2734
J. Phys. Chem. B 2008, 112, 2734-2737
Fluorescent II-VI Semiconductor Quantum Dots in Living Cells: Nonlinear Microspectroscopy in an Optical Tweezers System Patricia M. A. Farias,*,† Beate S. Santos,‡ Andre´ A. de Thomaz,§ Ricardo Ferreira,| Frederico D. Menezes,| Carlos L. Cesar,§ and Adriana Fontes*,† Departamento de Biofı´sica e Radiobiologia, UFPE, Recife, PE, Brazil, Departamento de Cieˆ ncias Farmaceˆ uticas, UFPE, Recife, PE, Brazil, Instituto de Fı´sica Gleb Wataghin, UNICAMP, Campinas, SP, Brazil, and Departamento de Quı´mica Fundamental, UFPE, Recife, PE, Brazil ReceiVed: July 25, 2007; In Final Form: December 12, 2007
In this work we used a setup consisting of an optical tweezers combined with a nonlinear microspectroscopy system to perform scanning microscopy and obtain emission spectra using two photon excited (TPE) luminescence of captured single living cells labeled with core-shell fluorescent semiconductor quantum dots (QDs). The QDs were obtained via colloidal synthesis in aqueous medium with an adequate physiological resulting pH. Sodium polyphosphate was used as the stabilizing agent. The results obtained show the potential presented by this system as well as by these II-VI fluorescent semiconductor quantum dots to perform spectroscopy in living trapped cells in any neighborhood and dynamically observe the cell chemical reactions in real time.
Introduction Luminescence provides a unique method for the investigation of basic physical properties of biological structures. The high sensitivity of fluorescence, combined with the advances in measurement techniques, permits detection of ultrasmall quantities of specific species present in the biological system. There are several different compounds commonly used to generate fluorescence, such as organic molecules and fluorescent proteins. However, all of these fluorophores present one or more of the following disadvantages: lack of brightness; broad emission bands; high photobleaching rate. Due to these facts, in the past decade a new class of fluorescent materials known as quantum dots (QDs) has been tested as biolabels. Quantum dots are nanometric inorganic crystals, which present special characteristics due to the fact that they are in quantum confinement regime.1-3 In the case of semiconductor quantum dots, one of these special characteristics is the capability of tuning their optical properties, particularly their emission spectra,4 by controlling the size of the particles. The first biological applications of quantum dots were reported in 1998 by Bruchez et al.5 and Chan at al.6 using CdSe QDs coated with silica and mercaptoacetic acid layers, respectively; both groups showed specific labeling by covalent coupling of ligands to the quantum dots surfaces. Subsequently, several authors have reported labeling of whole cells and tissue sections using several different surface modifications of QDs.7-11 By the attachment of biomolecules to nanometersized bits of semiconductors, a sensitive and potentially widely applicable method for detecting biomolecules and for scrutinizing biomolecular processes was developed. The quantum-dot-labeled molecules remain active for biochemical reactions, and the tagged species produce bright * To whom correspondence should be addressed. E-mail:
[email protected] (P.M.A.F.);
[email protected] (A.F.). Tel.: +55-81-21268535 (P.M.A.F.). Fax: +55- 81-21268560 (P.M.A.F.). † Departamento de Biofı´sica e Radiobiologia, UFPE. ‡ Departamento de Cie ˆ ncias Farmaceˆuticas, UFPE. § Instituto de Fı´sica Gleb Wataghin, UNICAMP. | Departamento de Quı´mica Fundamental, UFPE.
images.5,6 This methodology takes advantage of the efficient fluorescence and high photostability of the semiconductor dots, representing a new class of biological stains. The synthesis, passivation, and stabilization steps to obtain core-shell fluorescent semiconductor quantum dots, prepared for biological labeling, are generally succeeded by a functionalization procedure. The functionalizing agents may be small organic molecules (e.g., thiol-containing molecules) or larger biomolecules (e.g., proteins, nucleic acids, and polyethylene glycol). The specificity of the functionalizer depends on its nature. QDs can be functionalized to target particular agents like peptides and antigens of specific living cells for in vivo and/or in vitro experiments.12 The combined advantages of this new class of fluorescent biolabels allow its application in novel detection techniques that applied to biological samples enable long-term experiments. For example, Dahan et al. used the sustained strong signal of QDs to track dynamic cellular processes over time scales that are not viable using organic fluorophores.13 On the other hand, to detect fluorescent signals and to obtain spectroscopic data in biological systems it is necessary to couple these materials to microscopic detection setups. Recently, it has been described by some of us14 a new experimental setup that combines spectroscopic techniques with an optical tweezers system to analyze fluorescence induced by two photon excitation (TPE) and to observe Raman and hyper-Rayleigh scattering. In our work, we apply this setup to perform scanning microscopy and observe spectra using TPE luminescence of trapped single living cells (macrophages) conjugated with core-shell quantum dots of CdS/Cd(OH)2 and CdTe/CdS. Experimental Methods The lasers used in the microspectroscopy optical tweezers setup are a CW Ti:Sapphire laser (700-900 nm, model 3900S, Spectra Physics), a femtosecond Ti:Sapphire laser (700-900 nm, Tsunami, Spectra Physics), and a Nd:YAG laser (1064 nm, model 3800S, Spectra Physics). The CW Ti:Sapphire laser was used for the optical tweezers and the Raman excitation. The
10.1021/jp0758465 CCC: $40.75 © 2008 American Chemical Society Published on Web 02/08/2008
Fluorescent II-VI Semiconductor Quantum Dots
Figure 1. Schematic representation of the complete system for microspectroscopy (Raman, hyper-Raman, hyper-Rayleigh, and twophoton excited luminescence) combined with optical tweezers. BS1 is a metallic beam splitter, BS2 is a dichroic mirror of Nd:YAG, BP is the band-pass filter for Raman, SP is the short pass filter for hyperRayleigh and hyper-Raman, T is the telescope, and M is the monochromator.
