Fluorescence Dynamics of a FRET Probe Designed for Crowding

Article pubs.acs.org/JPCB

Fluorescence Dynamics of a FRET Probe Designed for Crowding Studies Megan Currie,† Hannah Leopold,† Jacob Schwarz,† Arnold J. Boersma,‡ Erin D. Sheets,*,† and Ahmed A. Heikal*,† †

Department of Chemistry and Biochemistry, Swenson College of Science and Engineering, University of Minnesota Duluth, Duluth, Minnesota 55812, United States ‡ Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute & Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands ABSTRACT: Living cells are crowded with macromolecules and organelles. As a result, there is an urgent need for molecular sensors for quantitative, site-specific assessment of the macromolecular crowding effects on a myriad of biochemical processes toward quantitative cell biology and biophysics. Here we investigate the excited-state dynamics and translational diffusion of a novel FRET sensor (mCerulean−linker−mCitrine) in a buffer (PBS, pH 7.4) at room temperature. Complementary experiments were carried out on free CFP, YFP, and the cleaved FRET probe as controls. The wavelength-dependent fluorescence lifetime measurements of the donor and acceptor in the FRET probe, using the timecorrelated single-photon counting technique, indicate an energy transfer efficiency of 6.8 ± 0.9% in PBS, with distinct excited-state dynamics from the recombinant CFP and YFP. The estimated mCeruleanmCitrine distance in this FRET probe is 7.7 ± 0.2 nm. The energy transfer efficiency increases (11.5 ± 0.9%) as the concentration of Ficoll-70 increases over the range of 0−300 g/L with an estimated mCerulean−mCitrine distance of 6.1 ± 0.2 nm. Complementary time-resolved anisotropy measurements suggest that the rotational diffusion of hetero-FRET in PBS is sensitive to the energy transfer from the donor to the acceptor. The results also suggest that the linker, −(GSG)6A(EAAAK)6A(GSG)6A(EAAAK)6A(GSG)6−, is rather flexible, and the observed rotational dynamics is likely to be due to a segmental mobility of the FRET pairs rather than an overall tumbling motion of a rigid probe. Comparative studies on a new construct of a FRET probe with a shorter, more flexible linker, mCerulean−(GSG)18−mCitrine, reveal enhanced energy transfer efficiency. On the millisecond time scale, fluorescence fluctuation analyses of the acceptor (excited at 488 nm) provide a means to examine the translational diffusion coefficient of the FRET probe. The results also suggest that the linker is flexible in this FRET probe, and the observed diffusion coefficient is faster than predicted as compared to the cleaved FRET probe. Our results serve as a point of reference for this FRET probe in a buffer toward its full potential as a sensor for macromolecular crowding in living cells and tissues.

I. INTRODUCTION Living cells are crowded by macromolecules such as proteins, DNA, microtubules, and a number of organelles.1,2 Such macromolecular crowding is believed to influence protein folding,3,4 diffusion and transport,5,6 and the kinetics of biochemical reactions.3,7−9 In addition, emerging evidence suggests a correlation between compartmentalized cellular crowding and cell pathophysiology and diseases.2 Yet, macromolecular crowding effects on cellular processes remain far from being fully understood. Part of the challenges toward such understanding is the multidimensional nature of crowding effects, which requires novel molecular sensors that are sensitive to crowding. In addition, different experimental approaches are needed to elucidate the extended nature of spatial and temporal scaling associated with crowding. Some of these methods used in crowding studies include NMR,10−12 fluorescence correlation spectroscopy,5,13,14 and time-resolved anisotropy measurements.15−18 © 2017 American Chemical Society

Fluorescence resonance energy transfer, or FRET, is a powerful tool19−23 for investigating conformational changes in biomolecules24 and intermolecular interactions.25−27 The energy transfer efficiency between a donor and acceptor (i.e., FRET pair) depends on (i) the spectral overlap between the donor’s emission and the acceptor’s absorption, (ii) the intermolecular donor−acceptor distance, and (iii) the relative orientation of the dipole moments of the FRET pair.28,29 FRET methods have been used successfully for intermolecular interactions in both solution and living cells.30 The energy transfer efficiency in FRET studies can be determined using steady-state spectroscopy of a solution in a cuvette,30 multichannel confocal microscopy,31 or total internal refraction fluorescence (TIRF) microscopy32 for intracellular investigaReceived: February 9, 2017 Revised: May 11, 2017 Published: May 18, 2017 5688

DOI: 10.1021/acs.jpcb.7b01306 J. Phys. Chem. B 2017, 121, 5688−5698

Article

The Journal of Physical Chemistry B tions. Fluorescence lifetime measurements have also been used to quantify FRET in both controlled environments30 and living cells.23,33 Recently, Boersma et al. have developed a novel, genetically encoded FRET sensor for macromolecular crowding.25 This sensor consists of mutated (A206 K) mCerulean (a donor) and mCitrine (an acceptor) that are separated by a flexible linker (−(GSG)6A(EAAAK)6A(GSG)6A(EAAAK)6A(GSG)6−). Importantly, this FRET probe (referred to as GE) can be genetically encoded into different cellular compartments toward site-specific crowding studies.25 The authors demonstrated the sensitivity of the FRET sensor to biomimetic crowding when expressed in living cells.25 The rationale is that the compact conformation of the FRET probe (i.e., smaller donor−acceptor distance and therefore enhanced energy transfer efficiency) will be favored in crowding environment due to (i) weak interactions between the crowding agents and the FRET sensor as well as (ii) steric hindrance due to the excluded volume of the crowding agents.25 Steady-state spectroscopy was used to quantify the energy transfer from the donor to the acceptor. However, it is not clear how the A206K mutations of mCerulean and mCitrine may influence the corresponding excited-state dynamics in the GE probe as compared with the parent CFP and YFP. The steady-state spectroscopy approach for determining the energy transfer efficiency in a FRET pair is known to suffer from complication due to spectral overlap as well as the sensitivity to the donor versus donor−acceptor concentrations.19−22 To overcome these challenges, a complementary fluorescence lifetime approach is needed to determine the energy transfer efficiency in the GE FRET probe. Finally, GFP mutants are known to undergo fluorescence blinking at the single molecule level, both in solutions34−39 and in living cells.40,41 Yet, it is not clear how the blinking and energy transfer may influence the fluorescence fluctuation analysis of the GE probe. In this contribution, we investigate the excited-state dynamics of a novel FRET probe using integrated fluorescence spectroscopy, at both the single-molecule and ensemble levels. The time-correlated single-photon counting (TCSPC) technique is used to examine the fluorescence lifetime of both the intact and cleaved GE sensor (Figure 1). As additional controls, similar measurements were carried out on recombinant CFP and YFP mutants in a buffer (PBS, pH 7.4) at room temperature using the same experimental approach to assess the conformational flexibility of the GE probe. Both fluorescence lifetime and anisotropy measurements are also used to determine the energy transfer efficiency of the GE construct as compared with mCerulean−(GSG)18−mCitrine (referred to as G18) to assess the effects of the linker length and flexibility on the FRET efficiency. To examine the crowding effects on the energy transfer efficiency, we conducted time-resolved fluorescence measurements of GE as a function of Ficoll-70 concentration (0−300 g/L) as compared with homogeneously viscous solution enriched with glycerol (0−760 g/L) as a control. Using fluorescence correlation spectroscopy (FCS), we also investigated the fluorescence fluctuation analysis to assess the translational diffusion in the presence of both fluorescent blinking and energy transfer in the GE probe at the single molecule level. Taken together, these fluorescence dynamics studies provide an essential point of reference for the GE probe toward its application in living cells using fluorescence lifetime imaging microscopy (FLIM).

