Peer reviewed version of the manuscript published in final form at DOI:
10.1039/C4NR04687K
Identification of intracellular gold nanoparticles using surface-enhanced
Raman scattering
Hai-nan Xie,a Yiyang Lin,a Manuel Mazo,a Ciro Chiappini,a Ana Sánchez-Iglesias,b Luis
M. Liz-Marzánbc and Molly M. Stevens*a
* Corresponding authors
a Department of Materials, Department of Bioengineering and Institute for Biomedical
Engineering, Imperial College London, London SW7 2AZ, UK
E-mail: [email protected]
b BioNanoPlasmonics Laboratory, CIC biomaGUNE, Paseo de Miramón 182, 20009
Donostia - San Sebastián, Spain
c Ikerbasque, Basque Foundation for Science, 48011 Bilbao, Spain
Abstract
The identification of intracellular distributions of noble metal nanoparticles is of great utility
for many biomedical applications. We present an effective method to distinguish intracellular
from extracellular nanoparticles by selectively quenching the SERS signals from dye
molecules adsorbed onto star-shaped gold nanoparticles that have not been internalized by
cells.
Rapid developments in the application of nanotechnology are providing increasing insight
into cellular-level biomedical applications. Indeed, various functional nanomaterials have
been extensively used as imaging agents,1–3 drug carriers4 and direct therapeutic sources.5 In
particular, gold nanoparticles (AuNPs) are one of the most widely employed nanomaterials
due to their unique optical properties, biocompatibility and the facile manipulation of their
surface chemistry.6 So far, different biomolecules such as drugs,7 siRNA,8 DNA9 and
proteins10 have been coupled with AuNPs for biomedical applications ranging from in vitro
biosensing to in vivo cancer treatments.6 Significant efforts are directed at the study of the
interactions of gold nanoparticles with cells (e.g., toxicity,11 internalization,12 membrane
penetration13 and endosomal escape8), to increase fundamental understanding and in turn help
to engineer highly effective nanomaterials for diagnostics and therapeutics. During the study
of cell–nanomaterial interactions, being able to elucidate the location of the nanomaterials
within cells is of high relevance toward the interpretation of the results. Transmission
electron microscopy (TEM) is a straightforward method to localize nanoparticles at the
cellular interface. However, imaging cell samples by TEM generally involves a complex
preparation process (e.g., cell fixation, staining and dehydration)14 and a low throughput.
Another commonly used method is confocal fluorescence microscopy which offers spatially
resolved information of fluorophore labelled biomolecules along the cells and provides three
dimensional distributions from the top membrane to the bottom membrane. Although
confocal microscopy is a powerful tool to track fluorescent biomolecules in real time, its
application to metal nanoparticles is greatly restricted since the fluorescence from organic
dyes is likely to be quenched through energy transfer unless the distance between dyes and
nanoparticle surface is carefully adjusted.15 Recently, dark-field microscopy has also been
utilized to follow the internalization of nanoparticles into cells.16 However, it requires
complicated modifications from the commercially available setups. Therefore, a new method
such as the one proposed here, which can discriminate the locations of noble metal
nanoparticles between intracellular and extracellular environments should be of great utility.
Raman scattering spectroscopy is able to provide a molecular fingerprint of the molecules
under study. However, only one in every 106–108 photons involves a change in energy thus
making Raman scattering an intrinsically weak effect.17 In the presence of noble metal
nanostructures such as gold or silver nanoparticles, Raman scattering of molecules that are on
or in close proximity to the nanostructure surface is significantly enhanced,17 a phenomenon
known as the surface-enhanced Raman scattering (SERS). The recent development of SERS
has provided a highly sensitive method for biosensing18 with a detection limit approaching
the single molecule level.19,20 Although the 3D localization of metal nanomaterials in cells
using SERS has been reported, it requires an expensive confocal Raman microscope and the
process is extremely time consuming.21
On the basis of the above considerations, this work aims at providing an alternative strategy
to identify intracellular metal nanoparticles based on SERS. As shown in Fig. 1a, the
principle of this strategy comprises the selective quenching of the SERS signals from
extracellular nanoparticles through a reaction with tris(2-carboxyethyl)phosphine (TCEP), a
reducing agent commonly used in biochemistry and molecular biology.22 Zhuang et al. have
demonstrated that TCEP can effectively quench the fluorescence of the cyanine dye Cy5
through 1,4-addition of the phosphine to the polymethine bridge of Cy5 forming a covalent
adduct.23 It is thus reasonable to anticipate that the perturbation of the conjugation structure
of the fluorophore will not only quench its fluorescence, but will also alter its Raman
scattering spectrum.24 Importantly, TCEP has been shown to be impermeant to phospholipid
bilayers, which are the main components of cell membranes.25 Consequently, this allows us
to selectively quench the Raman signals from molecules adsorbed onto metal nanoparticles
that have not been internalized by cells, while retaining the signal from nanoparticles inside
cells.
