Dissecting Nanosecond Dynamics in Membrane Proteins with Dipolar

Subscriber access provided by - Access paid by the | UCSB Libraries

Biophysical Chemistry, Biomolecules, and Biomaterials; Surfactants and Membranes

Dissecting Nanosecond Dynamics in Membrane Proteins with Dipolar Relaxation upon Tryptophan Photoexcitation Erik Frotscher, Georg Krainer, Michael Schlierf, and Sandro Keller J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b00834 • Publication Date (Web): 13 Apr 2018 Downloaded from http://pubs.acs.org on April 15, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Dissecting Nanosecond Dynamics in Membrane Proteins with Dipolar Relaxation upon Tryptophan Photoexcitation Erik Frotscher,1 Georg Krainer,1,2 Michael Schlierf,2 and Sandro Keller1*

1

Molecular Biophysics, Technische Universität Kaiserslautern (TUK), Erwin-SchrödingerStr. 13, 67663 Kaiserslautern, Germany

2

B CUBE – Center for Molecular Bioengineering, Technische Universität Dresden, Arnoldstr. 18, 01307 Dresden, Germany

*Corresponding author. E-mail: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 19

ABSTRACT The structural dynamics of proteins on the nanosecond timescale can be probed with dipolar relaxation in response to photoexcitation of intrinsic tryptophan (Trp) residues. For membrane proteins, however, the complexity due to overlapping contributions from the protein itself, the membrane mimic, and the aqueous solvent impairs detailed analysis and interpretation. To disentangle these contributions, we measured time-resolved emission spectra of Trp in the protein Mistic in detergent micelles of various polarities. By comparison with Trp analogues in water and micelles, we could dissect the contributions from hydration, micelle, and protein matrix to dipolar relaxation on the nanosecond timescale. Our results demonstrate that ultrafast, subnanosecond relaxation reports on the extent of Trp shielding from water, with micelle and protein moieties making additive contributions. By contrast, relaxation in the low nanosecond regime is due to dipolar rearrangement of micelle and protein moieties upon photoexcitation, thereby probing conformational dynamics around the intrinsic fluorophore.

TOC GRAPHIC


Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Proteins exhibit conformational dynamics on a broad range of timescales ranging from femtoseconds to seconds.1,2 A substantial part of functionally important protein dynamics occurs on the nanosecond timescale, including motions of side chains, secondary-structure elements, disordered regions, and individual protein domains.1,3 A powerful methodology for studying such small-scale protein dynamics in the nanosecond regime consists in monitoring dipolar relaxation upon photoexcitation. This can be accomplished with the aid of wavelength-selective steady-state fluorescence spectroscopy4–8 or, to afford higher information content, time-resolved fluorescence spectroscopy, where time-dependent fluorescence shifts (TDFSs) serve as sensitive metrics of dipolar relaxation.9–16 Of particular interest is the use of the indole moiety of tryptophan (Trp) residues as an intrinsic probe that reports on native protein dynamics without the need of labels.10 Yet, a major hurdle in the interpretation of TDFS of intrinsic protein fluorescence consists in the complexity and interdependence of subnanosecond and nanosecond phenomena that contribute to the overall relaxation process upon Trp photoexcitation.10 In aqueous environments, both hydration and protein dynamics contribute to dipolar relaxation of watersoluble proteins.17–19 The situation is further complicated in the case of the physiologically and therapeutically important class of membrane proteins, which reside in the heterogeneous, anisotropic milieu of a lipid-bilayer membrane. Even in the absence of proteins, TDFS studies exploiting extrinsic fluorescent probes have revealed complex relaxation dynamics within membranes20–22 and membrane mimics.23 Hence, for proteins embedded in a lipid-bilayer membrane or a membrane mimic, the situation is even more complex because the relaxation dynamics of the membrane (mimic) are intimately coupled to hydration and protein dynamics. Therefore, to exploit the great potential of TDFS experiments for enhancing our understanding of the functionally important nanosecond dynamics of membrane proteins, it is essential to disentangle the contributions from the solvent, the membrane (mimic), and the protein itself to dipolar relaxation upon photoexcitation. This appears also particularly 3 ACS Paragon Plus Environment


Page 4 of 19

valuable in the light of the rapidly increasing accessibility of time-resolved intrinsic protein fluorescence spectroscopy, which has become a mature tool for studying the complexities of membrane-protein dynamics. Instrumentation has greatly improved over the past years, with