Figure 1. Chemical structure and steady-state spectroscopy of a novel FRET probe. (A) A sketch describing the chemical structure of the FRET sensor, which consists of mutated CFP and YFP attached with a flexible linker (top). (B) The absorption spectrum of the GE sensor (black curve) as well as the emission spectra of both the donor (excited at 425 nm) and acceptor (excited at 500 nm) is also shown in a buffer (PBS, pH 7.4). The arrows indicate the excitation wavelengths used in the reported fluorescence lifetime and anisotropy measurements. The horizontal lines indicate the width of the bandwidth of the emission filters for the donor (blue curve) and acceptor (red curve) fluorescence detections. The absorption and emission bands are normalized for both the donor and acceptor.

II. MATERIALS AND METHODS Materials. The design of the genetically encoded FRETbased sensor, GE, was described in detail elsewhere.25 Briefly, mCerulean353 (cyan fluorescent protein) and mCitrine (yellow fluorescent protein) in this sensor serve as a donor and acceptor (i.e., FRET pair), respectively. As shown in Figure 1A, the mCerulean is located at the N terminus of a flexible linker (−(GSG) 6 A(EAAAK) 6 A(GSG) 6 A(EAAAK) 6 A(GSG) 6 −), while mCitrine is attached to the corresponding C terminus. αHelical peptides were included as part of the flexible linker in this FRET pair as compared with a random coil. To examine the linker-length effect on the FRET efficiency, we used a variant of the FRET sensor that had a more flexible, random coil linker (−(GSG)18−) as a comparison.42 In addition, mCerulean and mCitrine in this sensor were mutated (A206K) to minimize self-association.25 For purification, the E. coli strain BL21(DE)pLysS was transformed with the GE sensor plasmid, which was in a pRSET host vector, and grown to an OD600 of 0.6 in terrific broth supplemented with 0.4% (v/v) glycerol and 1 mg/mL 5689


Article

The Journal of Physical Chemistry B ampicillin at 30 °C.25 The cells were then incubated overnight with 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) at 25 °C to induce protein expression. Following centrifugation, the cells were resuspended in lysis buffer (10 mM sodium phosphate, pH 7.4, 100 mM sodium chloride, 0.1 mM phenylmethylsulfonyl fluoride, 1 mg/mL lysozyme) and then lysed with sonication. The lysate was again clarified (i.e., the supernatant was isolated) through centrifugation to which imidazole was added to a final concentration of 10 mM. The sensor was then purified using ProBond nickel-chelating resin (Life Technologies). The binding buffer was 10 mM sodium phosphate, pH 7.4, 100 mM sodium chloride, and 10 mM imidazole; the wash buffer was 50 mM sodium phosphate, pH 8.0, 300 mM sodium chloride, and 20 mM imidazole; and the elution buffer was 50 mM sodium phosphate, pH 8.0, 300 mM sodium chloride, and 250 mM imidazole. The purified fractions were then dialyzed against phosphate-buffered saline (pH 7.4). Fractions were stored at 4 °C and used within two weeks of preparation. The absorption and emission spectra were determined using a Beckman Coulter DU800 spectrophotometer and a Horiba Jobin Yvon Fluorolog, respectively. In Figure 1B, the absorption band at 280 nm (ε = 54 000 M−1 cm−1) was used to calculate the concentration of the isolated FRET sensor.25 While the donor-to-acceptor ratio is 1:1 in the GE sensor, the absorption band of the mutated CFP (454 nm) is 59.1% smaller than that of the corresponding acceptor (514 nm) due to differences in the extinction coefficient in the PBS buffer. The main emission peaks of the donor (475 nm) and acceptor (530 nm) are in agreement with previous studies.25 As a control for energy transfer, the flexible linker region of the GE sensor was cleaved using proteinase K, as described elsewhere.25 For this sensor cleavage, 0.56 ng of proteinase K (Sigma-Aldrich) was added per μmol of the purified GE sensor. After 1 min incubation at 25 °C, 20 μmol of PMSF per mg of proteinase K was added to terminate the cleavage reaction. The extent of cleavage was analyzed using SDS-PAGE and compared with intact protein. Recombinant CFP (31.3 kDa) and YFP (26.4 kDa) were purchased from BioVision and used without further purification as additional controls to assess the linker and mutations effects on the excited-state dynamics of the FRET probe. Comparison of the CFP and mCerulean sequences was conducted using a multiple sequence alignment, which reveals that the two sequences are approximately 76% similar. Notably, the amino acids that comprise the chromophore of CFP and mCerulean are AWG and SWG, respectively. A similar multiple sequence alignment of YFP and mCitrine indicates that the two sequences are only 52% similar to TYG and GYG fluorophore, respectively. These differences should be taken into consideration with regards to the excitedstate dynamics of the GE probe in comparison to free CFP and YFP. Rhodamine green 110 (Rh110; Invitrogen) and coumarin (ThermoFisher Scientific) were used as received in PBS and Tris buffers, respectively, as a reference for calibrating our experimental setup. For crowding studies, we used Ficoll-70 (Santa-Cruz Biotechnology) as a crowding agent in buffer at different concentrations (0−300 g/L). For homogeneous, viscous environment, however, we used glycerol (0−760 g/L) as a control. Time-Resolved Fluorescence and Anisotropy. Excitedstate dynamics of the GE sensor was carried out using timecorrelated single-photon counting (TCSPC) technique, and the

experimental setup was described in detail elsewhere.41,43 Briefly, femtosecond infrared laser pulses (850 or 930 nm) were generated using a titanium−sapphire solid-state laser system (Mira 900-F, Coherent). One-photon laser pulses (425 and 465 nm at 4.2 MHz) were generated and used for both fluorescence lifetime and anisotropy measurements on a droplet sample on a coverslip positioned on a 1.2NA 60× microscope objective. The filtered wavelength-dependent fluorescence was polarization-analyzed, detected by a microchannel plate (MCP) photomultiplier tube (R3809U, Hamamatsu), amplified, and routed to a synchronized SPC-830 module (Becker & Hickl) for data acquisition.44,45 For fluorescence lifetime measurements, a Glan-Thompson polarizer was used for magic-angle detection, and the acquired fluorescence decays were analyzed using the SPCImage software (Becker & Hickl), where the quality of the fit was judged using both χ2 and the residual.43,44 For anisotropy measurements, a polarizing beam splitter was used to isolate the parallel and perpendicular fluorescence polarizations (with respect to the laser polarization), which were detected simultaneously using two MCPs. Generally, the fluorescence intensity, F(t), of a given fluorophore can be described using a multiexponential decay model, depending on the chemical structure and the surrounding environment such that 3

F (t ) =

∑ αie−t /τ

i

(1)

i=1

where αi and τi are the amplitude fraction and fluorescence lifetime of the ith fluorophore. The measured fluorescence decay was deconvoluted with a computer-generated system response function. The measured fluorescence lifetime of the GE probe was used to calculate the energy transfer efficiency (E), which is dependent on the donor−acceptor distance (RDA) such that29 τ E = 1 − DA τD (2) where τD and τDA are the fluorescence lifetimes of the donor alone (cleaved probe) and the donor−acceptor in the FRET pair (intact probe), respectively. The Förster distance (R0) depends on the spectral overlap between the donor’s emission and acceptor’s absorption spectra as well as the orientation parameter between their dipole moments.29,31 Importantly, in our crowding experiments, we measured both τD and τDA under the same conditions, thereby eliminating changes in lifetime due to changes in refractive index. For time-resolved anisotropy, the measured parallel and perpendicular fluorescence decays were calculated using the anisotropy decay such that29 r (t ) =

I (t ) − GI⊥(t ) I (t ) + 2GI⊥(t )

=

∑ βi e−t/φ

i

i

(3)

The geometrical factor (G-factor) was also determined using the tail-matching approach.24,46 The factor of 2 in the denominator of eq 3 depends on the depolarization caused by high numerical aperture of the microscope objective.44 Because such depolarization does not affect the anisotropy decay time constant, we will use the traditional anisotropy decay equation (3), especially with our under-filled 1.2NA 60× microscope objective. The anisotropy decays were analyzed using OriginPro software. 5690


