Laser-Induced Population Inversion in Rhodamine 6G for Lysozyme

From the Bench pubs.acs.org/biochemistry

Laser-Induced Population Inversion in Rhodamine 6G for Lysozyme Oligomer Detection Piotr Hanczyc*,† and Lech Sznitko‡ †

Center for Polymers & Organic Solids, University of California, Santa Barbara, 2520A Physical Sciences Building North, Santa Barbara, California 93106, United States ‡ Advanced Materials Engineering and Modelling Group, Faculty of Chemistry, Wroclaw University of Technology, 50-370 Wroclaw, Poland ABSTRACT: Fluorescence spectroscopy is a common method for detecting amyloid fibrils in which organic fluorophores are used as markers that exhibit an increase in quantum yield upon binding. However, most of the dyes exhibit enhanced emission only when bound to mature fibrils, and significantly weaker signals are obtained in the presence of amyloid oligomers. In the concept of population inversion, a laser is used as an excitation source to keep the major fraction of molecules in the excited state to create the pathways for the occurrence of stimulated emission. In the case of the proteins, the conformational changes lead to the self-ordering and thus different light scattering conditions that can influence the optical signatures of the generated light. Using this methodology, we show it is possible to optically detect amyloid oligomers using commonly available staining dyes in which population inversion can be induced. The results indicate that rhodamine 6G molecules are complexed with oligomers, and using a laser-assisted methodology, weakly emissive states can be detected. Significant spectral red-shifting of rhodamine 6G dispersed with amyloid oligomers and a notable difference determined by comparison of spectra of the fibrils suggest the existence of specific dye aggregates around the oligomer binding sites. This approach can provide new insights into intermediate oligomer states that are believed to be responsible for toxic seeding in neurodegeneration diseases.

M

ultidisciplinary research of protein aggregation, a phenomenon that is related to numerous brain diseases such as Alzheimer’s disease, Parkinson’s disease, or amyotrophic lateral sclerosis (ALS), continuously provides new insights into the complex nature of the amyloid fibrils.1,2 The neurodegenerative amyloid diseases seem to be caused by smaller oligomers, intermediate in size between the monomers and the fibrils, that can act as seeds for fibril elongation.3 The analysis of the oligomers is more challenging than that of the fibrils as the aggregates are small and turn into fibrils in time.4 Study of the seeding process and early nucleation requires Xray crystallography5 or nuclear magnetic resonance (NMR),6 each of which can resolve the structure of the smallest aggregates, i.e., dimers or tetramers, but larger intermediates do not produce crystals for X-ray or require laborious work on data interpretation in the case of NMR analysis. On the other hand, electron microscopy and atomic force microscopy are often used to study fibrils but have no appropriate resolution to visualize oligomer structures in detail during the early stage of nucleation.7,8 All these techniques leave a gap in studies of fibrillization at different stages of misfolding. Some potential remains in the spectroscopic methodologies in which, for example, Fourier transform infrared spectroscopy can monitor aggregation kinetics and resolve the β-sheet content at different stages of misfolding and determine whether the position of the sheets is parallel or antiparallel in the oligomers and fibrils.9 © 2017 American Chemical Society

Another commonly used method is ultraviolet−visible (UV− vis) fluorescence spectroscopy where amyloids are stained with fluorophores10,11 or polymers,12,13 but most of the studied molecules show a characteristic lag phase in fluorescence kinetic experiments in which no distinct changes in quantum yield can be recorded during the initial aggregation stage (Figure 1a).14 Recently, a BD-oligo fluorophore was shown to exhibit a higher quantum yield in the presence of the Aβ oligomers than in the presence of fibrils,15 and together with crystallography results on the Orange G chromophore complexed with early formed aggregates,16 one could expect that binding of small molecules indeed occurs at the nucleation stage. Singh and co-workers,17 who used the laser-assisted methodology to study thioflavin T (ThT) dynamics when it is complexed with insulin fibrils, suggested that chromophores can exist in a set of different geometries driven by amyloid polymorphism where some are short-lived and weakly emissive. Thus, only the small and highly emissive fraction contributes to the characteristic increase in the intensity of dye emission, whereas weakly emissive species remain inaccessible to standard fluorescence spectroscopy. This leaves an open question of whether it is possible to uncover the weakly emissive states and by that gain Received: March 16, 2017 Revised: May 15, 2017 Published: May 18, 2017 2762