femtosecond laser was used to the TPE luminescence measurements. Both lasers were operating at 785 nm. The Nd:YAG laser was used for the optical tweezers. The laser beams were focused through a 100× oil immersion (SPlan) or 60× (Apochromatic) objective in an Olympus microscope (BH2, Olympus Optical Co., Ltd.). The quantum dots images were recorded by a camera (TK1085-U, JVC-Victor Co. of Japan, Ltd.). We used a translation motorized stage (Prior Scientific, model ProScan) to control the x, y, and z microscope movements, using a joystick or via computer software. The signals were collected in a backscattering geometry using the microscope objective. The collected signals were sent to a 30 cm monochromator (Acton Research, model 300i) equipped with a back-illuminated refrigerated CCD (Princeton Instruments LNCCD 1340/100 EB/ 1), as shown in Figure 1. The metallic beam-splitter BS1 reflectivity is about the same for all lasers of the system. A super notch filter (NF Kaiser Optics Super Notch Plus with 350 cm-1 rejection bandwidth) was used as a mirror to reflect both the femtosecond and the CW Ti:Sapphire laser beams to the microscope and to act as a notch filter for the backscattered light. The Nd:YAG laser beam is 100% transmitted to the microscope and rejected to the CCD by using a dichroic mirror (BS2). A set of telescopes is part of the system to capture and to generate the signal on the microscope focal plane. The hyperRaman, hyper-Rayleigh, and TPE fluorescence systems are confocal by themselves, since all of them employ the absorption of two photons, which only occurs at the laser focus which corresponds to the point of maximum intensity. The backscattered Raman signal spectra were collected in confocal configuration. We used a band-pass filter (BP-Omega Filters) to decrease the spectral width of the laser beam and another notch filter (NF) in front of the monochromator to avoid Rayleigh scattering signal. For the hyper-Raman, hyper-Rayleigh, and TPE luminescence experiments, there was no need to use either the band-pass or the additional notch filter. We used only a short pass color filter (SP-Newport), which transmits the visible light and cuts the emission at the infrared region. The images were obtained by using a translation motorized stage (Prior Scientific) and a CCD coupled to the monochromator. First we acquired the emission spectrum of the sample to choose the optimized region for constructing the images. After that, we set the CCD to integrate the intensities along this region and converted the intensities in megapixel units. These procedures were performed by scanning the labeled cells point by point, by moving the translation stage in steps of 1 µm. Therefore the megapixels obtained from TPE luminescence for each point were grouped to construct the images, by using a
J. Phys. Chem. B, Vol. 112, No. 9, 2008 2735 script build in the Mathematica software (Wolfram Research). The detection of spectra and the concomitant movement of the stage were done using a script build in LabView (National Instruments). The CdS/Cd(OH)2 quantum dots were obtained via colloidal synthesis in aqueous medium using the methodology described in a previous work.15 CdS suspensions were obtained by adding H2S into an aqueous solution containing Cd2+ ions and sodium polyphosphate as the stabilizing agent. After the formation of the particles (10 min), a layer of cadmium hydroxide was deposited on their surface. At this point the suspensions show a strong green emission. The particles were functionalized with glutaraldehyde molecules in a pH ) 7. The CdTe nanocrystals are obtained, under argon flow, by reacting Na2Te (obtained by the controlled reduction of Te(s) and NaBH4 in aqueous medium) with Cd(ClO4)2 in the presence of thiol-binding molecules, such as mercaptoacetic acid (AMA), at 80 °C. The particle’s size growth is controlled by a posterior heat treatment at 80 °C, and colloidal size selection procedures are performed to define smaller size dispersion. The CdS surface passivation layer presents a slow kinetic rate and results from the hydrolysis of the thiol group of the AMA molecules that are linked to the dangling bonds of the CdTe nanocrystals. The CdTe/CdS colloidal suspensions were buffered to a pH ) 7 prior to their conjugation with the living cells. The macrophages at the concentration of 7 × 105 cells/mL were diluted (1:2 µL) in the suspensions of functionalized quantum