Dynamic SERS Imaging of Cellular Transport Pathways with

---

LETTER pubs.acs.org/NanoLett

Dynamic SERS Imaging of Cellular Transport Pathways with Endocytosed Gold Nanoparticles Jun Ando,†,|| Katsumasa Fujita,† Nicholas I. Smith,‡ and Satoshi Kawata*,†,§ †

Department of Applied Physics and ‡Immunology Frontier Research Center, Osaka University, Suita, Osaka 565-0871, Japan § Nanophotonics Laboratory, RIKEN, Wako, Saitama 351-0198, Japan

bS Supporting Information ABSTRACT: Dynamic SERS imaging inside a living cell is demonstrated with the use of a gold nanoparticle, which travels through the intracellular space to probe local molecular information over time. Simultaneous tracking of particle motion and SERS spectroscopy allows us to detect intracellular molecules at 65 nm spatial resolution and 50 ms temporal resolution, providing molecular maps of organelle transport and lisosomal accumulation. Multiplex spectral and trajectory imaging will enable imaging of specific dynamic biological functions such as membrane protein diffusion, nuclear entry, and rearrangement of cellular cytoskeleton. KEYWORDS: Gold nanoparticle, surface-enhanced Raman scattering, plasmonics, Raman imaging, cellular transportation

S

urface-enhanced Raman spectroscopy, called SERS, has widely been used for high-sensitivity detection and identification of analyte molecules1,2 and has been extended to single molecule detection due to its extremely large enhancement.3 Biomedical applications of SERS have been actively investigated such as DNA/RNA detection,4,5 cancer diagnosis,6 and endogenous molecular detection7 by taking full advantage of its high sensitivity and molecular selectivity. The detection volume of SERS is locally confined, because enhancement of electric field for Raman scattering occurs at the vicinity of metallic nanostructure. In particular, tip-enhanced Raman microscopy (TERS) has achieved spatial resolution in the order of several tens of nanometers beyond the diffraction limit of optical microscope8 10 by confining the optical interaction at the tip end. For TERS, however, a metallic nanoprobe needs to scan over the sample surface with scanning probe microscopy, limiting its application to surface imaging. Our method proposed here is an extension of SERS and tipenhanced Raman microscopy to dynamic SERS imaging of a living cell. A gold nanoparticle is captured in a cell and works as a SERS probe, which travels around inside of the cell to analyze the biochemical composition along the cellular transport pathway. To follow the nanoparticle movement while detecting SERS signal, we combined a laser-beam scanning Raman microscope and dark-field microscope with a feedback system to keep the beam centered on the moving particle. In addition to the SERS analysis, we simultaneously recorded the movement of the gold nanoparticle, which characterizes a functional cellular response to the presence of the nanoparticle. The complementary information from SERS and trajectory analysis create molecular r 2011 American Chemical Society

mapping of cellular pathways over time. Compared to other works in particle tracking, this approach provides not only the particle trajectory but also molecular information of the cell through Raman spectroscopy.11 13 We constructed a laser microscope with laser-beam steering mechanism to track the nanoparticle motion. A CW Ti:Sapphire laser (3900S, Spectra Physics) is used for excitation of Raman scattering. The wavelength of the excitation laser was 676 nm, which is slightly far from the plasmon resonance wavelength of 550 nm for 50 nm gold nanoparticle. Even at this wavelength gold nanoparticles give high SERS signals, as shown later, due to the contribution of the interband transition of gold atoms to dielectric function of gold.14 The experimental results with the excitation laser at around 550 nm were dominated by the photoluminescence from the gold nanoparticle due to the plasmon resonance,15,16 which hindered the Raman signal from the molecules. The light was focused to a point in the sample through a water-immersion 1.2 NA objective lens (UPLSPO 60xW, Olympus), and Raman scattered signals were collected with the same objective lens, and finally imaged by an EM-CCD camera (iXon 885, Andor Technology) through the spectrophotometer (MK-300, Bunko Keiki). A halogen lamp illuminates the sample through a dark-field condenser for the darkfield imaging, and scattered light was detected by another EM-CCD camera (QuantEM:512SC, Photometrics). The particle coordinate was obtained from the Received: August 19, 2011 Revised: October 28, 2011 Published: November 07, 2011 5344