Article

The Journal of Physical Chemistry B Fluorescence Correlation Spectroscopy (FCS). For fluorescence fluctuation analysis, we used a home-built FCS as described elsewhere.43 Briefly, a fiber-coupled 488 nm laser (Coherent Sapphire 488-20) was steered toward an inverted microscope via the back exit port. A droplet of the fluorophore solution on a coverslip was excited using a 1.2NA microscope objective (60×), and the filtered fluorescence emission was focused on a 50 μm optical fiber that acted as a confocal pinhole. The fluorescence fluctuations were detected using an avalanche photodiode (APDs, SPCM CD-2969, PerkinElmer, Fremont, CA) followed by amplification and autocorrelation using external multiple-tau-digital correlator (ALV/6010-160, Langen/Hessen, Germany). The FCS setup was calibrated daily using Rh110 (PBS, pH 7.4) with a diffusion coefficient of 4.3 × 10−6 cm2/s.47 Rh110 was also used to assess the extent of the afterpulsing artifact associated with the avalanche photodiode. Based on the quality of autocorrelation curve fitting and knowledge of the triplet-state lifetime of Rh110, the starting point for Rh110 in a buffer that gave the best fit was used for all autocorrelation curves measured on the same day.48 The autocorrelation, G(τ), of fluorescence fluctuation of a single molecule diffusing through an open observation volume is49,50 G (τ ) =

⟨δF(t ) ⊗ δF(t + τ )⟩ ⟨F ⟩2

(4)

where δF(t) is the fluorescence fluctuation at time t and τ is the correlation lag time. For a Gaussian (lateral) and Lorentzian (axial) observation volume, the 3D autocorrelation function depends on both the average number of molecules (N) residing in the observation volume and the diffusion time (τD) such that41,43,49,51 G (τ ) =

1 (1 + τ /τD)−1(1 + τ /ω0 2τD)−1/2 N

(5)

where τD is the residence time of a molecule in the observation volume and ω0 is the structure parameter that describes the ratio of the axial (z0) to the lateral (ωxy) extension of the observation volume. In the presence of a photophysical process (jth) that causes an additional fluorescence fluctuation (e.g., triplet state or blinking), the corresponding autocorrelation will be modified accordingly.23 The autocorrelation curves were analyzed using OriginPro software. The measured diffusion time in a calibrated setup was then used to determine the translational diffusion coefficient (D) and the hydrodynamic radius, assuming the Stokes−Einstein model.41,43,49,51

Figure 2. One-photon excited-state dynamics of the GE probe is sensitive to both the excitation/detection wavelength and chemical structures as measured using the TCSPC technique. Time-resolved fluorescence on the GE sensor (curve 2; detected at magic angle) was carried out as a function of the excitation and detection wavelength such that we distinguish the excited-state dynamics of the donor and acceptor. Under 425 nm excitation, the fluorescence decays of both the donor (A: 475/50 nm) and acceptor (B: 530/40 nm) were recorded. In addition, both the donor and acceptor were excited at 465 nm, and corresponding fluorescence decays were also measured (C: 530/40 nm). Under the same conditions, the cleaved GE sensor (curve 3) as well as both the free CFP (curve 1, panel A) and YFP (curve 1, panels B and C) were measured as controls. See Table 1 for a summary of the fitting parameters. These fluorescence decays are shown on a logarithmic scale.

III. RESULTS AND DISCUSSION Wavelength Dependence of the Excited-State Dynamics of the GE Probe. To examine the effects of both the mutation and linker on the excited-state dynamics of the parent CFP and YFP in the GE probe, we carried out time-resolved fluorescence measurements using the TCSPC technique. These measurements will also provide a means to assess the energy transfer efficiency of the GE probe. Figure 2 shows typical fluorescence decays of both the intact and cleaved GE probe as a function of excitation and detection wavelengths. The donor (mCerulean) emission decays as a biexponential in the intact and cleaved GE probe, which is similar to that of the free CFP. However, the fitting parameters are distinct with an average lifetime of 2.82 ns (free CFP), 3.72 ns (GE probe), and 3.99 ns (cleaved GE probe) as shown in Table 1. There is a slight difference in the fluorescence lifetime

of the intact probe as compared with the cleaved counterpart, which we attribute to energy transfer. These results indicate that the mutated and tethered mCerulean in the GE probe is distinct from the free CFP. In addition, the average fluorescence lifetime of the donor in the intact GE probe undergoes energy transfer to the tethered acceptor (mCitrine) at an estimated efficiency of 6.8% in a buffer at room temperature. These calculations are based on the fluorescence lifetime of the intact (τDA) and cleaved (τD) GE sensor. In contrast, the fast decay component of the donor in the intact GE probe, with respect to the cleaved GE (i.e., donor alone), 5691


Article

The Journal of Physical Chemistry B

Table 1. A Summary of the Fitting Parameters of Time-Resolved Fluorescence of the FRET Probe, Cleaved Probe, Free CFP, and Free YFP As Measured Using the TCSPC Techniquea λx−λfl (nm): molecule 425−475/50: coumarin free CFP intact GE cleaved GE 425−530/40: Rh110 free YFP intact GE cleaved GE 465−530/40:c Rh 110 free YFP GE probe cleaved GE

α1

0.71

0.33

τ1 (ns)

0.28

0.67

α2

τ1 (ns)

α3

τ3b (ns)

⟨τfl⟩ (ns)

0.64 0.70

2.36 3.27

0.36 0.30 1.00

3.64 4.74 3.99

2.82 3.72 3.99

0.20

1.00

0.75

3.63

0.09 1.00 0.25

2.99 3.91 4.79

0.67 3.91 3.92

0.51

3.22

0.72

3.26

0.16 1.00 0.28

3.94 3.74 4.52

2.48 3.74 3.62

a

These measurements were carried out in a buffer (PBS, pH 7.4) at room temperature as a function of excitation and detection wavelengths. The laser repetition rate was 4.2 MHz. bIn multiexponential fluorescence decays, we assigned the longest fluorescence lifetime component as τ3 for easy comparison. cWhen exciting the GE probe at 465 nm, there is a spectral overlap of the absorption and emission of both the donor (CFP-like) and acceptor (YFP-like) as indicated in Figure 1B.

lifetime of 3.98 ns at room temperature in agreement with previous reports.43 Rotational Dynamics of the GE Probe. To assess the conformational flexibility of the GE probe, we carried out timeresolved fluorescence anisotropy of the intact and cleaved GE probe as a function of excitation/detection wavelengths in a buffer at room temperature. Our experimental setup was calibrated using the anisotropy of Rh110 (rotational time of ∼140 ps) and coumarin (rotational time of ∼98 ps) to assess the wavelength dependence of the G-factor. Representative time-resolved anisotropy of the intact and cleaved GE probe is shown in Figure 3. The anisotropy decays of free CFP and YFP (PBS, pH 7.4) are also shown as a function of excitation/ detection wavelengths at room temperature. Under 425 nm illumination, the polarization-analyzed fluorescence (475/50 nm) of the donor (mCerulean) was detected, and the corresponding anisotropy of the intact GE probe decays as a single exponential with a rotational time of 18.25 ns (Figure 3A and Table 2). The anisotropy of the donor in the cleaved GE probe sample also decays as a single exponential with a faster rotational time (14.57 ns). In contrast, the anisotropy of the free, recombinant CFP decays as a biexponential with an average rotational time of 22.17 ns (Table 2). The observed rotational time of the donor in the GE probe is too fast to be assigned to the overall rotation of the intact probe (∼64 kDa). As a result, we assign the observed rotational time to segmental mobility of the donor in the FRET probe due to a flexible linker. Using the Stokes−Einstein model, the rotating moiety of the recombinant CFP has an estimated hydrodynamic radius of 2.90 nm, which corresponds to a spherical volume of 102 nm3. In contrast, the calculated hydrodynamic radius of the intact GE probe (∼64 kDa) is 2.72 nm, which corresponds to 84.3 nm3 in Stokes−Einstein model. Using the Perrin equation, the hydrodynamic volume of recombinant CFP (49.9 nm3) and GE probe (102 nm3) was calculated based on their molecular weight. The corresponding hydrodynamic radii for a spherical shaped CFP and GE are 2.28 and 2.9 nm, respectively. The faster rotational time of the cleaved probe suggest a smaller moiety than both the intact