Fig. 1 (a) Schematic representation of the selective identification of intracellular noble metal
nanoparticles using SERS. In the presence of TCEP, SERS signals from dyes on nanostars
outside cells are selectively quenched. (b) Extinction spectrum and (c) representative TEM
image of gold nanostars used in this assay. Scale bar is 50 nm.
Optical bioimaging is greatly facilitated by using near infrared (NIR) light. The main reason
is that the penetration depth of photons in biological tissues is limited by the inherent
absorption and scattering by tissue components such as blood and water. These components
have reduced absorption and scattering effects in the NIR region which provides a relatively
transparent optical sensing window for bioimaging.26 Gold nanorods,27 nanocages,28
nanostars29 and hollow shells30 are known to have highly desirable plasmonic properties
within the NIR window. Gold nanostars were selected as SERS substrates for this work, since
they display a strong localized surface plasmon resonance (LSPR) at 795 nm (Fig. 1b,c),
which allows SERS excitation with a 785 nm laser. Superior Raman signals can be obtained
by a combination of highly efficient gold nanostars with the dye Alexa-750, which has both
absorption (Fig. 2a) and emission (Fig. 2b) bands within the NIR region, so that surface
enhanced resonance Raman scattering is registered when a NIR (785 nm) laser excitation is
used.17 This minimizes the required excitation time and the potential interference from cells
and cell medium due to the presence of the dominant Alexa-750 SERS signals.
Fig. 2 Absorption (a) and emission spectra (b) of Alexa-750 upon addition of 10 mM TCEP.
The excitation wavelength for emission was 720 nm. (c) SERS spectra of Alexa-750 after the
addition of 50 mM TCEP. The excitation wavelength was 785 nm and the integration time
was 1 s, with ∼1 mW power at the sample. (d) Intensity of the highlighted peak in (c) versus
time. In (a) and (b), measurements were taken every 2 min. In (a)–(c), the black lines
correspond to the sample at time 0 min.
To verify the reaction between Alexa-750 and TCEP (ESI, Scheme S1†), 5 μM Alexa-750
was reacted with 10 mM TCEP and the addition of the phosphine to the polymethine bridge
of the fluorophore was monitored by UV-Vis (Fig. 2a) and fluorescence spectroscopy (Fig.
2b). Both the absorption and emission bands of Alexa-750 decreased rapidly upon TCEP
addition and completely vanished after 1 hour. The slight blue-shift observed in the emission
band is due to the disruption of the molecule's conjugation. The conjugation of the
fluorophores to gold nanostars was achieved by first forming a cysteamine-Alexa-750
complex and then incubating it with polyvinylpyrrolidone (PVP) capped Au nanostars
overnight (ESI, Experimental†). The free amine group on the unreacted cysteamine was
blocked by carboxyl polyethylene glycol, which also provided greater stability to the
nanostars. The dye-coupling step was verified by successfully observing the strong SERS
signals of Alexa-750 under 785 nm excitation, as shown in Fig. S1, ESI.† In the presence of
50 mM TCEP, the SERS signal of the fluorophore gradually damped over time and was
completely quenched after 1 hour (Fig. 2c and d).