dx.doi.org/10.1021/nl202877r | Nano Lett. 2011, 11, 5344–5348

Nano Letters

LETTER

Figure 1. SERS analysis of the cellular pathway with an endocytosed gold nanoparticle. (a) An image of a J774A.1 macrophage cell taken by dark-field microscope. The white arrow indicates a gold nanoparticle seen as a small white spot. The gold nanoparticle is taken up by endocytosis of the cell. (b) SERS spectra, obtained from the nanoparticle indicated in panel a. Characteristic Raman peaks were observed at 977 cm 1 (assigned to the vibration mode of phosphate), 1457 cm 1 (vibration mode of CH2 and CH3), and 1541 cm 1 (vibration mode of Amide II). These three Raman peaks are overlaid with bars in red, green and blue. (c) Trajectory of the nanoparticle, marked by a white arrow in panel a, obtained from the dark-field images. (d) An RGB color map of the molecular distribution displayed on the nanoparticle trajectory. Green spots show the Raman intensity distribution of 1457 cm 1, blue spots 1541 cm 1, and red spots 977 cm 1. The green and blue color is highlighted during the linear paths, while the red color appears during the confined zone random walk. The spatial resolution is determined as ∼65 nm, resulting from the particle diameter ∼50 nm and measurement accuracy ∼15 nm.

dark-field images with an acquisition time of between 50 to 250 ms. The wavelength region of the halogen lamp was set to 400 650 nm, using a bandpass filter (PB0042, Asahi Spectra) to chromatically separate scattered light for dark-field imaging from incident and scattered light for Raman spectroscopy using a shortwave pass dichroic mirror (DMDS0640, Asahi Spectra). The laser spot was scanned using two galvanometer mirrors to direct the focal point to the position of interest, using feedback from the dark-field positioning signal (see Supporting Information). Camera exposures for both Raman spectroscopy and dark-field imaging were synchronized. The cell type studied with our method was J774A.1 macrophage cell, which was cultured on a glass bottom dish in a Dulbecco’s modified eagle’s medium (DMEM) solution. The 50 nm diameter gold nanoparticles (EMGC50, BBI) were added in DMEM solution at a concentration of 2  1010 particles/ml. The gold nanoparticles were not surface functionalized, aside from the citrate coating to prevent aggregation. A macrophage readily uptakes nanoparticles through endocytosis.17 DMEM is replaced with HEPES-buffered Tyrode solution. Figure 1a shows an image of a macrophage cell, observed with a dark-field optical microscope. A gold nanoparticle, seen as a small white spot (marked by white arrow), is endocytosed by the cell and continues to move in the cell after endocytosis. We observed SERS spectra from a moving nanoparticle marked by white arrow in Figure 1a. Figure 1b shows representative Raman spectra obtained during the measurement. In Figure 1b, we found three characteristic Raman peaks at 977, 1457, and 1541 cm 1. The peak at 1457 cm 1, which can be assigned to the vibration mode of CH2 and CH3 common in lipids and proteins,17,18 was observed at 0, 5, and 26.5 s after starting measurement. The peak at 1541 cm 1, which can be assigned to Amide II, present in most proteins,19,20 was observed at 32.00, 39.00, and 43.75 s, while the phosphate vibrational

mode PO32‑ at 977 cm 1 21,22 was seen at 70.25, 116.25, and 137.75 s. These three Raman peaks are marked with bars in green, blue, and red in Figure 1b, respectively. Figure 1c shows the trajectory of the nanoparticle marked by the white arrow in Figure 1a, obtained from the time-resolved dark-field images. Particle center is determined in each frame with centroid detection using Gaussian fitting. In the figure, the position marked (i, 0 s) is the starting point of the nanoparticle movement, and also the measurement of SERS spectroscopy. The particle moved almost linearly from position (i) to (iv, 29.0 s), via positions (ii, 5.0 s) and (iii, 26.5 s). Then, the particle returned back from position (iv, 29.0 s), along the same path via (v, 32.0 s) and (vi, 39.0 s), reaching (vii, 43.75 s), where it began to exhibit a random walk within a confined zone, including (viii, 70.25 s), (ix, 116.25 s) and (x, 137.75 s). Figure 1d shows a color-coded map of the molecular distribution along the nanoparticle trajectory of Figure 1c. A color movie of the particle movement is also provided (see Supporting Information movie). The color of each spot in Figure 1d represents the addition of Raman peak intensities at 977, 1457, and 1541 cm 1, colored in red, green and blue respectively, which rise as SERS-active cell molecules near the nanoparticle increase in number or proximity. The spatial resolution (which corresponds to the spatial range in which the existence of molecules can be detected) is ∼65 nm, resulting from the particle diameter ∼50 nm and measurement accuracy ∼15 nm of the particle position (see Supporting Information). This resolution, which is much shorter than the diffraction limit of light, is necessary in order to observe the transportation of the particle by dynein and kinesin, both ∼60 to 80 nm long,23,24 along the microtubule, and ∼25 nm in diameter.23 The time interval to detect the Raman signal was set to be 250 ms to follow the typical kinesin movement.12,13 5345