yields an estimated energy transfer efficiency of 18.0%. In addition, the absorption and emission spectra of both the donor and acceptor of the GE probe suggest a Förster distance of 4.99 nm for this new FRET pair. Taken together, we estimate a mCerulean−mCitrine distance of 7.7 ± 0.2 nm. Under 425 nm excitation of the donor in the GE probe, the free YFP emission (530/40 nm) decays as a triple exponential with a dominant fast decay component (Table 1), which is also distinct from the mutated acceptor (mCitrine) of the GE probe. The fast component of the free YFP under 425 nm illumination is attributed to a protonated electronic state as in the parent wild-type GFP.35,37,40,52 These results indicate a significant effect of the A206K mutation and/or the linker on the excitedstate dynamics of the donor−acceptor pair (mCerulean− linker−mCitrine) in this newly designed FRET probe. The amplitude of the fast decay component of the free YFP is significantly reduced under 465 nm excitation, which is within the main absorption band on the deprotonated (bright) state of the GFP mutants. Under 465 nm excitation of the GE probe, the fluorescence (530/40 nm) decays as a single exponential with a decay constant of 3.74 ns (Table 1), which is slightly longer than previously reported lifetime (3.61 ± 0.03 ns) for free Citrine (pH 9.0) under 490 nm excitation.35 The observed slightly longer lifetime of the acceptor (mCitrine) in this FRET probe might be attributed to the spectral overlap (Figure 1B) with the donor (mCerulean), which has a longer lifetime. Because the fluorescence lifetime of free Citrine decreases at low pH,35 the pH sensitivity can be ruled out here at pH 7.4 for the mCitrine in the GE probe. Under the same conditions, the fluorescence of the cleaved GE decays as a biexponential with an estimated average lifetime of 3.62 ns (Table 1). As an additional control, the fluorescence of free YFP (excited at 465 nm and detected at 530/40 nm) decays as a triple exponential with an average lifetime of 2.48 ns (Table 1). Our fluorescence lifetimes of free CFP and YFP are in general agreement with related literature.53,54 Under the same conditions, the fluorescence decay of Rh110 (PBS, pH 7.4) indicates an excited-state 5692


Article


Under 425 nm illumination, the polarization-analyzed fluorescence (530/40 nm) of the FRET probe was detected, and the corresponding anisotropy of the intact GE probe decays as a biexponential with an overall rotational time of 19.6 ns (β2 = 0.066) along with a faster rotational component (φ1 = 2.10 ns, β1 = 0.209). The fast rotational time is too fast to be assigned as a tumbling motion of ∼64-kDa size protein (Figure 3B and Table 2). Such a fast component is absent from the cleaved GE probe. As a result, we attribute the faster rotational time of the GE probe under 425 nm illumination to energy transfer from the donor to the acceptor. Once again, the slow rotational time of the intact GE probe under these conditions is relatively slower than that of the cleaved counterpart (Table 1). It is worth mentioning that a direct excitation of the free YFP was relatively weak and the biexponential anisotropy decay with a low signal-to-noise (S/N) ratio as shown in Figure 3B. Under 465 nm excitation and 530/40 nm detection, the anisotropy of both the intact and cleaved GE probe decays as a biexponential (Figure 3C), and the fitting parameters are summarized in (Table 2). This might not be surprising considering the fact that both the donor and acceptor of the intact GE probe can be excited under 465 nm illumination with some spectral overlap of their emission (see Figure 1B). It is worth noting that the observed rotational time of the acceptor (mCitrine) of the cleaved GE probe seems in agreement with that of the free Citrine in a buffer (pH 9.0) at room temperature.35 The initial anisotropy of all the molecules used in this report seems to be smaller than the theoretical value (r0 = 0.4) as shown in Table 2. Following the conventional interpretation of such deviation, we attribute this difference to the presence of ultrafast processes that compete with excited-state relaxation via fluorescence pathways.36 We then used these initial anisotropy values to calculate the angle between the absorbing and emitting dipoles of the intact and cleaved GE probe (Table 2). In addition, the average rotational times of the GE probe were used to calculate the hydrodynamic volume of the tumbling moiety as well as the corresponding radius using Stokes− Einstein model for a spherical rotor. Using the Stokes−Einstein model for rotational diffusion of these molecular systems, we calculated the projected rotational times (φ) and the hydrodynamic radii. For the GE probe (64 kDa), we calculated a rotational time of 22.1 ns in buffer at room temperature and a hydrodynamic volume of 102 nm3 (or 2.90 nm radius). Assuming a monomeric recombinant CFP (31.3 kDa), we calculated the projected rotational time (10.8 ns), hydrodynamic volume (49.9 nm3), and radius (2.28 nm). These projected values are slightly different from those of recombinant YFP (26.4 kDa), where rotational time (9.11 ns), hydrodynamic volume (42.1 nm3), and radius (2.16 nm) were calculated in a buffer at room temperature. On the basis of our time-resolved anisotropy results, we conclude that the linker in the GE probe is rather long and flexible because the rotational time of either the donor or acceptor in this probe seems too fast for the overall tumbling motion of 64 kDa probe. This conclusion seems to be consistent with the relatively low energy transfer efficiency based on fluorescence lifetime measurements. Warren et al.55 have shown analytically how the energy transfer rate constant in homo-FRET of an oligomer of N fluorophores could be estimated using biexponential anisotropy decay. Using the same analytical approach, we used the observed biexponential anisotropy decays of our GE probe

Figure 3. One-photon rotational dynamics of the GE probe is sensitive to both the excitation-detection wavelength and chemical structures as demonstrated using time-resolved anisotropy. Under 425 nm excitation, the anisotropy decays of both the donor (A: 475/50 nm) and acceptor (B: 530/40 nm) were recorded. In addition, both the donor and acceptor were excited at 465 nm and corresponding anisotropy decays of both intact (curve 2, panels A−C) and cleaved (curve 3, panels A−C) GE probe were also measured (C: 530/40 nm). Under the same conditions, the free CFP (curve 4, panel A) and YFP (curve 4, panel C) were measured as controls, along with coumarin (curve 1) (425 nm excitation) and Rh110 (curve 1) (465 nm excitation). See Table 2 for a summary of the fitting parameters. These fluorescence decays are shown on a logarithmic scale.

counterpart and the free CFP in a buffer at room temperature. In addition, the free CFP seems to rotate slower than that of the cleaved donor in the GE probe, which is counterintuitive if we assumed that the cleavage site is in the middle of the peptide linker. In light of these findings, it is possible recombinant CFP might exist as a dimer in buffer (pH 7.4). 5693


Article


Table 2. Fitting Parameters of Time-Resolved Anisotropy of the GE Probe, Cleaved GE Probe, Free CFP, and Free YFP As Measured Using the TCSPC Technique with Emission-Polarization Analysisa λx−λfl (nm): molecule 425−475/50: coumarin free CFP intact GE cleaved GE 425−530/40: Rh110 free CFP free YFP intact GE cleaved GE 465−530/40: Rh110 free YFP intact GE cleaved GE

β1

φ1 (ns)

β2

φ2 (ns)

r0

θb (deg)

Vrotc (nm3)

Rrot (nm)