To confirm that TCEP cannot penetrate the cell membrane and the SERS signals from
internalized nanoparticles could be retained, a combination of TEM and SERS imaging was
employed. HeLa cells were incubated with Alexa-750 functionalized gold nanostars (1.2 ×
10−11 M) for 48 hours, followed by TCEP treatment (50 mM) for 1 hour. The extracellular
nanostars were removed by washing with phosphate buffered saline (PBS) for three times
before the standard TEM cell sample preparation.14 The cellular uptake of gold nanostars was
confirmed by TEM (ESI, Fig. S2†). By using an indexed TEM grid, the sample area where
gold nanostars located was identified under Raman microscopy as well as TEM and SERS
signals of Alexa-750 are found (ESI, Fig. S2†). This confirms that TCEP cannot penetrate the
cell membrane and the fluorophores inside cells remained Raman active. It is evidenced from
the above results that TCEP could quench SERS signal from the extracellular nanostars but
not from the intracellular ones. To demonstrate the in vitro applications of the system, after
being incubated with nanostars, cells were fixed with 10% (v/v) formalin on MgF2 slides,
which have negligible Raman background at 785 nm excitation as compared to glass or
quartz (ESI, Fig. S3†). Raman mapping was performed using a 63× water immersion
objective at 785 nm excitation while fixed cells were kept in 1× PBS. Laser power and
integration time not only determine the SERS intensity but may also affect the scanned
nanostructures and attached fluorophores.31 As shown in Fig. S4,† less than 10% signal
reduction was observed after the second scan on the same area when the laser power was
∼0.5 mW with 5 seconds integration time. Therefore, these optimized parameters were used
throughout the whole study. Fig. 3a shows the white light image of the mapped area in which
nanostar aggregates can be identified as a darker area. However, it is impossible to identify
whether the nanostars are inside or outside the cells solely based on this image. Fig. 3b
demonstrates the reconstructed SERS map based on the intensity of the 1300 cm−1 peak of
the highlighted area in Fig. 3a. The signal intensities of the attached fluorophores are revealed
by the brightness level of each pixel. It should be noted that at this stage, nanostars, both
inside and outside the cells, contributed to this reconstructed map. One hour after addition of
50 mM TCEP, a second SERS map was recorded from the same area. Based on the previous
SERS experiments (Fig. 2d, S2†) and also on literature data,25 we find that all the
contributions from nanostars outside the cells were eliminated in Fig. 3c and thus the
remaining signals shown in this map originated only from internalized nanostars. Image
analysis was performed on the basis of the greyscale value changes by the comparison of the
valid pixels between Fig. 3b and c. SERS maps were converted by Photoshop® to greyscale
values, varying from black at the weakest intensity (0) to white at the strongest intensity
(255)32 by measuring changes in pixel brightness. Any grey values above zero in the
converted images were treated as valid pixels. As shown in Fig. 3d, the brightness was
reduced to ∼20% of the original image after the addition of TCEP (the calculation detail is
described in the ESI†). Since TCEP cannot penetrate through the cell membrane, this means
that ∼70% of the nanostars were located outside of the cells considering the 10% reduction
induced by the laser thus providing a clear indication that ∼30% of the nanostar distribution
was intracellular. More than 10 cells were imaged in the same way from different areas of the
cell culture and similar results were obtained. To be noted, aggregation of gold nanostars in
the intracellular environment was observed (Fig. 3a). This further enhanced the SERS signals
from the intracellular nanostars by generating hot spots17 and facilitated the localization of
the particles. It is worth mentioning that this method could distinguish the nanostars
physically binding to the cell membrane from those inside the HeLa cells, which cannot be
achieved by ICP-MS and ordinary SERS mapping. This method is however not intended to
provide the exact locations of individual nanoparticles within cells, but rather a verification of
whether they are intracellular or extracellular. In our experiment, the samples were washed
three times before Raman imaging and the majority of unbound nanostars were removed.
Hence the extracellular nanostars in the samples located mostly on the cell membrane (Fig. 3)
and displayed ‘identical distributions’ to the nanostars inside the cells. By selectively
quenching through TCEP, the SERS signal from the extracellular nanostars was eliminated
while those from inside the cells remained.