DOI: 10.1021/acs.biochem.7b00243 Biochemistry 2017, 56, 2762−2765

From the Bench

Biochemistry

Figure 1. (a) Schematic drawing representing fluorescence kinetic experiments at different stages of amyloid fibril formation. (b) Laser setup used for the analysis of amyloid−rhodamine 6G thin films.

the horizontal size of the exciting beam, while the cylindrical lens was responsible for focusing light in the vertical direction. Therefore, in the plane of the samples, the excitation beam was formed into a 3 mm long stripe with a height of 0.5 mm. Such pumping geometry provides the largest gain in the direction of the longer stripe dimension, so it is possible to collect the emitted light waveguiding in the sample layer from its edges. For spectral analysis, we have used the Andor Shamrock 163 fiber spectrometer with a spectral resolution of 0.02 nm coupled with a personal computer for data acquisition and directly triggered by the Nd:YAG laser. A 1 % stock solution of the dye was dispersed in solutions with three different samples [(1) monomers, (2) fibrils, and (3) oligomers] in volume ratios of 1:10. Fresh monomer lysozyme protein (1) from hen egg white (Sigma-Aldrich) was dissolved at pH 2 (0.01 M HCl) to a concentration of 20 mg/mL and equilibrated with rhodamine 6G. Only the freshly prepared solution was drop-casted to prevent any potential aggregation. Lysozyme fibrils (2) were prepared by keeping the monomer stock solution (20 mg/mL) at an elevated temperature (65 °C) for 6 days. Formed aggregates were spun at 3000 rpm for 3 min, and the collected precipitant was dispersed again in the acidic pH 2 buffer containing rhodamine dye. The protocol was repeated three times prior to preparation of the solid state samples. Lysozyme oligomers (3) were prepared by aggregating monomer protein as described for lysozyme fibrils (2) and spun at 25000 rpm for 30 min to separate the gradient soluble oligomers from the fibrils and insoluble clumps that precipitated. Next, the collected supernatant with oligomers and remaining monomers was transferred to a column filter (Amicon Millipore) with 50 kDa pores and spun at 22000 rpm for 10 min. The size of the pores allows us to separate unaggregated monomers (14.3 kDa) that are small enough to pass the filter from the remaining solute with soluble oligomers

access to the early stages of aggregation where knowledge about fluorophore−oligomer interactions remains elusive. Here we propose a simple extension of UV−vis fluorescence methodology to overcome the limitation regarding spectroscopic detection of oligomers by implementing a stronger excitation source that is capable of exciting weakly emissive (aggregated) species of the chromophores. In this work, we expand our previous research on dye-doped fibrils18,19 to amyloid oligomers and show that stimulated emission spectra can be significantly different depending on the spatial organization of chromophores that is driven by protein structure and conformation. Nanosecond laser pulses that can induce population inversion at relative low pumping energies (∼1 mJ/cm2) in rhodamine 6G (Rh6G) were used for that purpose (shown in Figure 1b). The Surelite II Nd:YAG laser (by Continuum) operating at wavelength of 355 nm was used for Horizon I Optical Parametric Oscillator (OPO) pumping. The utilization of OPO allows us to change the laser wavelength in the whole visible range of light. Light at a wavelength of 532 nm with vertical polarization (supporting efficient pumping of Rh6G molecules) was extracted from OPO and directed through a half-wave plate and Glan-Laser polarizer optical system. The Glan-Laser polarizer was used to continue to polarize light vertically, while rotation of the half-wave plate allowed us to change the azimuth of polarization of incident light. Therefore, the polarization of the outgoing beam was vertical, but according to Malus law, it was possible to change precisely the intensity of the outgoing beam by rotation of the half-wave plate azimuth. Then the beam was passing through a beam expanding (afocal) system providing the 5-fold spatial extension of the beam. The cylindrical lens was then used to provide vertical light that was focused on the sample. The utilization of the adjusted slit before the cylindrical lens allowed us to control 2763

DOI: 10.1021/acs.biochem.7b00243 Biochemistry 2017, 56, 2762−2765

From the Bench

Biochemistry with molecular weights between 74 and 590 kDa according to the method described by Frere et al.20 Finally, a dye:volume ratio of 1:10 was added to isolated and purified oligomers and prepared for measuring the stimulated emission. Thirty microliters of solutions 1−3 were poured onto glass substrates (0.5 cm × 0.5 cm) and allowed to evaporate under ambient conditions. In such small plates, the surface tension at the glass edge is higher than in drop-casted samples, which favors a homogeneous material distribution over the casted area and helps to avoid “coffee ring” formation at the edges. To minimize sample differences that may affect the optical signal, the thin films were additionally pressed with the silicon mold (2.8 MPa) to make the surface flat and uniform and to create equal conditions for light propagation in each sample. ATR-FTIR spectra in selected amide I (1600−1700 cm−1) and II (1450−1600 cm−1) regions of monomer protein and aggregates confirmed that doping with rhodamine 6G, solvent evaporation, or high-pressure treatment of thin films has no influence on the protein structure or conformation (Figure 2).