∼280-nm light-emitting diodes (LEDs) for pulsed Trp excitation now being available even in off-the-shelf, bench-top fluorescence spectrometers fitted with time-correlated single-photon counting (TCSPC). Here, we developed a rigorous approach for dissecting the different contributions to nanosecond dipolar relaxation in membrane proteins by exploiting a native single-Trp protein and two Trp analogues as simple model systems. The bacterial membrane-associated protein Mistic24 has been characterized in great detail in detergent micelles of various chain lengths and carrying different polar headgroups.25–27 Water-soluble N-acetyl-L-tryptophanamide (NATA)28–30 and micelle-embedded

DL-tryptophan

octyl ester (TOE),23,31–33 served as model

compounds for probing indole fluorescence in an isotropic aqueous solution and in an anisotropic membrane mimic, respectively. We reconstructed time-resolved emission spectra (TRES) from fluorescence decays at different emission wavelengths, which were then parametrized in the form of multiexponential decay functions (Figure 1a,b; eqs 1 and 2). TRES revealed stark differences between, on the one hand, NATA in aqueous solution (Figure 1c) and, on the other hand, TOE and Mistic in micelles of n-dodecyl-β-D-maltopyranoside (DDM) (Figure 1d,e). While TRES of NATA were almost identical across the entire time window, prominent TDFSs were observed for TOE and Mistic, with TRES shifting to lower emission energies with increasing time after excitation. To follow the temporal change quantitatively, we used the spectral barycenter, νbc(t), (eq 3) as a measure of average emission energy (Figure 1f). As expected from visual inspection of TRES, νbc(t) was virtually constant for NATA but monotonically decreased over time for TOE and Mistic in DDM, thereby evidencing relaxation processes on the nanosecond timescale.29,34 Notably, previous TRES of TOE in DDM have shown a similar TDFS.23 4 ACS Paragon Plus Environment

Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. TDFSs of intrinsic Trp fluorescence. (a) Representative set of fluorescence decays of Mistic in DDM showing emission intensities, I, as functions of time, t, and emission wavelength, λ. (b) Triexponential decay functions derived from data in (a) by reconvolution. (c–e) TRES of (c) NATA in buffer, (d) TOE in DDM, and (e) Mistic in DDM. TRES show peak-normalized I values as functions of λ and wavenumber, ν, at different time points after excitation as indicated by the color code. (f) Spectral barycenter in units of wavelengths, λbc, and wavenumbers, νbc, as a function of time, t, for NATA, TOE, and Mistic. 8 µM TOE or 4 µM Mistic in 5 mM micellar DDM or 8 µM NATA at 20°C. Excitation was at (281 ± 5) nm.

The spectral barycenter extrapolated to t = 0, ν0, showed that, at ultrashort times, emission occurred at higher energies for Mistic (29 070 cm−1) than for TOE (28 560 cm−1) and NATA (26 970 cm−1). Importantly, the true time-zero emission spectrum of Trp is estimated to have a v0 ≈ 30 000 cm−1 (eq 4),35 which is significantly higher than the extrapolated ν0 values found here. Thus, a substantial fraction of the overall relaxation occurs before the accessible time window, which ranges from ∼50 ps to ∼15 ns. Ultrafast relaxation on the femto- to picosecond 5 ACS Paragon Plus Environment


Page 6 of 19

timescale is mainly due to reorientation of water dipoles.3,17,18,36 Consequently, ν0 reports on the extent of ultrafast, hydration-associated relaxation, which is expected to be more pronounced in well-hydrated environments. Indeed, ν0 assumed the lowest value in the case of NATA, where the fluorophore was shielded neither by a micelle nor by a protein matrix, so that relaxation was completed within the ultrafast regime.10,29,37 The higher ν0 values observed for TOE and, in particular, Mistic demonstrated that the indole fluorophore experienced a less polar environment resulting from additional hydrophobic burial by the micellar environment and, even more, by the protein matrix, respectively. On the nanosecond timescale, νbc(t) then decreased much more slowly for Mistic than for TOE, as quantified by the normalized correlation function defined as C(t) = (νbc(t)−νbc(15 ns))/(ν0−νbc(15 ns)).21 Note that C(t) represents the amplitude-normalized temporal change in νbc(t) within the first 15 ns after excitation, and integration of C(t) yields the correlation time, τcor, as a metric of the relaxation kinetics within this time window. Together, C(t) and τcor allow for a straightforward quantification of nanosecond dynamics without requiring the assumption of a specific model function to fit temporal changes in νbc(t). Importantly, τcor was found to be much higher for Mistic in DDM (5.9 ns) than for TOE in DDM (2.8 ns), thus demonstrating that the protein matrix slowed down dipolar relaxation also on the nanosecond timescale. To dissect the influence of hydration as well as micelle and protein dynamics in greater detail, we further tested the roles of micellar hydrophobic thickness and detergent headgroup hydration during dipolar relaxation. To this end, we used a homologous series of non-ionic alkyl maltoside detergents bearing 8, 10, and 12 carbon atoms in their alkyl chains (i.e., n-octyl-β-D-maltopyranoside