Fig. 3 (a) White light image where the mapped area is indicated by a box. (b, c)
Reconstructed SERS map based on the intensities of the highlighted peak in Fig. 2c at 1300
cm−1 before (b) and after (c) the addition of 50 mM TCEP. (d) Greyscale values of the valid
pixel areas before (left) and after (right) the addition of TCEP. 785 nm excitation wavelength,
5 s integration time, ∼0.5 mW power at the sample. Scale bars are 8 μm.
The technique of selectively quenching the SERS signal by TCEP can potentially be applied
to investigate the intactness of the cell membrane. As shown above, TCEP is generally
impermeable to cell membranes. However, TCEP can enter the cell when the membrane is
damaged. Upon entry, we hypothesized that the SERS signals from internalized Raman
probes could be quenched which in turn could be utilized to monitor the perturbation of cell
membrane. To this end, HeLa cells were incubated with Alexa-750 functionalized Au
nanostars and fixed on MgF2 slides. The SERS signals from outside the cells were quenched
upon incubating with TCEP for 1 hour, while the signals from inside the cells were retained
(Fig. 4a). Perturbation of the cell membrane was then triggered by adding 1% v/v Triton X100, which is known to increase the permeability of the cell membrane.33 After 10 minutes,
the HeLa cells were immersed with TCEP for another 1 hour to examine the integrity of cell
membrane. As shown by Raman mapping on the same area (Fig. 4b), more than 95% of
SERS signals are quenched compared to that in Fig. 4a confirming the expected increase in
the permeability of the cell membrane.
Fig. 4 Reconstructed SERS maps based on the intensities of the highlighted peak in Fig. 2c at
1300 cm−1 overlapped with white light images before (a) and after (b) the disruption of cell
membrane with Triton X-100. 785 nm excitation wavelength, 5 s integration time, ∼0.5 mW
power at the sample. Scale bar is 10 μm.
Importantly, this method can be applied utilizing different Raman reporters for localizing
noble metal nanoparticles which was confirmed by using Cy5, which is resonant with 633 nm
but not with 785 nm excitation. The SERS responses of Cy5-functionalized gold nanostars
were assessed before and after addition of TCEP, showing that also in this case the
fluorescence and SERS intensities decreased with time after addition of TCEP (ESI, Fig. S5).
Although gold nanostars were chosen in this study because of their high signal enhancement
efficiency, this method could also be applied with other plasmonic nanostructures with
efficient SERS capability.17
In summary, a general method is presented to localize gold nanoparticles by selectively
quenching the SERS signals originating from dye-conjugated nanoparticles outside cells.
This approach is expected to help understanding actual intracellular nanoparticle
distributions. We also anticipate this localization strategy will provide a means for assessing
the internalization efficiency of various cargos coupled with noble metal nanoparticles, such
as DNA/RNA, proteins, or drugs, which are of major relevance when studying endocytosis.
As demonstrated here, it could also prove very useful as a way of checking cell membrane
integrity.
Acknowledgements
This research was supported by the European Commission FP7 “Self-assembled virus-like
vectors for stem cell phenotyping (SAVVY) project”, project number: 310445.