dx.doi.org/10.1021/nl202877r |Nano Lett. 2011, 11, 5344–5348

Nano Letters

Figure 2. Analysis of time-resolved SERS spectra obtained with a nanoparticle traveling through a macrophage cell. (a) Raman peaks at 1075 and 1236 cm 1 are continuously observed. The gray bars highlight those peaks. Brown bars highlight 20 peaks that appear only transiently. Temporal resolution of Raman spectroscopy measurement was 50 ms. (b) Schematics showing relative movement of the nanoparticle and molecules. Stationary peaks are due to biomolecules moving together with the nanoparticle, while the peaks appearing transiently relate to molecules that rapidly change distance from the particle. (c) Trajectory of a gold nanoparticle during SERS measurement for panel a. The particle moved linearly during the experiment.

The linear path from positions (i) toward (iv) is occupied by CH2 and CH3 (green spots in Figure 1d), while the return path from (iv) to (vii) shows Amide II (blue spots). Such a directed motion trajectory has been reported with quantum dots and has been explained as functional organelle transport.12,13 It is also known that the different biological schemes are used for transporting organelles with different motor proteins, kinesin and dynein in opposite direction along the microtubule,25 which may also be reflected in our result that shows different directional trajectories with different Raman spectra. When a nanoparticle was confined in (vii) (x), the phosphate (PO32‑) at 977 cm 1 appears dominantly (shown in red in Figure 1d). Phosphate resulting from the digestion and turnover of nucleic acids is found in lysosomes,26 which are known accumulation points of nanoparticles.27 Transmission electron microscopy also revealed the introduction of gold nanoparticles captured in intracellular organelles (see Supporting Information Figure S2). The analysis above on the SERS spectrum and trajectory provides the evidence that the particles enter the lysosomes and then selectively enhances the phosphate band. SERS spectra (Figure 2a) obtained with higher temporal resolution at 50 ms indicates that there are two kinds of nanoparticle biomolecule interaction. Raman peaks at 1075 and 1236 cm 1 marked with gray bars remained over 1.4 s while 20 other peaks appeared only transiently and are marked by brown bars. Stationary peaks indicate that the corresponding biomolecules moved together with the nanoparticle, presumably in contact, while the peaks appearing transiently indicate that the nanoparticle passed by the corresponding biomolecules (Figure 2b). The particle moved approximately linearly during

LETTER

the experiment from 0.0 to 1.4 s (Figure 2c), while it began to exhibit a random walk after 1.4 s with disappearance of peaks at 1075 and 1236 cm 1. In conclusion, Raman intensities at 1075 and 1236 cm 1 are related to molecules used for transport. Three typical trajectories of a nanoparticle endocytosed in a macrophage cell are illustrated in Figures 3a c. The particle shown in Figure 3a undergoes relatively linear motion, while the motion of another nanoparticle (Figure 3b) is well-confined in a limited volume; a less well-confined “random walk” path is also observed (Figure 3c). The peaks in four selected spectra detected during linear motion (Figure 3d) are relatively identical to the averaged spectrum from the linear motion (Figure 3g). Spectra selected from the period of well-confined motion (Figure 3e) also exhibit similarity to the average spectrum for that period (Figure 3h), indicating that nanoparticles carried the molecules or that their movement is confined in the region of certain molecules. In contrast, for the particle trajectory undergoing random walk motion in Figure 3c, selected spectra during the course of the trajectory (Figure 3f) varied with a large deviation so that the averaging is less meaningful in this case (Figure 3i). Combining the trajectory measurement and spectral analysis, the random walk of a nanoparticle in a living cell was found to be governed not by biological function, but rather by thermal energy28 or inhomogeneity of the intracellular space.29,30 A gold nanoparticle works as a nanosized light source (65 nm) in a cell, visits local molecular sites, and reports the enhanced Raman signal near the particle. The spectral changes resolved over 50 ms time-scales shows that for some types of motion the particle movement is governed by cellular functions, such as transport, accumulation and digestion, while for others it appears governed by the physics of the fluid and thermal energy. There is a possibility of covering the gold nanoparticle by phospholipid layers, since the gold nanoparticle is introduced into a cell through endocytosis. We have confirmed through experiment that the detected peaks of Raman spectra do not represent phopholipid layers.31 One of the possible explanations is that the particle moves out of the transportation organelle and therefore lipid molecules were not observed. However, we did observe motion-dependent spectral change, indicating that the change of molecules around the particle causes the change in spectral shape. The particle is transported by some type of cellular function, according to the particle trajectory analysis. Therefore, the SERS spectra will reflect the molecular environmental change during cellular transport, such as the presence or absence of transport proteins and molecules in the organelle. The number of peaks observed in the SERS spectra is quite low, indicating that the numbers of molecules are limited. The nanoparticle may detect only a very limited number of molecules existing on the particle surface due to its highly confined detection area, although there are many molecules existing around the particle, which can also explain why the phospholipid layer is not detected. This type of discussion can be made more quantitative by combining information from particle trajectory analysis and SERS measurement. Our proposed technology will provide us with new insight on the working of dynamic cellular functions in the cell. For future perspectives, the ultimate limits of this new type of imaging should allow time-resolved spectral imaging at less than 10 ms per frame by optimizing the excitation wavelength, particle shape and other considerations.32 35 Even without such optimization, the signal-to-noise in data presented here shows that there 5346

