3, 288–294. [CrossRef] 15. Yamamoto, N.; Ohta, K.; Tamura, A.; Tominaga, K. Broadband Dielectric Spectroscopy on Lysozyme in the Sub-Gigahertz to Terahertz Frequency Regions: Effects of Hydration and Thermal Excitation. J. Phys. Chem. B 2016, 120, 4743–4755. [CrossRef] [PubMed] 16. Karplus, M.; Kuriyan, J. Molecular dynamics and protein function. Proc. Natl. Acad. Sci. USA 2005, 102, 6679–6685. [CrossRef] [PubMed] 17. Yada, H.; Nagai, M.; Tanaka, K. Origin of the fast relaxation component of water and heavy water revealed by terahertz time-domain attenuated total reflection spectroscopy. Chem. Phys. Lett. 2008, 464, 166–170. [CrossRef] 18. van Exter, M.; Fattinger, C.; Grischkowsky, D. Terahertz time-domain spectroscopy of water vapor. Opt. Lett. 1989, 14, 1128–1130. [CrossRef] 19. Wang, S.; Zhang, X.C. Pulsed terahertz tomography. J. Phys. D Appl. Phys. 2004, 37, R1. [CrossRef] 20. He, Y.; Ku, P.I.; Knab, J.R.; Chen, J.Y.; Markelz, A.G. Protein Dynamical Transition Does Not Require Protein Structure. Phys. Rev. Lett. 2008, 101, 178103. [CrossRef] [PubMed] 21. Brooks, B.R.; Bruccoleri, R.E.; Olafson, B.D.; States, D.J.; Swaminathan, S.; Karplus, M. CHARMM: A program for macromolecular energy, minimization, and dynamics calculations. J. Comput. Chem. 1983, 4, 187–217. [CrossRef] 22. Brooks, B.R.; Brooks, C.L.; MacKerell, A.D.; Nilsson, L.; Petrella, R.J.; Roux, B.; Won, Y.; Archontis, G.; Bartels, C.; Boresch, S.; et al. CHARMM: The Biomolecular Simulation Program. J. Comput. Chem. 2009, 30, 1545–1614. [CrossRef] [PubMed] 23. Chapagain, P.P.; Regmi, C.K.; Castillo, W. Fluorescent protein barrel fluctuations and oxygen diffusion pathways in mCherry. J. Chem. Phys. 2011, 135, 235101. [CrossRef] 24. Jo, S.; Kim, T.; Iyer, V.G.; Im, W. CHARMM-GUI: A web-based graphical user interface for CHARMM. J. Comput. Chem. 2008, 29, 1859–1865. [CrossRef] 25. He, Y.; Chen, J.; Knab, J.; Zheng, W.; Markelz, A. Evidence of Protein Collective Motions on the Picosecond Timescale. Biophys. J. 2011, 100, 1058–1065. [CrossRef] 26. Ding, T.; Middelberg, A.P.J.; Huber, T.; Falconer, R.J. Far-infrared spectroscopy analysis of linear and cyclic peptides, and lysozyme. Vib. Spectrosc. 2012, 61, 144–150. [CrossRef] 27. Heyden, M.; Tobias, D.J.; Matyushov, D.V. Terahertz absorption of dilute aqueous solutions. J. Chem. Phys. 2012, 137, 235103. [CrossRef] 28. Nakagawa, H.; Joti, Y.; Kitao, A.; Kataoka, M. Hydration Affects Both Harmonic and Anharmonic Nature of Protein Dynamics. Biophys. J. 2008, 95, 2916–2923. [CrossRef] 29. Doster, W.; Cusack, S.; Petry, W. Dynamical transition of myoglobin revealed by inelastic neutron scattering. Nature 1989, 337, 754–756. [CrossRef] 30. Zaccai, G. How soft is a protein? A protein dynamics force constant measured by neutron scattering. Science 2000, 288, 1604–1607. [CrossRef] 31. Heyden, M.; Tobias, D.J. Spatial Dependence of Protein-Water Collective Hydrogen-Bond Dynamics. Phys. Rev. Lett. 2013, 111, 218101. [CrossRef] 