0.398 0.035

0.098 1.60

0.248 0.276 0.280

22.17 18.25 14.57

0.398 0.283 0.276 0.280

40.1 46.5 46.8 46.6

0.454 102 84.3 67.3

0.475 2.90 2.72 2.52

0.284

0.18

0.284

46.5

0.832

0.583

0.058 0.066 0.011

0.854 2.104 0.886

0.247 0.209 0.284

29.33 19.61 15.02

0.305 0.275 0.295

45.3 46.9 45.8

136 90.6 68.4

3.19 2.79 2.55

0.374 0.022 0.054 0.018

0.171 0.677 1.804 1.99

0.259 0.242 0.291

31.16 19.69 15.86

0.374 0.281 0.296 0.309

41.5 46.6 45.8 45.1

0.573 144 90.9 73.3

0.573 3.25 2.79 2.60

a

These measurements were carried out in a buffer (PBS, pH 7.4) at room temperature as a function of excitation and detection wavelengths. The laser repetition rate was 4.2 MHz. bThe angle (θ) is the angle between the absorbing and emitting dipole moments of a given fluorophore.58 cIn case of multiexponential anisotropy decay, the hydrodynamic volume (Vrot) of the rotating moiety of a given fluorophore was calculated using the corresponding rotational time of the slowest anisotropy decay component. We also assumed a water viscosity of 0.89 cP at room temperature (298 K).

Figure 4 shows representative fluorescence and anisotropy decays of G18 and the parent GE probe with −(GSG)6A(EAAAK)6A(GSG)6A(EAAAK)6A(GSG)6− linker. The fluorescence of G18 decays (Figure 4A) indicates an excited-state lifetime of 3.02 ns with an estimated energy transfer efficiency of 22%, which is significantly larger than that with the longer linker (GE). The enhanced energy-transfer efficiency, which is based on the fluorescence lifetime of the intact and cleaved G18 probe, is attributed to the shorter, flexible linker. Interestingly, the corresponding time-resolved anisotropy measurements also exhibit distinct rotational dynamics as compared with the parent GE sensor (Figure 4B). The fast rotational component of the G18 probe provides an upper estimate of 45% for the energy transfer efficiency due to the shorter linker. Using the Stokes−Einstein model, we estimated the hydrodynamic volume of 84 nm3 for G18, which corresponds to a radius of 2.71 nm. Based on the molecular weight for G18 (58 kDa), the Perrin equation yields a hydrodynamic volume of 93 nm3, which corresponds to a radius of 2.8 nm for a spherically shaped probe. The enhanced energy transfer efficiency of G18, using lifetime and anisotropy measurements, seems in a general agreement with steady-state-based studies.42 Crowding Effects of the Excited-State Dynamics of the GE Probe. We also investigated the effect of homogeneous (glycerol-enriched buffer) and heterogeneous (Ficoll-70enriched buffer) viscosity on the energy transfer efficiency of GE using time-resolved fluorescence (see Figure 5 for representative results). In Ficoll-crowded solution, the energy transfer efficiency increases with the concentration of the crowding agent over the range of 0−300 g/L. In contrast, slight reduction of the energy transfer efficiency of GE was observed as the glycerol concentration increases over the range 0−760 g/ L. The estimated bulk viscosity over the concentration range used in these measurements is 1−18 cP and 1−28 cP for glycerol and Ficoll-70 solutions, respectively. Our results show that the homogeneously viscous environment (e.g., glycerolenriched buffer) reduces conformational fluctuations of the

(Figure 3B and Table 2) for an approximate estimate of the energy transfer rate (kET) using the equation55 φ1−1 = (φ2−1 + NkET)

(6)

Assuming N = 2 for our probe (mCerulean−linker−mCitrine), we estimate an energy transfer rate of 0.23 ± 0.02 ns−1. In these calculations, we used the fast (φ1) and slow (φ2) rotational times in the biexponential anisotropy decays (Table 2). Assuming that the fluorescence lifetime of the donor alone is 3.99 ns, the anisotropy-based energy transfer rate yields ∼47% energy transfer efficiency according to the equation29 E ET =

kET kET + (τfl D)−1

(7)

The calculated 47% energy transfer efficiency using this approach is too high as compared with that determined using fluorescence lifetime measurements, which is 6.8%−18.0% (see above). To reconcile such difference, we must emphasize that the slow rotational time constant here does not represent the overall rotational time of the GE probe as mentioned above in departure from the assumption made by Warren et al.55 The disagreement between the estimated energy transfer efficiency using fluorescence lifetime and anisotropy approaches attests to the fact that the homo-FRET model developed by Warren et al.55 might not be directly applicable to hetero-FRET. Finally, the weighted above energy transfer efficiency (47%) by the amplitude fraction of the fast rotational time of the GE probe (0.27 ± 0.05) would yield an energy transfer efficiency of 12 ± 3%, which is in reasonable agreement with both our fluorescence lifetime-based approach (above) and the steadystate approximation approach by Boersma et al.25 Linker Effect on Excited-State Dynamics of the G18 Probe. To examine the linker effects on the excited-state dynamics and energy transfer efficiency, we carried out complementary time-resolved fluorescence and anisotropy measurements on a newly developed FRET probe (G18), namely, mCerulean(GSG)18mCitrine.42 5694


Article


Figure 5. Energy transfer efficiency of GE probe is sensitive to the environment. (A) The energy transfer efficiency of GE increases as the concentration of Ficoll-70 increases over the range of 0−300 g/L at room temperature. (B) In glycerol-enriched buffer, however, the energy transfer efficiency slightly decreases with the increase of the glycerol concentration (0−760 g/L). These results were recorded under the excitation (425 nm) and detection (475/50 nm) of the donor in GE probe. The fluorescence lifetime of the cleaved GE was used as the fluorescence lifetime of the donor alone in our calculations of the energy transfer efficiency. The dotted line indicates the data trend for guidance and not a fit.

Figure 4. Linker effects on the excited-state dynamics of novel FRET probes. Time-resolved fluorescence (A) and anisotropy (B) are shown for both the parent GE probe and a newly engineered derivative G18, mCerulean−(GSG)18−mCitrine, with a shorter linker. Under 425 nm excitation and 475/50 nm detection, the G18 probe with a shorter linker exhibits a significant reduction of the fluorescence lifetime (A), which is attributed to efficient energy transfer between donor and acceptor. In addition, the anisotropy decay of the G18 probe and enhanced energy transfer efficiency decays as biexponential anisotropy with enhanced fast rotational component. More details will be published elsewhere.

fluctuation analysis. These studies should also help understand the translational diffusion mechanism associated with the GE probe, designed for intracellular crowding studies. Together with the rotational diffusion, FCS studies will help us address the spatial and temporal scaling of the diffusion processes of this crowding probe. Figure 6 shows typical fluorescence fluctuation autocorrelation curves of the intact and cleaved GE probe. Under 488 nm illumination, the free YFP autocorrelation curve was also measured and is shown in Figure 6. The experimental setup was calibrated using Rh110 (PBS, pH 7.4) with a diffusion coefficient of 4.3 × 10−6 cm2/s.47 Accordingly, our observation volume in FCS has a lateral extension of ∼332 nm, which is close to the diffraction limit. The fitting parameters of these autocorrelation curves are summarized in Table 3. Our results indicate that the intact GE probe, under 488 nm illumination, diffuses slower (0.44 ms) than both the cleaved counterpart (0.23 ms) and the free recombinant YFP (0.30 ms) in a buffer at room temperature (Figure 6 and Table 3). On the basis of these autocorrelation curves, we estimated a diffusion coefficient of the GE probe to be (6.25 ± 0.9) × 10−7 cm2/s as compared with 1.19 × 10−6 cm2/s for the cleaved GE counterpart and 9.11 × 10−7 cm2/s for free YFP (Table 3). In agreement with our time-resolved anisotropy and the rotational times, the observed faster diffusion coefficient of the cleaved GE probe, as compared with the free recombinant YFP,

