Single Particle Plasmon Sensors as Label-Free Technique To Monitor

Letter pubs.acs.org/NanoLett

Single Particle Plasmon Sensors as Label-Free Technique To Monitor MinDE Protein Wave Propagation on Membranes Christina Lambertz,† Ariadna Martos,‡ Andreas Henkel,† Andreas Neiser,† Torben-Tobias Kliesch,§ Andreas Janshoff,§ Petra Schwille,‡ and Carsten Sönnichsen*,† †

Institute of Physical Chemistry, University of Mainz, Duesbergweg 10-14, D-55128 Mainz, Germany Department of Cellular and Molecular Biophysics, Max Planck Institute of Biochemistry, Am Klopferspitz 18, D-82152 Martinsried, Germany § Institute of Physical Chemistry, University of Goettingen, Tammannstrasse 6, D-37077 Goettingen, Germany ‡

S Supporting Information *

ABSTRACT: We use individual gold nanorods as pointlike detectors for the intrinsic dynamics of an oscillating biological system. We chose the pattern forming MinDE protein system from Escherichia coli (E. coli), a prominent example for selforganized chemical oscillations of membrane-associated proteins that are involved in the bacterial cell division process. Similar to surface plasmon resonance (SPR), the gold nanorods report changes in their protein surface coverage without the need for fluorescence labeling, a technique we refer to as NanoSPR. Comparing the dynamics for fluorescence labeled and unlabeled proteins, we find a reduction of the oscillation period by about 20%. The absence of photobleaching allows us to investigate Min proteins attaching and detaching from lipid coated gold nanorods with an unprecedented bandwidth of 100 ms time resolution and 1 h observation time. The long observation reveals small changes of the oscillation period over time. Averaging many cycles yields the precise wave profile that exhibits the four phases suggested in previous reports. Unexpected from previous fluorescence-based studies, we found an immobile static protein layer not dissociating during the oscillation cycle. Hence, NanoSPR is an attractive label-free real-time technique for the local investigation of molecular dynamics with high observation bandwidth. It gives access to systems, which cannot be fluorescently labeled, and resolves local dynamics that would average out over the sensor area used in conventional SPR. KEYWORDS: Optical dark-field spectroscopy, gold nanorod, plasmon sensor, Min system, protein coverage, spatiotemporal dynamics

T

NanoSPR allowed us to monitor in real-time the local protein concentration within the Min oscillation with a precision comparable to confocal microscopy, however, in the absence of fluorescent labels. To the best of our knowledge, this is the first time that the self-assembly process of a biological system is followed without the need of fluorescent tags on the molecular level. Moreover, we were able to compare the oscillation period of labeled and unlabeled proteins and to follow the dynamics with unprecedented bandwidth, which in our case is over 1 h with 100 ms time resolution. In addition, the NanoSPR results show coverage oscillations on top of a static background that is absent in fluorescence-based studies and is indicative of a layer of permanently bound proteins. The layer was also visible in ellipsometry measurements. This discrepancy between fluorescence microscopy and label-free methods might require revisiting the biophysics of the Min cycle with regard to protein layer organization.

he plasmon resonance, that is apparent color, of gold nanoparticles reacts to adsorption and desorption of macromolecules, such as proteins,1 a technique we refer to as NanoSPR in analogy to the renowned surface plasmon resonance (SPR) method on gold films. Both SPR and NanoSPR rely on local changes in refractive index introduced by the sampled molecules as the principle of measurement, which eliminates the need for fluorescence labels. The small detection volume around gold nanoparticles and the possibility to monitor individual nanoparticles allows the study of local attachment/detachment of proteins with high precision.2,3 NanoSPR has been applied successfully, for example, to simultaneously detect the presence of multiple analytes,4 determine protein affinities5,6 and even to monitor individual protein binding events.2,3 Here, we show that NanoSPR can also capture dynamic biomolecular systems with high temporal and spatial resolution. As an example for intrinsic dynamics, we chose the pattern forming MinDE protein system from Escherichia coli (E. coli), a paradigm for self-organized chemical oscillations of membrane-associated proteins that are involved in the bacterial cell division process.7 © XXXX American Chemical Society