32. Stadler, A.M.; Garvey, C.J.; Bocahut, A.; Sacquin-Mora, S.; Digel, I.; Schneider, G.J.; Natali, F.; Artmann, G.M.; Zaccai, G. Thermal fluctuations of haemoglobin from different species: Adaptation to temperature via conformational dynamics. J. R. Soc. Interface 2012, 9, 2845–2855. [CrossRef] [PubMed] Photonics 2021, 8, 302 13 of 14 33. Doster, W. The dynamical transition of proteins, concepts and misconceptions. Eur. Biophys. J. 2008, 37, 591–602. [CrossRef] [PubMed] 34. Chen, G.; Fenimore, P.W.; Frauenfelder, H.; Mezei, F.; Swenson, J.; Young, R.D. Protein fluctuations explored by inelastic neutron scattering and dielectric relaxation spectroscopy. Philos. Mag. 2008, 88, 3877–3883. [CrossRef] 35. Kawaguchi, S.; Kambara, O.; Shibata, M.; Kandori, H.; Tominaga, K. Low-frequency dynamics of bacteriorhodopsin studied by terahertz time-domain spectroscopy. Phys. Chem. Chem. Phys. 2010, 12, 10255–10262. [CrossRef] 36. Yamamoto, K.; Tominaga, K.; Sasakawa, H.; Tamura, A.; Murakami, H.; Ohtake, H.; Sarukura, N. Terahertz Time-Domain Spectroscopy of Amino Acids and Polypeptides. Biophys. J. 2005, 89, L22–L24. [CrossRef] 37. Perticaroli, S.; Nickels, J.D.; Ehlers, G.; Sokolov, A.P. Rigidity, Secondary Structure, and the Universality of the Boson Peak in Proteins. Biophys. J. 2014, 106, 2667–2674. [CrossRef] 38. Mori, T.; Jiang, Y.; Fujii, Y.; Kitani, S.; Mizuno, H.; Koreeda, A.; Motoji, L.; Tokoro, H.; Shiraki, K.; Yamamoto, Y.; et al. Detection of boson peak and fractal dynamics of disordered systems using terahertz spectroscopy. Phys. Rev. E 2020, 102, 022502. [CrossRef] 39. Schirò, G.; Fichou, Y.; Gallat, F.-X.; Wood, K.; Gabel, F.; Moulin, M.; Härtlein, M.; Heyden, M.; Colletier, J.-P.; Orecchini, A.; et al. Translational diffusion of hydration water correlates with functional motions in folded and intrinsically disordered proteins. Nat. Commun. 2015, 6, 6490. [CrossRef] 40. Wood, K.; Frölich, A.; Paciaroni, A.; Moulin, M.; Härtlein, M.; Zaccai, G.; Tobias, D.J.; Weik, M. Coincidence of Dynamical Transitions in a Soluble Protein and Its Hydration Water: Direct Measurements by Neutron Scattering and MD Simulations. J. Am. Chem. Soc. 2008, 130, 4586–4587. [CrossRef] 41. Møller, U.; Cooke, D.G.; Tanaka, K.; Jepsen, P.U. Terahertz reflection spectroscopy of Debye relaxation in polar liquids [Invited]. J. Opt. Soc. Am. B 2009, 26, A113–A125. [CrossRef] 42. Penkov, N.; Shvirst, N.; Yashin, V.; Fesenko, E.; Fesenko, E. Terahertz Spectroscopy Applied for Investigation of Water Structure. J. Phys. Chem. B 2015, 119, 12664–12670. [CrossRef] 43. Vinh, N.Q.; Sherwin, M.S.; Allen, S.J.; George, D.K.; Rahmani, A.J.; Plaxco, K.W. High-precision gigahertz-to-terahertz spec- troscopy of aqueous salt solutions as a probe of the femtosecond-to-picosecond dynamics of liquid water. J. Chem. Phys. 2015, 142, 164502. [CrossRef] 44. Twardowski, J.; Anzenbacher, P.; Masson, M.R. Raman and IR Spectroscopy in Biology and Biochemistry; Ellis Horwood: Chichester, UK, 1994. 