tethered donor−acceptor pair and therefore reduces or stabilizes the FRET efficiency. In contrast, the excluded volume and confinement in crowded, heterogeneous environments (Ficoll-70-enriched buffer) favors smaller donor−acceptor distance and therefore enhances FRET efficiency. Using comparative studies on the cleaved counterpart of GE allowed us to account for the effect of refractive index changes in these complex environments on the fluorescence lifetime of these FRET probes. In addition, the fluorescence lifetime of the cleaved GE was used as the fluorescence lifetime of the donor alone in our calculations of the corresponding energy transfer efficiency shown in Figure 5. The general trends observed here using fluorescence lifetime to calculate energy transfer efficiency in crowded and glycerol environments are in general agreement with steady-state spectroscopy.25 Fluorescence Fluctuation Analysis of the FRET Probe. To examine whether the diffusing moiety of the GE probe is the same on the picosecond−nanosecond (rotation) and microsecond−millisecond (translation) time scales, we carried out fluorescence fluctuation correlation analysis using FCS. In addition, these measurements would elucidate how the energy transfer in the GE probe might interfere with the known fluorescence blinking in a wide range of GFP mutants.34−37,40,41,56 It is conceivable that both energy transfer and blinking would contribute to the overall fluorescence 5695


Article


Assuming a hydration of 0.23 g of H2O per gram of protein, we used the molecular weight of both the free YFP and GE probe to calculate the corresponding hydrodynamic radius. For free YFP (26.4 kDa), we calculated a hydrodynamic radius of 2.15 nm as compared with 2.90 nm for the intact GE probe. On the basis of these calculations and the Stokes−Einstein model, one would expect a translational diffusion coefficient of 1.1 × 10−6 and 8.5 × 10−7 cm2/s for free YFP and intact GE probe. The estimated diffusion coefficient of recombinant YFP (26.4 kDa) might suggest dimer formation, even at the nanomolar concentration used in FCS, as previously proposed in a buffer using ultrafast spectroscopy.57 The autocorrelation curves measured here suggest that fluorescence blinking prevails in the GE probe, both cleaved and intact, as well as the free YFP (Figure 6 and Table 3). The population fraction of the blinking YFP (14%) is in agreement with previous studies.34,35,38−40,56 In addition, the blinking time constant of the intact GE probe (470 μs) seems significantly slower than that of the free YFP (120 μs).

Figure 6. Fluorescence fluctuation analysis of the GE probe, at the single-molecule level, is sensitive to chemical structures as demonstrated using fluorescence correlation spectroscopy. Normalized autocorrelation curves of Rh110 (curve 1), free YFP (curve 4), and both the intact (curve 3) and cleaved (curve 2) GE sensor are shown in a buffer (PBS, pH 7.4) at room temperature. These autocorrelation curves were recorded using 488 nm excitation, 1.2NA objective (60×), 525/30 nm emission filter, and a 50 μm optical fiber as a confocal pinhole. See Table 3 for fitting parameters of these correlation curves.

IV. CONCLUSIONS We investigated the excited-state dynamics and translational diffusion of a FRET sensor, mCeruleanlinkermCitrine, in a buffer (PBS, pH 7.4) at room temperature using integrated fluorescence spectroscopy methods. The wavelength-dependence of the excited-state dynamics of the donor and acceptor of the GE probe seems different from the free CFP and YFP. These fluorescence lifetime measurements also indicate an energy transfer efficiency of 6.8 ± 0.9% with an estimated donor−acceptor distance of 7.7 ± 0.2 nm. In Ficoll-70-induced crowding (300 g/L), the energy transfer efficiency of GE increased up to 11.5 ± 0.9% with an estimated donor−acceptor distance of 6.1 ± 0.2 nm in such heterogeneous environment. In contrast, a slight decrease in the energy transfer efficiency of GE in glycerol-enriched buffer was observed and attributed to reduced conformational fluctuation of the donor−acceptor pair in viscous solutions. Such modest energy transfer efficiency can be attributed to the extended and flexible linker, namely −(GSG)6A(EAAAK)6A(GSG)6A(EAAAK)6A(GSG)6−, which links the donor (mCerulean) with the acceptor (mCitrine). Such structural attributes are supported by the rotational and translational diffusion using time-resolved anisotropy and fluorescence fluctuation analysis, respectively. This becomes rather significant as these studies move forward into investigating the effects of macromolecular crowding using the rotational and translational diffusion of these FRET probes.

seems counterintuitive (Figure 6 and Table 3) and might suggest dimerization of YFP as previously reported. 57 Currently, to the best of our knowledge, there is no theoretical model that describes the translational diffusion of two proteins attached to each other with a flexible linker. Therefore, we used our experimentally measured diffusion coefficients to calculate the hydrodynamic radius of our probe, assuming the Stokes− Einstein model. On the time scale of our observations of fluorescence fluctuation due to translational diffusion, our results suggest that the two FRET pairs (donor and acceptor) are diffusing fairly independently, which further justifies our use of this model. Based on our calculations, the GE probe yields a hydrodynamic radius of 3.92 nm as compared with 2.06 nm for the cleaved GE counterpart. The estimated translational diffusion and hydrodynamic radius of the GE probe seem in agreement of the free Citrine in a buffer (pH 9.0) using FCS.35 This indicates that the GE probe has a flexible linker, and the diffusion is limited by the displacement of the solvent by the acceptor rather than an intact, rigid, spherically shaped probe. Such interpretation is consistent with our time-resolved anisotropy measurements on an ensemble of the GE probe at the nanosecond time scale.

Table 3. Fitting Parameters of Fluorescence Fluctuation Autocorrelation of the GE Probe, Cleaved GE Probe, Free CFP, and Free YFP as Measured Using FCSa λx−λfl (nm): molecule

N

τD (ms)

f jb

τj (ms)

488−530/40: Rh110 free YFP GE probe cleaved GE

9.3 5.2 13.9 3.9

0.064 0.30 0.44 0.23

0.15 0.14 0.33 0.34

0.016 0.012 0.049 0.044

DTc (cm2/s) 4.3 9.11 6.25 1.19

× × × ×

10−6 10−7 10−7 10−6

aFCSd (nm) 0.57 2.69 3.92 2.06

a

These measurements were carried out in a buffer (PBS, pH 7.4) at room temperature using a 488 nm laser, 1.2NA 60× objective (water immersion), and 50 μM confocal pinhole. bThe fraction and corresponding time constant of the jth photophysical process depends on the fluorophore. In the case of rhodamine, such jth photophysical process is attributed to a triplet state as compared with blinking in CFP, YFP, and GE probe (intact and cleaved).48 cThe translational diffusion coefficient (D) was calculated using the diffusion time (τD) measured using our calibrated FCS setup.43 dThe hydrodynamic radius was calculated for a given fluorophore assuming a spherical shape as in Stokes−Einstein model, T = 298 K, and a viscosity of 0.89 cP. 5696


Article


0718741), and the National Institutes of Health (AG030949). Additionally, the financial support was provided by a Chancellor’s Small Grant, the Department of Chemistry and Biochemistry, the Swenson College of Science and Engineering, all of which are from the University of Minnesota Duluth. M.C., J.S., and H.L. were supported by teaching fellowships from the University of Minnesota Duluth Department of Chemistry and Biochemistry.