dx.doi.org/10.1021/nl202877r |Nano Lett. 2011, 11, 5344–5348

Nano Letters

LETTER

Figure 3. Three different types of motion by trajectory and SERS analysis. A gold nanoparticle showing linear motion (a), confined motion (b), and random walk motion (c). Each trajectory has 70 dots, obtained with temporal resolution of 50 ms. Typical SERS spectra are shown, obtained during linear motion (d), confined motion (e), and during random walk motion (f). Averaged SERS spectra are shown, obtained during linear motion (g), confined motion (h), and during random walk motion (i). The difference in spectra shows that linear and confined motion are governed by relevant molecules in the cell, while random walk motion appears to be unrelated to specific molecules in the cell.

is sufficient headroom to increase the imaging speed. Additionally, surface functionalization of gold nanoparticles will further extend our analytic capabilities for the investigation of specific biological events and functions.11,19,36 38

’ ASSOCIATED CONTENT

bS

Supporting Information. Experimental details, sample preparation, instrumentation used, and experimental results. Optical setup and procedures for Raman spectroscopy and particle tracking, TEM image of a macrophage cell cultured with gold nanoparticles, and movie showing molecular mapping of a living cell with use of a gold nanoparticle, which is taken up to the interior of the living cell, providing local molecular distribution on the pathway in the cell. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: (+) 81-6-6879-7845. Fax: (+) 81-6-6879-7876. E-mail: [email protected]. )

Present Addresses

Sodeoka Live Cell Chemistry Project, ERATO, Japan Science and Technology Agency, Wako, Saitama 351-0198, Japan.

’ ACKNOWLEDGMENT The authors thank K. Hamada, S. Kawano, H. Niioka, and Y. Yonemaru for technical support. This research was supported

by JST as a CREST project and Osaka University Photonics Advanced Research Center.

’ REFERENCES (1) Albrecht, M. G.; Creighton, J. A. J. Am. Chem. Soc. 1977, 99, 5215–5217. (2) Moskovits, M. Rev. Mod. Phys. 1985, 57, 783–826. (3) Nie, S.; Emory, S. R. Science 1997, 275, 1102–1106. (4) Bell, S. E. J.; Sirimuthu, N. M. S. J. Am. Chem. Soc. 2006, 128, 15580–15581. (5) Barhoumi, A.; Zhang, D.; Tam, F.; Halas, N. J. Am. Chem. Soc. 2008, 130, 5523–5529. (6) Feng, S.; Lin, J.; Cheng, M.; Li, Y.-Z.; Chen, G.; Huang, Z.; Yu, Y.; Chen, R.; Zeng, H. Appl. Spectrosc. 2009, 63, 1089–1094. (7) Kneipp, K.; Haka, A. S.; Kneipp, H.; Badizadegan, K.; Yoshizawa, N.; Boone, C.; Shafer-Peltier, K. E.; Motz, J. T.; Dasari, R. R.; Feld, M. S. Appl. Spectrosc. 2002, 56, 150–154. (8) Hayazawa, N.; Inouye, Y.; Sekkat, Z.; Kawata, S. Opt. Commun. 2000, 183, 333–336. (9) Stockle, R. M.; Suh, Y. D.; Deckert, V.; Zenobi, R. Chem. Phys. Lett. 2000, 318, 131–136. (10) Kawata, S.; Inouye, Y.; Verma, S. Nat. Photonics 2009, 3, 388– 394. (11) Kusumi, A.; Sako, Y.; Yamamoto, M. Biophys. J. 1993, 65, 2021–2040. (12) Courty, S.; Dini, C.; Belliche, Y.; Cappello, G.; Dahan, M. Nano Lett. 2006, 6, 1491–1495. (13) Ruan, G.; Agrawal, A.; Marcus, A. I.; Nie, S. J. Am. Chem. Soc. 2007, 129, 14759–14766. (14) Johnson, P. B.; Christy, R. W. Phys. Rev. B 1972, 6, 4370–4379. (15) Bouhelier, A.; Beversluis, M. R.; Novotny, L. Appl. Phys. Lett. 2003, 83, 5041–5043. 5347