45. Genzel, L.; Kremer, F.; Poglitsch, A.; Bechtold, G. Relaxation processes on a picosecond time scale in hemoglobin and poly(L- alanine) observed by millimeter-wave spectroscopy. Biopolymers 1983, 22, 1715–1729. [CrossRef] 46. Genzel, L.; Santo, L.; Shen, S.C. Far-Infrared Spectroscopy of Biomolecules. In Spectroscopy of Biological Molecules: Theory and Applications—Chemistry, Physics, Biology, and Medicine; Sandorfy, C., Theophanides, T., Eds.; Springer: Dordrecht, The Netherlands, 1984; pp. 609–619. 47. Poglitsch, A.; Kremer, F.; Genzel, L. Picosecond relaxations in hydrated lysozyme observed by mm-wave spectroscopy. J. Mol. Biol. 1984, 173, 137–142. [CrossRef] 48. Powell, J.W.; Edwards, G.S.; Genzel, L.; Kremer, F.; Wittlin, A.; Kubasek, W.; Peticolas, W. Investigation of far-infrared vibrational modes in polynucleotides. Phys. Rev. A 1987, 35, 3929–3939. [CrossRef] 49. Powell, J.W.; Peticolas, W.L.; Genzel, L. Observation of the far-infrared spectrum of five oligonucleotides. J. Mol. Struct. 1991, 247, 107–118. [CrossRef] 50. Wittlin, A.; Genzel, L.; Kremer, F.; Häseler, S.; Poglitsch, A.; Rupprecht, A. Far-infrared spectroscopy on oriented films of dry and hydrated DNA. Phys. Rev. A 1986, 34, 493–500. [CrossRef] [PubMed] 51. Zakaria, H.A.; Fischer, B.M.; Bradley, A.P.; Jones, I.; Abbott, D.; Middelberg, A.P.; Falconer, R.J. Low-frequency spectroscopic analysis of monomeric and fibrillar lysozyme. Appl. Spectrosc. 2011, 65, 260–264. [CrossRef] 52. Kaatze, U. Complex permittivity of water as a function of frequency and temperature. J. Chem. Eng. Data 1989, 34, 371–374. [CrossRef] 53. Stadler, A.M.; Digel, I.; Embs, J.P.; Unruh, T.; Tehei, M.; Zaccai, G.; Buldt, G.; Artmann, G.M. From powder to solution: Hydration dependence of human hemoglobin dynamics correlated to body temperature. Biophys. J. 2009, 96, 5073–5081. [CrossRef] 54. Frolov, E.N.; Gvosdev, R.; Goldanskii, V.I.; Parak, F.G. Differences in the dynamics of oxidized and reduced cytochrome c measured by Mössbauer spectroscopy. JBIC J. Biol. Inorg. Chem. 1997, 2, 710–713. [CrossRef] 55. Paciaroni, A.; Orecchini, A.; Cornicchi, E.; Marconi, M.; Petrillo, C.; Haertlein, M.; Moulin, M.; Sacchetti, F. Coupled thermal fluctuations of proteins and protein hydration water on the picosecond timescale. Philos. Mag. 2008, 88, 4071–4077. [CrossRef] 56. Chen, S.-H.; Liu, L.; Fratini, E.; Baglioni, P.; Faraone, A.; Mamontov, E. Observation of fragile-to-strong dynamic crossover in protein hydration water. Proc. Natl. Acad. Sci. USA 2006, 103, 9012–9016. [CrossRef] [PubMed] 57. Duan, C.; Adam, V.; Byrdin, M.; Ridard, J.; Kieffer-Jaquinod, S.; Morlot, C.; Arcizet, D.; Demachy, I.; Bourgeois, D. Structural Evidence for a Two-Regime Photobleaching Mechanism in a Reversibly Switchable Fluorescent Protein. J. Am. Chem. Soc. 2013, 135, 15841–15850. [CrossRef] [PubMed] 58. Doster, W. The protein-solvent glass transition. Biochim. Biophys. Acta Proteins Proteom. 2010, 1804, 3–14. [CrossRef] 59. Bragger, J.M.; Dunn, R.V.; Daniel, R.M. Enzyme activity down to-100 degrees C. Biochim. Biophys. Acta Protein Struct. Molec. Enzym. 