On the nanosecond time scale, time-resolved anisotropy also indicated a fast rotational decay component that we attributed to energy transfer in the intact GE probe. Importantly, these fast rotational dynamics indicate that both the donor and acceptor in the intact GE probe undergo segmental mobility that is faster that the overall rotation of a 64 kDa probe. Both the efficiency and rate constant of energy transfer of this FRET probe determined by the lifetime measurements seem in a general agreement with that determined by the anisotropy measurements. Our preliminary excited-state dynamics results on a new FRET probe with a shorter linker, G18, indicate enhanced energy transfer efficiency as compared with the parent GE counterpart. As a result, G18 has a greater potential as a sensor for FRET studies in crowded environments using FLIM and anisotropy measurements that are best suited for live cell studies. In addition, our comparative linker-dependent measurements would provide a model system for developing a theory for hetero-FRET analysis using time-resolved anisotropy measurements. At the single-molecule level, fluorescence fluctuation analysis using FCS yield an overall translational diffusion coefficient of the GE probe, 6.25 × 10−7 cm2/s, which seems to be comparable to that of the cleaved counterpart or free recombinant YFP (i.e., potential dimerization) in a buffer (PBS, pH 7.4) at room temperature. These single-molecule studies seem to support our ensemble findings on the nanosecond time scale concerning the flexible, extended structure of the GE probe, implying that the donor in the GE probe diffuses almost freely without much restriction from the acceptor-linker attachment. In addition to translational diffusion, energy transfer in the GE probe provides an additional mechanism for enhancing fluorescence blinking, which is known to occur in a wide range of GFP mutants. Taken together, our results provide complementary information for this genetically encoded FRET probe to make fluorescence lifetime and polarization imaging microscopy accessible for our ultimate goal of site-specific crowding studies in living cells. Currently, we are investigating how macromolecular crowding might influence the multiscale molecular dynamics of this and other derivatives of FRET probe.

■

■

REFERENCES

(1) Ellis, R. J. Macromolecular Crowding: Obvious but Underappreciated. Trends Biochem. Sci. 2001, 26, 597−604. (2) Ellis, R. J.; Minton, A. P. Cell Biology: Join the Crowd. Nature 2003, 425, 27−28. (3) Minton, A. P. Influence of Macromolecular Crowding Upon the Stability and State of Association of Proteins: Predictions and Observations. J. Pharm. Sci. 2005, 94, 1668−1675. (4) Tokuriki, N.; Kinjo, M.; Negi, S.; Hoshino, M.; Goto, Y.; Urabe, I.; Yomo, T. Protein Folding by the Effects of Macromolecular Crowding. Protein Sci. 2004, 13, 125−133. (5) Dauty, E.; Verkman, A. S. Molecular Crowding Reduces to a Similar Extent the Diffusion of Small Solutes and Macromolecules: Measurement by Fluorescence Correlation Spectroscopy. J. Mol. Recognit. 2004, 17, 441−7. (6) Dix, J. A.; Verkman, A. S. Crowding Effects on Diffusion in Solutions and Cells. Annu. Rev. Biophys. 2008, 37, 247−63. (7) Minton, A. P. Confinement as a Determinant of Macromolecular Structure and Reactivity. Biophys. J. 1992, 63, 1090−1100. (8) Minton, A. P. The Influence of Macromolecular Crowding and Macromolecular Confinement on Biochemical Reactions in Physiological Media. J. Biol. Chem. 2001, 276, 10577−80. (9) Minton, A. P.; Zhou, X. H.; Rivas, G. Macromolecular Crowding and Confinement: Biochemical, Biophysical, and Potential Physiological Consequences. Annu. Rev. Biophys. 2008, 37, 375−397. (10) Wang, Y.; Benton, L. A.; Singh, V.; Pielak, G. J. Disordered Protein Diffusion under Crowded Conditions. J. Phys. Chem. Lett. 2012, 3, 2703−2706. (11) Wang, Y.; Li, C.; Pielak, G. J. Effects of Proteins on Protein Diffusion. J. Am. Chem. Soc. 2010, 132, 9392−7. (12) Wang, Y.; Sarkar, M.; Smith, A. E.; Krois, A. S.; Pielak, G. J. Macromolecular Crowding and Protein Stability. J. Am. Chem. Soc. 2012, 134, 16614−16618. (13) Goins, A. B.; Sanabria, H.; Waxham, M. N. Macromolecular Crowding and Size Effects on Probe Microviscosity. Biophys. J. 2008, 95, 5362−73. (14) Neuweiler, H.; Löllmann, M.; Doose, S.; Sauer, M. Dynamics of Unfolded Polypeptide Chains in Crowded Environment Studied by Fluorescence Correlation Spectroscopy. J. Mol. Biol. 2007, 365, 856− 869. (15) Lavalette, D.; Tetreau, C.; Tourbez, M.; Blouquit, Y. Microscopic Viscosity and Rotational Diffusion of Proteins in a Macromolecular Environment. Biophys. J. 1999, 76, 2744−51. (16) Zorrilla, S.; Hink, M. A.; Visser, A. J. W. G.; Lillo, M. P. Translational and Rotational Motions of Proteins in a Protein Crowded Environment. Biophys. Chem. 2007, 125, 298−305. (17) Zorrilla, S.; Rivas, G.; Acuna, A. U.; Lillo, M. P. Protein SelfAssociation in Crowded Protein Solutions: A Time-Resolved Fluorescence Polarization Study. Protein Sci. 2004, 13, 2960−9. (18) Zorrilla, S.; Rivas, G.; Lillo, M. P. Fluorescence Anisotropy as a Probe to Study Tracer Proteins in Crowded Solutions. J. Mol. Recognit. 2004, 17, 408−16. (19) Müller, S. M.; Galliardt, H.; Schneider, J.; Barisas, B. G.; Seidel, T. Quantification of Forster Resonance Energy Transfer by Monitoring Sensitized Emission in Living Plants Cells. Front. Plant Sci. 2013, 4, 1−19. (20) Majoul, I.; Straub, M.; Duden, R.; Hell, S. W.; Soling, H.-D. Fluorescence Resonance Energy Transfer Analysis of Protein - Protein

AUTHOR INFORMATION

Corresponding Authors

*(A.A.H.) E-mail [email protected], Ph 218-726-7036. *(E.D.S.) E-mail [email protected], Ph 218-726-6046. ORCID

Arnold J. Boersma: 0000-0002-3714-5938 Ahmed A. Heikal: 0000-0002-0370-4033 Author Contributions

M.C., H.L., and J.S. contributed equally to this project. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Ben Orpen, Claire Baetzold, Sheila Paintsil, Kaelt Simpson, and Nathan Korson for their technical help during the course of these studies. A.J.B. acknowledges the financial support of The Netherlands Organization for Scientific Research Vidi grant. E.D.S. and A.A.H. also acknowledge the financial support provided by the University of Minnesota Grant-in-Aid, the National Science Foundation (MCB 5697