dx.doi.org/10.1021/nl202877r |Nano Lett. 2011, 11, 5344–5348

Nano Letters

LETTER

(16) Geddes, C. D.; Parfenov, A.; Gryczynski, I.; Lakowicz, J. R. Chem. Phys. Lett. 2003, 380, 269–272. (17) Fujita, K.; Ishitobi, S.; Hamada, K.; Smith, N. I.; Taguchi, A.; Inouye, Y.; Kawata, S. J. Biomed. Opt. 2009, 14, 024038. (18) Tang, H.-W.; Yang, X. B.; Kirkham, J.; Smith, D. A. Anal. Chem. 2007, 79, 3646–3653. (19) Xie, W.; Wang, L.; Zhang, Y.; Su, L.; Shen, A.; Tan, J.; Hu, J. Bioconjugate Chem. 2009, 20, 768–773. (20) Willets, K. A. Anal. Bioanal. Chem. 2009, 394, 85–94. (21) Tu, A. T. Raman Spectroscopy in Biology; J. Wiley and Sons: New York, 1982. (22) Twardowski, J.; Nowak, I.; Stufkens, D. J.; Snoeck, T. L. Biochim. Biophys. Acta 1984, 790, 70–77. (23) Hirosawa, N. Science 1998, 279, 519–526. (24) Kolomeisky, A. B.; Fisher, M. E. Annu. Rev. Phys. Chem. 2007, 58, 675–695. (25) M€uller, M. J. I.; Klumpp, S.; Lipowsky, R. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 4609–4614. (26) Pisoni, R. L. J. Biol. Chem. 1991, 266, 979–985. (27) Shi, H.; He, X.; Yuan, Y.; Wang, K.; Liu, D. Anal. Chem. 2010, 82, 2213–2220. (28) Einstein, A. Theory of the Brownian Movement; Furth, R., Ed., Dover: Mineola, New York, 1956. (29) Van Kampen, N. G. J. Phys. Chem. Solids 1988, 49, 673–677. (30) Siggia, E. D.; Lippincott-Schwartz, J.; Bekiranov, S. Biophys. J. 2000, 79, 1761–1770. (31) Kundu, J.; Levin, C. S.; Halas, N. Nanoscale 2009, 1, 114–117. (32) Jackson, J. B.; Westcott, S. L.; Hirsch, L. R.; West, J. L.; Halas, N. J. Appl. Phys. Lett. 2003, 82, 257–259. (33) McFarland, A. D.; Young, M. A.; Dieringer, J. A.; Van Duyne, R. P. J. Phys. Chem. B 2005, 109, 11279–11285. (34) Nehl, C. L.; Liao, H.; Hafner, J. H. Nano Lett. 2006, 6, 683–688. (35) Lin, J. Y.; Hansan, W.; Yang, J.-C.; Odom, T. W. J. Phys. Chem. C 2010, 114, 7432–7435. (36) Kumar, S.; Harrison, N.; Richards-Kortum, R.; Sokolov, K. Nano Lett. 2007, 7, 1338–1343. (37) Brown, S. D.; Nativo, P.; Smith, J.-A.; Stirling, D.; Edwards, P. R.; Venugopal, B.; Flint, D. J.; Plumb, J. A.; Graham, D.; Wheate, N. J. J. Am. Chem. Soc. 2010, 132, 4678–4684. (38) El-Sayed, I. H.; Huang, X.; El-Sayed, M. A. Nano Lett. 2005, 5, 829–834.

5348

dx.doi.org/10.1021/nl202877r |Nano Lett. 2011, 11, 5344–5348