Notes and references
1. J. Ando, T.-a. Yano, K. Fujita and S. Kawata, Phys. Chem. Chem. Phys., 2013, 15,
13713–13722
2. A. Jakhmola, N. Anton and T. F. Vandamme, Adv. Healthcare Mater., 2012, 1, 413–
431
3. N. Lee, S. H. Choi and T. Hyeon, Adv. Mater., 2013, 25, 2641–2660
4. H. Tian, J. Chen and X. Chen, Small, 2013, 9, 2034–2044
5. B. Van de Broek, N. Devoogdt, A. D'Hollander, H.-L. Gijs, K. Jans, L. Lagae, S.
Muyldermans, G. Maes and G. Borghs, ACS Nano, 2011, 5, 4319–4328
6. K. Saha, S. S. Agasti, C. Kim, X. Li and V. M. Rotello, Chem. Rev., 2012, 112,
2739–2779
7. K. Cho, X. Wang, S. Nie, Z. Chen and D. M. Shin, Clin. Cancer Res., 2008, 14,
1310–1316
8. E. Zhao, Z. Zhao, J. Wang, C. Yang, C. Chen, L. Gao, Q. Feng, W. Hou, M. Gao and
Q. Zhang, Nanoscale, 2012, 4, 5102–5109
9. L. He, M. D. Musick, S. R. Nicewarner, F. G. Salinas, S. J. Benkovic, M. J. Natan and
C. D. Keating, J. Am. Chem. Soc., 2000, 122, 9071–9077
10. C. M. Niemeyer and B. Ceyhan, Angew. Chem., Int. Ed., 2001, 40, 3685–3688
11. C. J. Murphy, A. M. Gole, J. W. Stone, P. N. Sisco, A. M. Alkilany, E. C. Goldsmith
and S. C. Baxter, Acc. Chem. Res., 2008, 41, 1721–1730
12. M. Ahmed, Z. Deng, S. Liu, R. Lafrenie, A. Kumar and R. Narain, Bioconjugate
Chem., 2009, 20, 2169–2176
13. J. Lin, H. Zhang, Z. Chen and Y. Zheng, ACS Nano, 2010, 4, 5421–5429
14. Ž. Krpetić, S. Saleemi, I. A. Prior, V. Sée, R. Qureshi and M. Brust, ACS Nano, 2011,
5, 5195–5201
15. U. H. F. Bunz and V. M. Rotello, Angew. Chem., Int. Ed., 2010, 49, 3268–3279
16. N. Fairbairn, A. Christofidou, A. G. Kanaras, T. A. Newman and O. L. Muskens,
Phys. Chem. Chem. Phys., 2013, 15, 4163–4168
17. K. A. Willets and R. P. Van Duyne, Annu. Rev. Phys. Chem., 2007, 58, 267–297
18. D. Graham and K. Faulds, Chem. Soc. Rev., 2008, 37, 1042–1051
19. S. Nie and S. R. Emory, Science, 1997, 275, 1102–1106
20. K. Kneipp, Y. Wang, H. Kneipp, L. T. Perelman, I. Itzkan, R. R. Dasari and M. S.
Feld, Phys. Rev. Lett., 1997, 78, 1667–1670
21. S. McAughtrie, K. Lau, K. Faulds and D. Graham, Chem. Sci., 2013, 4, 3566–3572
22. D. E. Shafer, J. K. Inman and A. Lees, Anal. Biochem., 2000, 282, 161–164
23. J. C. Vaughan, G. T. Dempsey, E. Sun and X. Zhuang, J. Am. Chem. Soc., 2013, 135,
1197–1200
24. K. Kneipp, H. Kneipp, I. Itzkan, R. R. Dasari and M. S. Feld, Chem. Rev., 1999, 99,
2957–2976
25. D. J. Cline, S. E. Redding, S. G. Brohawn, J. N. Psathas, J. P. Schneider and C.
Thorpe, Biochemistry, 2004, 43, 15195–15203
26. R. Weissleder, Nat. Biotechnol., 2001, 19, 316–317
27. J. Pérez-Juste, I. Pastoriza-Santos, L. M. Liz-Marzán and P. Mulvaney, Coord. Chem.
Rev., 2005, 249, 1870–1901
28. S. E. Skrabalak, J. Chen, Y. Sun, X. Lu, L. Au, C. M. Cobley and Y. Xia, Acc. Chem.
Res., 2008, 41, 1587–1595
29. S. Barbosa, A. Agrawal, L. Rodríguez-Lorenzo, I. Pastoriza-Santos, R. A. AlvarezPuebla, A. Kornowski, H. Weller and L. M. Liz-Marzán, Langmuir, 2010, 26, 14943–
14950
30. H.-n. Xie, I. A. Larmour, Y.-C. Chen, A. W. Wark, V. Tileli, D. W. McComb, K.
Faulds and D. Graham, Nanoscale, 2013, 5, 765–771
31. H.-n. Xie, I. A. Larmour, V. Tileli, A. L. Koh, D. W. McComb, K. Faulds and D.
Graham, J. Phys. Chem. C, 2011, 115, 20515–20522
32. L. Vincent, IEEE Trans. Image Process., 1993, 2, 176–201
33. D. Koley and A. J. Bard, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 16783–16787