(OM),

n-decyl-β-D-maltopyranoside

(DM),

and

DDM,

respectively) and a zwitterionic C12 dimethylamine oxide detergent (lauryldimethylamineN-oxide, (LDAO)). We have previously shown25 that long alkyl chains such as in DDM and ionic headgroups such as in LDAO contribute favorably to the conformational stability of Mistic; by contrast, the protein is highly susceptible to unfolding when solubilized in the non6 ACS Paragon Plus Environment

Page 7 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ionic short-chain detergent OM. Furthermore, single-molecule Förster resonance energy transfer (FRET) spectroscopy on a labeled variant of Mistic has revealed that polar interactions with ionic headgroups result in much faster protein-folding rates on the microsecond timescale as compared with non-ionic detergents.27 Following νbc(t) of TOE and Mistic in different detergents (Figure 2a) yielded two initial observations: First, differences in νbc(t) among detergents were much smaller for TOE than for Mistic. Second, for all detergents tested, νbc(t) and, thus, emission energies of Mistic’s Trp were always higher than those of their TOE counterparts.

Figure 2. Dipolar relaxation of TOE and Mistic in different detergents. a) Spectral barycenter in units of wavenumbers, νbc, and wavelengths, λbc, as a function of time, t, for TOE and Mistic in different detergents. b) Difference in ν0 between Mistic and TOE, ∆ν0. c) Correlation time within 0–15 ns, τcor, for Mistic. Panels b and c show mean values from three experiments with standard deviations often smaller than symbols. 8 µM TOE or 4 µM Mistic in 5 mM micellar detergent at 20°C. Excitation was at (281 ± 5) nm.

In order to probe the influence of micellar properties on hydration-associated dipolar relaxation and nanosecond dynamics, respectively, we evaluated ν0 and τcor as functions of detergent polarity, which increases upon going from long-chain non-ionic DDM to shortchain non-ionic OM and further to zwitterionic LDAO. For TOE, ν0 decreased from DDM (28 560 cm−1) and DM (28 460 cm−1) to OM (28 210 cm−1) and LDAO (28 190 cm−1). This slight but systematic trend was expected because short-chain detergents provide less effective hydrophobic burial within micelles than long-chain detergents38 and because zwitterionic headgroups allow deeper water penetration into micelles.39–41 This interpretation of ν0 is 7 ACS Paragon Plus Environment


Page 8 of 19

further corroborated by quenching studies33 on detergent-solubilized TOE showing greater accessibility to the water-soluble quencher acrylamide in micelles containing a zwitterionic phosphocholine as compared with non-ionic headgroups. In contrast with ν0, we found no significant influence of the micellar environment on the nanosecond kinetics of dipolar relaxation up to 15 ns, as reflected in τcor ≈ 2.7 ns in all cases. In summary, differences in detergent polarity clearly manifested in ultrafast hydration dynamics and, thus, ν0, but nanosecond dipolar relaxation in micelles was independent of detergent polarity. In the case of Mistic’s lone Trp residue, ν0 showed a similar trend as observed for TOE, with the values in DDM (29 070 cm−1), DM (29 020 cm−1), and LDAO (28 690 cm−1) being ~500 cm−1 larger than the corresponding TOE values (Figure 2b). Thus, additional solvent shielding by the protein matrix added a virtually constant contribution to the extent of shielding provided by the micellar environment. For OM, by contrast, the difference in ν0 between Mistic and TOE was considerably smaller, amounting to only ~300 cm−1. This is readily explained by the low equilibrium stability of Mistic in this environment, with the freeenergy difference between the folded and unfolded states being < 7 kJ mol−1.25 Its marginal stability in OM renders Mistic highly susceptible to unfolding and, consequently, results in significant ground-state heterogeneity, as reflected here in a substantially lower degree of solvent shielding at ultrashort times. Again, a different picture emerged on the nanosecond timescale, where detergent polarity turned out to be the main determinant of τcor. Intriguingly, τcor was much larger in all non-ionic detergents (~6 ns in DDM, DM, and OM) than in zwitterionic LDAO (3.7 ns) (Figure 2c). As our TOE experiments showed that τcor is indifferent to detergent properties per se, the significant variations found for Mistic must be ascribed to conformational dynamics at the protein level. Hence, Mistic displays not only faster microsecond, large-scale folding dynamics27 in LDAO (~30 µs) than in DDM (~700 µs) but also faster nanosecond, small-scale conformational adaptations in response to photoexcitation in zwitterionic than in non-ionic micelles. The generally faster conformational 8 ACS Paragon Plus Environment