Figure 2. FTIR spectra of native lysozyme (bottom) and aggregated samples in the fibril state (middle) and oligomers (top). All examined samples were doped with rhodamine 6G. Figure 3. (a) Steady state emission from rhodamine 6G mixed with monomer lysozyme (blue), amyloid oligomers (red). and fibrils (black). (b) Stimulated emission spectra of the same samples in the solid state.

There is a distinct spectral shift in the amide I region (∼25 cm−1) upon aggregation related to misfolding where lysozyme α-helices (1652 cm−1) turn into β-sheets. However, differentiation between oligomers (1627 cm−1) and fibrils (1630 cm−1) remain rather intuitive because of the very small spectral shift. Fluorescence experiments confirm the low sensitivity of rhodamine 6G to structural differences that appear upon misfolding as only minor spectral variations were recorded between monomers (λspon. em. max. = 523 nm) and fibrils or oligomers (λspon. em. max. = 525 nm) with a nearly identical quantum yield in each case (Figure 3a). However, differences become pronounced when population inversion is induced by the Nd:YAG pulsed laser (Figure 3b). Laser excitation (∼1 mJ/ cm2) keeps the fluorophores in the population inversion state, which allows for structural analysis of the doped protein structures above the emission amplification threshold. The attraction of Rh6G molecules to the binding sites is accelerated by the solvent evaporation that induces molecular crowding of the proteins and dye molecules.18 Analyzing the spectral features in the thin film state is particularly convincing because noninteracting dye molecules form microcrystals that are not contributing to the signal from the protein−Rh6G complex. The photons pass through the crystals, and no light amplification can be observed from that species.19

The monomer, native lysozyme protein retains the α-helical structure in the thin film (Figure 2). The rhodamine molecules are evenly distributed in the layer, and the stimulated emission profile showed spectral features that can be associated with the dimeric form of the chromophore21 with a λstim. em. max. of 582 nm (Figure 3b). In the misfolded protein, the α-helix conformation changes into a β-sheet creating grooves at the surface that favors the interactions with the dye molecules. Together with the accelerated attraction between the amyloids and chromophores caused by solvent evaporation, it creates optimal conditions for local aggregation of rhodamine 6G around the binding sites. Thus, the large spectral red-shifts comparing to dimers of ∼23 and ∼35 nm in the presence of fibrils (λstim. em. max. = 605 nm) and oligomers (λstim. em. max. = 617 nm), respectively, can be explained only by the formation of higher-order aggregates that accumulated at the amyloid βsheet surface.18 This effect clearly demonstrates that dye molecules are interacting with the lysozyme oligomers, and the ∼10 nm spectral shift compared to cylindrical fibrils indicates that fluorophores adopt different spatial geometries in the presence of the spherical oligomers. Rhodamine 6G aggregates of certain sizes and geometries, deposited at the oligomer β2764