Received: February 4, 2016 Revised: April 29, 2016

A

DOI: 10.1021/acs.nanolett.6b00507 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters Our results demonstrate that label-free NanoSPR permits one to follow the intrinsic dynamics of molecular systems with unprecedented spatial resolution. As a label-free technique, it facilitates sample preparation, avoids label-induced artifacts, and gives access to systems, which cannot be fluorescently labeled. Strikingly, photobleaching is not an issue and it is possible to follow protein adsorption and desorption for hours with a temporal resolution down to milliseconds.2 Comparative studies of NanoSPR with fluorescence-based confocal microscopy allow the exploration of the different physical mechanism of signal generation to obtain additional insights into the molecular processes. The so-called Min system, comprised of the proteins MinD, MinE, and MinC, defines the spatial positioning of the cell division site in the bacterium E. coli.8,9 Powered by ATP, these three proteins self-assemble on the inner membrane and oscillate from one pole of the cell to the other.10−12 As a result of this dynamic behavior, the concentration of the division inhibitor MinC exhibits a time-average local minimum at midcell, thus targeting the formation of the division septum there.13,14 Much about the detailed picture about the Min system has emerged from the reconstitution and observation of the propagating waves of labeled MinD, MinE and/or MinC proteins on lipid bilayers in confocal or total internal reflection (TIR) fluorescence microscopy in vitro.15−17 Both techniques reveal the protein density distribution of the labeled species within the focal plane with a lateral and axial resolution of a few hundreds of nanometers. Whereas the lateral dimension of the Min waves (about 50 μm) is well resolved, the axial extend of a Min protein layer (99.99%, Figure 2c). The cycle time we measure falls within the range previously reported22 (40−120 s). A longer period for labeled proteins would be expected from slower diffusion due to their increased hydrodynamic radius, however, the situation might be more involved in the highly coupled and cooperative Min system.15,23,24 The median amplitude of the resonance wavelength oscillation is about 0.5 nm for all three conditions (Figure 2d). Therefore, fluorescence labeling does B


Letter

Nano Letters

Figure 2. Min wave dynamics observed by NanoSPR. (a,b) Evolution of the plasmon resonance wavelength shift Δλ of an individual gold nanorod over time induced by protein coverage changes during the Min wave for labeled MinE (a) and unlabeled MinE (b), a wave that is invisible in fluorescence mode. (c,d) Values for the wave period (c) and oscillation amplitude (d) for Min waves observed on several particles. These experiments were performed with proteins containing no fluorescence labels as well as labeled MinD or MinE proteins, respectively. Crosses represent the results from one nanorod, black symbols indicate the median and standard error of the median (s.e.m.). The wave periods show an increase (slowing down) for labeled proteins.

Figure 3. Min wave dynamics observed over several individual cycles. (a) The lack of photobleaching in NanoSPR allows to record a long (1 h) time trace of a Min wave on an individual plasmon sensor, here with 100 ms time resolution (inset). Slight signal drift is superimposed on the individual oscillations. (b) The time evolution of the wave period extracted from the data in (a) (with a sliding window of 514 s) shows the high periodicity and a very slow but continuous speeding up of the oscillation over time. (c) Average wave profile extracted by superimposing several cycles. The blue symbols show the signal development averaged over 26 individual cycles (light gray lines) with the line width corresponding to the s.e.m. The wave profile shows four phases of Langmuir-type characteristics I−IV. These phases are compatible with the Min oscillation process proposed by Loose et al.15 displayed in (d): (I) MinD from solution attaches to the membrane, (II) MinE is recruited, (III) the MinE/MinD complex leaves the surface and (IV) is recycled into its constituents. This cycle is repeated everywhere on the membrane.

not impact the amount of proteins assembling and disassembling on top of the lipid bilayer. The lack of bleaching allows us to record Min oscillations over an extended period of time with high temporal resolution. Figure 3a shows an observation over 1 h with 100 ms time resolution. The oscillation is remarkably stable with just a slight signal drift, allowing us to look for changes in the oscillation period over time. Indeed, we observe a slight but continuous decrease in the oscillation period by about 10% from 40 to 36 s within 1 h (Figure 3b), that is, the oscillation becomes faster. This somewhat unexpected result could be caused either from concentration changes (such as depletion of ATP) or, most likely, a slight temperature increase in the sample caused by light absorption. The long observation time and signal stability allows one to superimpose many individual turnover cycles to obtain a more accurate average wave profile (Figure 3c). This waveform shows four distinct phases, I−IV, each with Langmuir-type shape, which is in good agreement with the molecular mechanism reported by Loose et al.15 (Figure 3d): (I) MinD from the bulk solution attaches to the membrane (from 0−8 s), (II) MinE is recruited (from 8−29 s), (III) the MinE/MinD complex leaves the surface (from 29−38 s), and is (IV) recycled into its constituents (from 38−45 s). Previous reports had suggested a complete dissociation of the membrane bound proteins at one point in the oscillation cycle.11,16,25 When we investigated the amplitude of the oscillation, we found to our surprise that the modulation depth is only about 10% of the maximum Δλres after addition of