Page 9 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


dynamics across various timescales in zwitterionic LDAO micelles could explain why Mistic is exceptionally stable in this detergent,25 as the folded polypeptide chain enjoys greater conformational freedom and, thus, suffers from a smaller entropic penalty upon folding. In summary, time-resolved fluorescence spectroscopy on the single-Trp protein Mistic in various membrane-mimetic environments in combination with the model compounds NATA and TOE has enabled detailed analysis and interpretation of the (sub)nanosecond dynamics of the overlapping contributions from hydration, micelle, and intrinsic protein motions to dipolar relaxation: (i) Ultrafast dipolar relaxation, as reflected in ν0, reports on protein and micellar hydration around the Trp and, therefore, largely depends on the extent of shielding of the fluorophore, with micelle and protein moieties contributing additively to ν0. (ii) Slow relaxation on the nanosecond timescale is due to dipolar rearrangement of micelle and protein moieties upon photoexcitation, which can be quantified in terms of τcor for the low nanosecond regime. Thus, ν0 probes water accessibility and τcor conformational dynamics around the intrinsic fluorophore. Together, ν0 and τcor values extracted from TRES allow a molecularly detailed interpretation of TDFSs observed for intrinsic protein fluorophores without the need of labels. Therefore, high-content analysis of time-resolved intrinsic protein fluorescence constitutes a powerful method for monitoring nanosecond membrane-protein dynamics that is capable of disentangling contributions from the solvent, the membrane mimic, and the protein itself. This lays the foundation for a better understanding of the intricate relationships among structural dynamics, stability, and function of the many physiologically and therapeutically relevant proteins that populate the heterogeneous, anisotropic milieu of a lipid-bilayer membrane.



Page 10 of 19

EXPERIMENTAL METHODS Materials n-Dodecyl-β-D-maltopyranoside (DDM), n-decyl-β-D-maltopyranoside (DM), and n-octylβ-D-maltopyranoside (OM) were from Glycon Biochemicals (Luckenwalde, Germany. Lauryldimethylamine N-oxide (LDAO) was from Anatrace (Maumee, USA) and DL-tryptophan

octyl ester (TOE) hydrochloride from Santa Cruz Biotechnology (Dallas,

USA). N-acetyl-L-tryptophanamide (NATA) and 1,4-dithiothreitol (DTT) were from Sigma– Aldrich (Steinheim, Germany). Tris(hydroxymethyl)-aminomethane (Tris) was from Carl Roth (Karlsruhe, Germany) and NaCl from VWR (Darmstadt, Germany). All chemicals were purchased in the highest purity available. A stock solution of TOE at a concentration of 2 mM was prepared in EtOH and diluted with detergent-containing buffer. Protein production and purification and detergent exchange were performed as described elsewhere.25

Fluorescence spectroscopy Samples contained 8 µM NATA in buffer (50 mM Tris, 50 mM NaCl, pH 7.4) or 8 µM TOE or 4 µM Mistic in 5 mM micellar detergent corresponding to 5.15 mM DDM, 7.03 mM DM, 26.6 mM OM, or 6.84 mM LDAO.25 For Mistic samples, 5 mM DTT was freshly added prior to measurements to prevent dimerization through the single cysteine residue. For each probe and detergent, three independent samples were measured, which demonstrated excellent reproducibility with relative differences in νbc(t) values of

Dissecting Nanosecond Dynamics in Membrane Proteins with Dipolar

Recommend Documents