2000, 1480, 278–282. [CrossRef] Photonics 2021, 8, 302 14 of 14 60. Malardier-Jugroot, C.; Head-Gordon, T. Separable cooperative and localized translational motions of water confined by a chemically heterogeneous environment. Phys. Chem. Chem. Phys. 2007, 9, 1962–1971. [CrossRef] [PubMed] 61. Shinyashiki, N.; Yamamoto, W.; Yokoyama, A.; Yoshinari, T.; Yagihara, S.; Kita, R.; Ngai, K.L.; Capaccioli, S. Glass Transitions in Aqueous Solutions of Protein (Bovine Serum Albumin). J. Phys. Chem. B 2009, 113, 14448–14456. [CrossRef] 62. Schiro, G.; Caronna, C.; Natali, F.; Cupane, A. Direct Evidence of the Amino Acid Side Chain and Backbone Contributions to Protein Anharmonicity. J. Am. Chem. Soc. 2010, 132, 1371–1376. [CrossRef] [PubMed] 63. Fenimore, P.W.; Frauenfelder, H.; McMahon, B.H.; Young, R.D. Proteins are paradigms of stochastic complexity. Phys. A 2005, 351, 1–13. [CrossRef] 64. Fenimore, P.W.; Frauenfelder, H.; McMahon, B.H.; Young, R.D. Bulk-solvent and hydration-shell fluctuations, similar to alpha- and beta-fluctuations in glasses, control protein motions and functions. Proc. Natl. Acad. Sci. USA 2004, 101, 14408–14413. [CrossRef] 65. Schenk, A.; Ivanchenko, S.; Röcker, C.; Wiedenmann, J.; Nienhaus, G.U. Photodynamics of Red Fluorescent Proteins Studied by Fluorescence Correlation Spectroscopy. Biophys. J. 2004, 86, 384–394. [CrossRef] 66. Bell, A.F.; Stoner-Ma, D.; Wachter, R.M.; Tonge, P.J. Light-driven decarboxylation of wild-type green fluorescent protein. J. Am. Chem. Soc. 2003, 125, 6919–6926. [CrossRef] 67. Liu, Y.; Kim, H.R.; Heikal, A.A. Structural basis of fluorescence fluctuation dynamics of green fluorescent proteins in acidic environments. J. Phys. Chem. B 2006, 110, 24138–24146. [CrossRef] [PubMed] 68. Habuchi, S.; Cotlet, M.; Gensch, T.; Bednarz, T.; Haber-Pohlmeier, S.; Rozenski, J.; Dirix, G.; Michiels, J.; Vanderleyden, J.; Heberle, J.; et al. Evidence for the isomerization and decarboxylation in the photoconversion of the red fluorescent protein DsRed. J. Am. Chem. Soc. 2005, 127, 8977–8984. [CrossRef] [PubMed] 69. Jiménez-Banzo, A.; Nonell, S.; Hofkens, J.; Flors, C. Singlet Oxygen Photosensitization by EGFP and its Chromophore HBDI. Biophys. J. 2008, 94, 168–172. [CrossRef] 70. Jimenez-Banzo, A.; Ragas, X.; Abbruzzetti, S.; Viappiani, C.; Campanini, B.; Flors, C.; Nonell, S. Singlet oxygen photosensitisation by GFP mutants: Oxygen accessibility to the chromophore. Photochem. Photobiol. Sci. 2010, 9, 1336–1341. [CrossRef] 71. Vegh, R.B.; Bravaya, K.B.; Bloch, D.A.; Bommarius, A.S.; Tolbert, L.M.; Verkhovsky, M.; Krylov, A.I.; Solntsev, K.M. Chromophore Photoreduction in Red Fluorescent Proteins Is Responsible for Bleaching and Phototoxicity. J. Phys. Chem. B 2014, 118, 4527–4534. [CrossRef]
Journal
Photonics – Multidisciplinary Digital Publishing Institute
Published: Jul 29, 2021
Keywords: protein collectivity; terahertz; thermal stability; fluorescent proteins; photobleaching