ATP, MinD, and MinE (Figure 4a, statistics for Δλres after addition of Min system with unlabeled and labeled MinD in Figure S1). A possible explanation for the small modulation depth is the existence of an immobile, static protein layer on top of the lipid bilayer that does not dissociate. This “static layer” could not be released from the lipid bilayer in our case (Figure S2). Control experiments excluded impurities (Figure S3), the presence of unintended interactions of proteins with nanoparticles (Figure S4), and identified the origin of the offset in MinD (Figure S4). To exclude nanoparticle-induced artifacts, we performed ellipsometric measurements on the same supported lipid bilayer system in the absence of nanoparticles. Similar to NanoSPR, ellipsometry is a label-free technique that also relies on refractive index changes. The ellipsometry data were analyzed according to Faiss et al.26 and show a nonremovable signal after addition of the proteins to the membrane (Figure 4b) that would be consistent with a static, immobile protein layer. Because such an immobile, static layer is not prominent in previous fluorescence based investigations,15,16 the origin of the observed signal offset in NanoSPR and ellipsometry remains elusive. Our results show the potential of using single plasmonic nanoparticles (NanoSPR) as novel label-free sensor platform C


Nano Letters

■

Letter

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Addresses

(A.M.) Coriolis Pharma, Am Klopferspitz 19, D-82152 Martinsried, Germany. (A.H.) Data Center, University of Mainz, Anselm-Franz-vonBentzelweg 12, D-55128 Mainz, Germany. (A.N.) Institute for Nuclear Physics, University of Mainz, Johann-Joachim-Becher-Weg 45, D-55128 Mainz, Germany. Author Contributions

C.L. initiated and coordinated the project and performed the single particle optical dark-field spectroscopy measurements in the laboratory and with guidance of C.S. A.M. expressed, purified, and labeled the proteins and performed fluorescence measurements in the laboratory and with guidance of P.S. A.H. developed the microscope setup and performed calculations. A.N. developed software for data evaluation. T.-T.K. performed ellipsometry measurements in the laboratory and with guidance of A.J. All authors continuously discussed the project and worked on the manuscript. All authors have given approval to the final version of the manuscript. Funding

Figure 4. Static protein layer. (a) Complete timetraces of the experiment including the as-prepared bilayer show the Min oscillation on top of a large offset. Removing proteins by rinsing with buffer, ADP, or MinE fails to reduce the signal below the oscillation minima. The oscillation takes place on top of a persistent static layer. (b) The static layer can be observed by ellipsometry as well. The data shows the buildup of a monolayer on top of the lipid bilayer followed by higher order attachments. The latter are washed away by rinsing with buffer, the former cannot be removed with either buffer or ADP.

This work was financially supported by the ERC Grant 259640 -“SingleSens” (to C.S.) and by the DFG Leibniz Prize and the HFSP through Grant RGP0050/2010-C102 (to P.S.). A.J. gratefully acknowledges financial support from the DFG through SFB 803 (B08). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Beatrix Scheffer for her assistance with MinDeGFP purification and Falk Butter (Proteomics Core Facility at IMB) for quantification of MinD purity by proteomics. Note: The former name of C.L. was Christina Rosman.

for the real-time investigation of dynamic biological processes out of equilibrium with high spatial resolution. NanoSPR removes the need for attaching fluorescent tags to proteins, thus eliminating a tedious and risky step in loosing much of the functionality of the native biological system. In fact, even small fluorescent dyes could influence molecular dynamics rendering label-free techniques especially attractive for the investigation of dynamics. Probably the most important advantage of NanoSPR is its very high observation bandwidth that allows one to follow molecular dynamics with millisecond time resolution over hours. The different contrast mechanism compared to fluorescence microscopy gives a valuable tool to cross-check existing models and investigate nonfluorescent systems. The fact that NanoSPR helped to identify a discrepancy in protein mobility compared to existing models demonstrates once more the benefit of using complementary techniques on the same system.