dots and incubated for 14 h at 10 °C. Murine peritoneal macrophages were collected from mice Balb/c and cultivated in RPMI 1640 (Sigma) culture medium complemented with antibiotic and 10% fetal bovine serum at 37 °C, into an incubator with 95% air and 5% CO2. Results and Discussion In a preliminary phase, we tested the capability of our homemade system on known samples to compare our results with the ones described in the literature.16-18 A set of Raman and TPE luminescence spectra were acquired for a range of samples such as bulk silicon, TiO2 (anatase) microspheres, ZnSe microparticles, single trapped dyed 2.5 and 6 µm microspheres (Polysciences and Molecular Probes), and a single captured red blood cell.16 The microspheres and the ZnSe microparticles were diluted in water and placed in a Neubauer chamber. Figure 2a shows the Raman spectrum for bulk silicon, and Figure 2b shows TPE luminescence for a ZnSe microparticle both obtained with 1 s acquisition time. After testing the performance of the whole system, we obtained the hyper-Rayleigh and TPE luminescence spectra for trapped samples (using the optical tweezers) and images for immobilized samples (Figures 3-6). Figure 3a shows the TPE luminescence spectrum for a single trapped stained 10 µm microsphere (Polysciences) showing emission in the green region, and Figure 3b shows the image obtained for the same microsphere using the experimental setup. In Figure 4a the hyper-Rayleigh spectrum for a ZnSe microparticle is depicted, while Figure 4b shows its respective image. Figure 5 shows the spectrum (a) and the image (b) of living macrophages conjugated with CdS/Cd(OH)2 quantum dots, while Figure 6 shows the same set of data for macrophages conjugated with CdTe/CdS quantum dots. The spectra were obtained using living trapped cells. For all images, the luminescent CdS/Cd(OH)2 quantum dots were functionalized with a glutaraldehyde solution (0.01% v/v). For CdTe/CdS quantum dots, the organic functionalizing agent was mercaptoacetic acid (AMA).
2736 J. Phys. Chem. B, Vol. 112, No. 9, 2008
Farias et al.
Figure 2. (a) Raman spectrum of bulk silicon. (b) TPE luminescence spectrum of a single ZnSe microparticle. Both spectra were acquired under 785 nm excitation.
Figure 3. (a) TPE luminescence spectrum of a single trapped stained microsphere. (b) Image obtained from TPE luminescence of a single microsphere. The spectra were acquired under 785 nm excitation.
Figure 4. Single ZnSe microparticle: (a) hyper-Rayleigh TPE emission spectrum; (b) respective hyper-Rayleigh image. The spectra were acquired under 785 nm excitation.
Figure 5. Macrophage labeled with CdTe/CdS quantum dots: (a) TPE luminescence spectrum; (b) image obtained from TPE luminescence. The spectra were acquired under 785 nm excitation.
The living cells continued alive for hours after the incubation with the quantum dots and continued to perform their biological functions such as endocytosis even after several days, showing the low level of cytotoxicity of these materials. The cell viability was monitored continuously using a conventional optical microscope. The efficiency of the quantum dots conjugation to the cells was also monitored by conventional fluorescence as well as by laser scanning confocal microscopies (Figure 7a,b). All the results show that the images of the macrophages obtained using the TPE fluorescence setup are in good agree-
ment with the images obtained using confocal scanning microscopy. The overall fluorescence analysis of the images shows a homogeneous marking pattern for both QDs. This was expected due to the low specificity of the functionalizing agents used in the labeling. Prior to incorporation to the biological systems the original CdS/Cd(OH)2 and CdTe/CdS emission spectra show a narrow bandwidth (fmhw ∼ 50 nm) with maxima ranging from 480 to 520 nm depending on the particle size and nature. Analyzing the TPE emission spectra of the marked macrophages, we observe a very rich and large band profile.
Fluorescent II-VI Semiconductor Quantum Dots
J. Phys. Chem. B, Vol. 112, No. 9, 2008 2737
Figure 6. Macrophage labeled with CdS/Cd(OH)2 quantum dots: (a) TPE luminescence spectrum; (b) image obtained from TPE luminescence. The spectra were acquired under 785 nm excitation.
Figure 7. Confocal microscopy images of macrophages marked with (a) CdS/Cd(OH)2 and (b) CdTe/CdS (fluorescence excited at 543 nm combined with differential interface contrast).