DOI: 10.1021/acs.biochem.7b00243 Biochemistry 2017, 56, 2762−2765

From the Bench

Biochemistry

(10) Younan, N. D., and Viles, J. H. (2015) Biochemistry 54, 4297− 4306. (11) Bäcklund, F. G., Pallbo, J., and Solin, N. (2016) Biopolymers 105, 249−259. (12) Hanczyc, P., Justyniarski, A., Gedefaw, D. A., Andersson, M., Samoc, M., and Müller, C. (2015) RSC Adv. 5, 49363−49368. (13) Wigenius, J., Persson, G., Widengren, J., and Inganäs, O. (2011) Macromol. Biosci. 11, 1120−1127. (14) Arosio, P., Knowles, T. P., and Linse, S. (2015) Phys. Chem. Chem. Phys. 17, 7606−7618. (15) Teoh, C. L., Su, D., Sahu, S., Yun, S.-W., Drummond, E., Prelli, F., Lim, S., Cho, S., Ham, S., Wisniewski, T., and Chang, Y.-T. (2015) J. Am. Chem. Soc. 137, 13503−13509. (16) Landau, M., Sawaya, M. R., Faull, K. F., Laganowsky, A., Jiang, L., Sievers, S. A., Liu, J., Barrio, J. R., and Eisenberg, D. (2011) PLoS Biol. 9, e1001080. (17) Singh, P. K., Mora, A. K., and Nath, S. (2015) Chem. Commun. 51, 14042−14045. (18) Hanczyc, P., Sznitko, L., Zhong, C., and Heeger, A. J. (2015) ACS Photonics 2, 1755−1762. (19) Sznitko, L., Hanczyc, P., Mysliwiec, J., and Samoc, M. (2015) Appl. Phys. Lett. 106, 023702. (20) Frare, E., Mossuto, M. F., de Laureto, P. P., Tolin, S., Menzer, L., Dumoulin, M., Dobson, C. M., and Fontana, A. (2009) J. Mol. Biol. 387, 17−27. (21) Martínez Martínez, V., López Arbeloa, F., Bañuelos Prieto, J., and López Arbeloa, I. (2005) J. Phys. Chem. B 109, 7443−7450.

sheet surface, can be then detected because of their longer emission lifetimes, compared to those of the monomers, that create additional and more probable channels for the occurrence of stimulated emission.21 Thus, the long-living states can help to detect amyloid oligomers with high accuracy and provide a comprehensive set of structural information about amyloid intermediates at the nucleation stage. In conclusion, a simple extension of fluorescence spectroscopy was introduced where the laser-assisted methodology shows the ability to differentiate lysozyme monomers, oligomers, and fibrils in the solid thin films by monitoring the position of amplified emission peaks of rhodamine 6G aggregates. Because the optical signatures are mediated by the protein state, we presume that this approach can be also applied for recognition of oligomers in the liquid phase, whereas the interpretation of results may be not trivial because of the progressing seeding and instability of the intermediate forms. Unlike standard fluorescence where dyes exhibit enhanced emission when they are bound to amyloids, in the laser-assisted methodology any fluorophore for which population inversion occurs can be successfully applied to detect oligomeric and fibrillar species by monitoring the peak position above the light amplification threshold. The proposed approach can be of particular interest for resolving the spectral properties of intermediate forms of oligomers that do not produce crystals or are too large for NMR analysis. The insight into the nucleation phase obtained in the process of population inversion can be helpful for understanding the seeding mechanism that is considered to be the key stage in neurodegeneration.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Piotr Hanczyc: 0000-0002-1460-8477 Funding

P.H. acknowledges the support from the Swedish Research Council (VR) in the form of an international postdoc grant and Department of the Navy via Office of Naval Research Award N00014-14-1-0580. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Annamalai, K., Gührs, K. H., Koehler, R., Schmidt, M., Michel, H., Loos, C., Gaffney, P. M., Sigurdson, C. J., Hegenbart, U., Schönland, S., and Fändrich, M. (2016) Angew. Chem., Int. Ed. 55, 4822−4825. (2) Landreh, M., Sawaya, M., Hipp, M., Eisenberg, D., Wüthrich, K., and Hartl, F. (2016) J. Intern. Med. 280, 164−176. (3) Verma, M., Vats, A., and Taneja, V. (2015) Annals of the Indian Academy of Neurology 18, 138−145. (4) Klimov, D. K., and Thirumalai, D. (2003) Structure 11, 295−307. (5) Nelson, R., Sawaya, M. R., Balbirnie, M., Madsen, A. Ø., Riekel, C., Grothe, R., and Eisenberg, D. (2005) Nature 435, 773−778. (6) Comellas, G., and Rienstra, C. M. (2013) Annu. Rev. Biophys. 42, 515−536. (7) Gras, S. L., Waddington, L. J., and Goldie, K. N. (2011) Methods Mol. Biol. 752, 197−214. (8) Adamcik, J., and Mezzenga, R. (2012) Curr. Opin. Colloid Interface Sci. 17, 369−376. (9) Sarroukh, R., Goormaghtigh, E., Ruysschaert, J.-M., and Raussens, V. (2013) Biochim. Biophys. Acta, Biomembr. 1828, 2328−2338. 2765

DOI: 10.1021/acs.biochem.7b00243 Biochemistry 2017, 56, 2762−2765