■

■

REFERENCES

(1) Willets, K. A.; Van Duyne, R. P. Annu. Rev. Phys. Chem. 2007, 58, 267−297. (2) Ament, I.; Prasad, J.; Henkel, A.; Schmachtel, S.; Sönnichsen, C. Nano Lett. 2012, 12, 1092−1095. (3) Zijlstra, P.; Paulo, P. M. R.; Orrit, M. Nat. Nanotechnol. 2012, 7, 379−382. (4) Rosman, C.; Prasad, J.; Neiser, A.; Henkel, A.; Edgar, J.; Sönnichsen, C. Nano Lett. 2013, 13, 3243−3247. (5) Ahijado-Guzmán, R.; Prasad, J.; Rosman, C.; Henkel, A.; Tome, L.; Schneider, D.; Rivas, G.; Sönnichsen, C. Nano Lett. 2014, 14, 5528−5532. (6) Beuwer, M. A.; Prins, M. W. J.; Zijlstra, P. Nano Lett. 2015, 15, 3507−3511. (7) Martos, A.; Jimenez, M.; Rivas, G.; Schwille, P. Trends Cell Biol. 2012, 22, 634−643. (8) Rothfield, L.; Taghbalout, A.; Shih, Y. L. Nat. Rev. Microbiol. 2005, 3, 959−968. (9) Lutkenhaus, J. Annu. Rev. Biochem. 2007, 76, 539−562. (10) Margolin, W. Curr. Opin. Microbiol. 2001, 4, 647−652. (11) Loose, M.; Kruse, K.; Schwille, P. Annu. Rev. Biophys. 2011, 40, 315−336. (12) Hu, Z. L.; Lutkenhaus, J. Mol. Cell 2001, 7, 1337−1343. (13) Raskin, D. M.; de Boer, P. A. J. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 4971−4976. (14) Hu, Z. L.; Lutkenhaus, J. Mol. Microbiol. 1999, 34, 82−90.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b00507. Methods, statistics on signal offset after addition of the proteins, attempts of protein removal, SDS-Page and mass spectrometry data demonstrating MinD purity, and individual interaction of MinD and MinE with lipid bilayer covered nanoparticles. (PDF) D


Letter

Nano Letters (15) Loose, M.; Fischer-Friedrich, E.; Ries, J.; Kruse, K.; Schwille, P. Science 2008, 320, 789−792. (16) Loose, M.; Fischer-Friedrich, E.; Herold, C.; Kruse, K.; Schwille, P. Nat. Struct. Mol. Biol. 2011, 18, 577−583. (17) Kretschmer, S.; Schwille, P. Life (Basel, Switz.) 2014, 4, 915−28. (18) Giepmans, B. N. G.; Adams, S. R.; Ellisman, M. H.; Tsien, R. Y. Science 2006, 312, 217−224. (19) Shaner, N. C.; Steinbach, P. A.; Tsien, R. Y. Nat. Methods 2005, 2, 905−909. (20) Nusz, G. J.; Curry, A. C.; Marinakos, S. M.; Wax, A.; Chilkoti, A. ACS Nano 2009, 3, 795−806. (21) Baciu, C. L.; Becker, J.; Janshoff, A.; Sönnichsen, C. Nano Lett. 2008, 8, 1724−1728. (22) Martos, A.; Petrasek, Z.; Schwille, P. Environ. Microbiol. 2013, 15, 3319−3326. (23) Schweizer, J.; Loose, M.; Bonny, M.; Kruse, K.; Mönch, I.; Schwille, P. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 15283−15288. (24) Bonny, M.; Fischer-Friedrich, E.; Loose, M.; Schwille, P.; Kruse, K. PLoS Comput. Biol. 2013, 9, e1003347. (25) Schwille, P.; Diez, S. Crit. Rev. Biochem. Mol. Biol. 2009, 44, 223−242. (26) Faiss, S.; Schuy, S.; Weiskopf, D.; Steinem, C.; Janshoff, A. J. Phys. Chem. B 2007, 111, 13979−13986.

E


Single Particle Plasmon Sensors as Label-Free Technique To Monitor

Recommend Documents