This modification suggests deactivation processes different from the previously observed exciton-hole recombination process, and these may include energy transfer from the nanoparticles to the biological moiety and trap emission due to different defects in the passivation layer (most probably, some of these defects may be induced by the chemical changes resulting from the QDs/biological system interaction). A more detailed explanation concerning the QDs/biological system interactions is beyond the scope of this work and will be discussed elsewhere. What is evident with these results is the possibility not only of constructing images but also of real time monitoring the spectral changes in trapped single cells. The possibility of associating a targeting molecule to the QDs can be also easily used to study specific biological components in the single cell using the proposed experimental setup. Concluding Remarks We presented a setup of optical tweezers combined with linear and nonlinear microspectroscopy that is capable to measure Raman, hyper-Raman, hyper-Rayleigh, and luminescence signals, excited by two-photon absorption. We were able to obtain a variety of spectra and images acquired from hyper-Rayleigh and TPE luminescence of samples trapped and not trapped by optical tweezers. We also were able to obtain spectra and images of macrophage living cells, labeled with colloidal aqueous CdTe/ CdS and CdS/Cd(OH)2 quantum dots synthesized by our group. The results obtained show the potential presented by this system as well as by the quantum dots as fluorescent labels to perform spectroscopy in a living trapped cell in any neighborhood and dynamically observe cell chemical reactions in real time. These spectroscopic techniques coupled to the optical tweezers in the same system can provide a thorough analysis from visible to infrared regions of all kinds of living cells and other single trapped particles. Acknowledgment. We thank Domingo Scordo, M.D., Ph.D., HISTO-LABO S.R.L., Buenos Aires, Argentina, for helpful
discussions and the Brazilian Research Support Agencies Capes, CNPq, Facepe, Renami, and Fapesp. This work is also linked to the CEPOF (Optics and Photonics Research Center, FAPESP). References and Notes (1) Borchert, H.; Talapin, D. V.; McGinley, C.; Adam, S.; Lobo, A.; de Castro, A. R. B.; Moller, T.; Weller, H. J. Chem. Phys. 2003, 119, 18001807. (2) Alivisatos, A. P. Science 1996, 271, 933-937. (3) Nowak, C.; Do¨llefeld, H.; Eychmu¨ller, A.; Friedrich, J.; Kolmakov, A.; Lofken, J. O.; Riedler, M.; Wark, A.; Weller, H.; Wolff, M.; Moller, T. J. Chem. Phys. 2001, 114, 489-494. (4) Eychmuller, A. J. Phys Chem. B 1996, 59, 13226-13239. (5) Bruchez, M. P.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013-2016. (6) Chan, W. C. W.; Nie, S. Science 1998, 281, 2016-2018. (7) Ballou, B.; Lagerholm, B. C.; Ernst, L. A.; Bruchez, M. P.; Waggoner, A. S. Bioconjugate Chem. 2004, 15, 79-86. (8) Gao, X.; Nie, S. Trends Biotechnol. 2003, 21, 371-373. (9) A ¨ kerman, M. E.; Chan, W. C. W.; Laakkonen, P.; Bhatia, S. N.; Ruoslahti, E. Appl. Biol. Sci. 2002, 99, 12617-12621. (10) Chan, W. C. W.; Maxwell, D. J.; Gao, X.; Bailey, R. E.; Han, M.; Shuming, N. Curr. Opin. Biotechnol. 2002, 13, 40-46. (11) Mattheakis, L. C.; Dias, J. M.; Choi, Y.; Gong, J.; Bruchez, M. P.; Liu, J.; Wang, E. Anal. Biochem. 2004, 327, 200-208. (12) Medintz, I. L.; Uyeda, H. T.; Goldman, E. R.; Mattoussi, H. Nat. Mater. 2005, 4, 435. (13) Dahan, M.; Levi, S.; Luccardini, C.; Rostaing, P.; Riveau, B.; Triller, A. Science 2003, 302, 442. (14) Fontes, A.; Ajito, K.; Neves, A. A. R.; Moreira, W. L.; Thomaz, A. A.; Barbosa, L. C.; de Paula, A. M.; Cesar, C. L. Phys. ReV. E 2005, 72, 12903. (15) Petrov, D. V.; Santos, B. S.; Pereira, G. A. L.; Donega´, C. M. J. Phys. Chem. B 2002, 106, 5325-5334. (16) Ajito, K.; Torimitsu, K. Appl. Spectrosc. 2002, 56, 541-544. (17) Xie, C. G.; Dinno, M. A.; Li, Y. Q. Opt. Lett. 2002, 27, 249-251. (18) Wood, B. R.; Tait, B.; McNaughton, D. Biochim. Biophys. Acta 2001, 1539, 58-70.