Article

The Journal of Physical Chemistry B Interactions in Single Living Cells by Multifocal Multiphoton Microscopy. Rev. Mol. Biotechnol. 2002, 82, 267−277. (21) Jares-Erijman, E. A.; Jovin, T. M. FRET Imaging. Nat. Biotechnol. 2003, 21, 1387−1395. (22) Day, R. N.; Tao, W.; Dunn, K. W. A Simple Approach for Measuring FRET in Fluorescent Biosensors Using Two-Photon Microscopy. Nat. Protoc. 2016, 11, 2066. (23) Koushik, S. V.; Vogel, S. S. Energy Migration Alters the Fluorescence Lifetime of Cerulean: Implications for Fluorescence Lifetime Imaging Forster Resonance Energy Transfer Measurements. J. Biomed. Opt. 2008, 13, 031204−031204−9. (24) Davey, A. M.; Krise, K. M.; Sheets, E. D.; Heikal, A. A. Molecular Perspective of Antigen-Mediated Mast Cell Signaling. J. Biol. Chem. 2008, 283, 7117−7127. (25) Boersma, A. J.; Zuhorn, I. S.; Poolman, B. A Sensor for Quantification of Macromolecular Crowding in Living Cells. Nat. Methods 2015, 12, 227−229. (26) Gnutt, D.; Gao, M.; Brylski, O.; Heyden, M.; Ebbinghaus, S. Excluded-Volume Effects in Living Cells. Angew. Chem., Int. Ed. 2015, 54, 2548−2551. (27) Biswas, S.; Chowdhury, P. K. Unusual Domain Movement in a Multidomain Protein in the Presence of Macromolecular Crowders. Phys. Chem. Chem. Phys. 2015, 17, 19820−19833. (28) Forster, T. Intermolecular Energy Migration and Fluorescence. Ann. Phys. 1948, 437, 55−75. (29) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, 2006; p xxvi, 954 pp. (30) Wu, P. G.; Brand, L. Resonance Energy Transfer: Methods and Applications. Anal. Biochem. 1994, 218, 1−13. (31) Biskup, C.; Zimmer, T.; Kelbauskas, L.; Hoffmann, B.; Klöcker, N.; Becker, W.; Bergmann, A.; Benndorf, K. Multi-Dimensional Fluorescence Lifetime and FRET Measurements. Microsc. Res. Technol. 2007, 70, 442−451. (32) Hildebrandt, L. L.; Preus, S.; Birkedal, V. Quantitative Single Molecule FRET Efficiencies Using Tirf Microscopy. Faraday Discuss. 2015, 184, 131−142. (33) Thaler, C.; Koushik, S. V.; Puhl, H. L.; Blank, P. S.; Vogel, S. S. Structural Rearrangement of CaMKIIα Catalytic Domains Encodes Activation. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 6369−6374. (34) Haupts, U.; Maiti, S.; Schwille, P.; Webb, W. W. Dynamics of Fluorescence Fluctuations in Green Fluorescent Protein Observed by Fluorescence Correlation Spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 1998, 95, 13573−13578. (35) Heikal, A. A.; Hess, S. T.; Baird, G. S.; Tsien, R. Y.; Webb, W. W. Molecular Spectroscopy and Dynamics of Intrinsically Fluorescent Proteins: Coral Red (Dsred) and Yellow (Citrine). Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 11996−12001. (36) Heikal, A. A.; Hess, S. T.; Webb, W. W. Multiphoton Molecular Spectroscopy and Excited-State Dynamics of Enhanced Green Fluorescent Protein (EGFP): Acid−Base Specificity. Chem. Phys. 2001, 274, 37−55. (37) Heikal, A. A.; Hess, S. T.; Webb, W. W. Multiphoton Molecular Spectroscopy and Excited-State Dynamics of Enhanced Green Fluorescent Protein (EGFP): Acid-Base Specificity. Chem. Phys. 2001, 274, 37−55. (38) Hess, S. T.; Heikal, A. A.; Webb, W. W. Fluorescence Photoconversion Kinetics in Novel Green Fluorescent Protein pH Sensors (Phluorins). J. Phys. Chem. B 2004, 108, 10138−10148. (39) Schwille, P.; Kummer, S.; Heikal, A. A.; Moerner, W. E.; Webb, W. W. Fluorescence Correlation Spectroscopy Reveals Fast Optical Excitation-Driven Intramolecular Dynamics of Yellow Fluorescent Proteins. Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 151−156. (40) Schwille, P.; Haupts, U.; Maiti, S.; Webb, W. W. Molecular Dynamics in Living Cells Observed by Fluorescence Correlation Spectroscopy with One- and Two-Photon Excitation. Biophys. J. 1999, 77, 2251−2265. (41) Heikal, A. A. Time-Resolved Fluorescence Anisotropy and Fluctuation Correlation Analysis of Major Histocompatibility Complex Class I Proteins in Fibroblast Cells. Methods 2014, 66, 283−291.

(42) Liu, B.; Åberg, C.; van Eerden, F. J.; Marrink, S. J.; Poolman, B.; Boersma, A. J. Design and Properties of Genetically Encoded Probes for Sensing Macromolecular Crowding. Biophys. J. 2017, 112, 1929− 1939. (43) Yu, Q.; Proia, M.; Heikal, A. A. Integrated Biophotonics Approach for Noninvasive and Multiscale Studies of Biomolecular and Cellular Biophysics. J. Biomed. Opt. 2008, 13, 041315. (44) Becker, W. Advanced Time-Correlated Single-Photon Counting Techniques; Springer: 2005. (45) O’Connor, D. V.; Phillips, D. Time-Correlated Single Photon Counting; Academic Press: 1984. (46) Davey, A. M.; Walvick, R. P.; Liu, Y.; Heikal, A. A.; Sheets, E. D. Membrane Order and Molecular Dynamics Associated with IgE Receptor Cross-Linking in Mast Cells. Biophys. J. 2007, 92, 343−355. (47) Gendron, P.-O.; Avaltroni, F.; Wilkinson, K. J. Diffusion Coefficients of Several Rhodamine Derivatives as Determined by Pulsed Field Gradient−Nuclear Magnetic Resonance and Fluorescence Correlation Spectroscopy. J. Fluoresc. 2008, 18, 1093. (48) Widengren, J.; Mets, U.; Rigler, R. Fluorescence Correlation Spectroscopy of Triplet States in Solution: A Theoretical and Experimental Study. J. Phys. Chem. 1995, 99, 13368−13379. (49) Hess, S. T.; Huang, S.; Heikal, A. A.; Webb, W. W. Biological and Chemical Applications of Fluorescence Correlation Spectroscopy: A Review. Biochemistry 2002, 41, 697−705. (50) Kubin, R. F.; Fletcher, A. N. Fluorescence Quantum Yields of Some Rhodamine Dyes. J. Lumin. 1982, 27, 455−462. (51) Heikal, A. A.; Multiparametric, A. Fluorescence Approach for Biomembrane Studies. In Advances in Planar Lipid Bilayers and Liposomes; Iglic, A., Ed.; Elsevier: 2011; Vol. 13, pp 169−197. (52) Hur, K.-H.; Mueller, J. D. Quantitative Brightness Analysis of Fluorescence Intensity Fluctuations in E. Coli. PLoS One 2015, 10, e0130063. (53) Markwardt, M. L.; Kremers, G.-J.; Kraft, C. A.; Ray, K.; Cranfill, P. J. C.; Wilson, K. A.; Day, R. N.; Wachter, R. M.; Davidson, M. W.; Rizzo, M. A. An Improved Cerulean Fluorescent Protein with Enhanced Brightness and Reduced Reversible Photoswitching. PLoS One 2011, 6, e17896. (54) Nakabayashi, T.; Oshita, S.; Sumikawa, R.; Sun, F.; Kinjo, M.; Ohta, N. Ph Dependence of the Fluorescence Lifetime of Enhanced Yellow Fluorescent Protein in Solution and Cells. J. Photochem. Photobiol., A 2012, 235, 65−71. (55) Warren, S. C.; Margineanu, A.; Katan, M.; Dunsby, C.; French, P. M. W. Homo-FRET Based Biosensors and Their Application to Multiplexed Imaging of Signalling Events in Live Cells. Int. J. Mol. Sci. 2015, 16, 14695−14716. (56) Hess, S. T.; Sheets, E. D.; Wagenknecht-Wiesner, A.; Heikal, A. A. Quantitative Analysis of the Fluorescence Properties of Intrinsically Fluorescent Proteins in Living Cells. Biophys. J. 2003, 85, 2566−80. (57) Shi, X.; Basran, J.; Seward, H. E.; Childs, W.; Bagshaw, C. R.; Boxer, S. G. Anomalous Negative Fluorescence Anisotropy in Yellow Fluorescent Protein (Yfp 10c): Quantitative Analysis of FRET in Yfp Dimers. Biochemistry 2007, 46, 14403−14417. (58) Hussain, I.; et al. Identification of a Novel Aspartic Protease (Asp 2) as Beta-Secretase. Mol. Cell. Neurosci. 1999, 14, 419−27.

5698


Fluorescence Dynamics of a FRET Probe Designed for Crowding

Recommend Documents