Protein-Mediated Sandwich Strategy for Surface-Enhanced Raman

Anal. Chem. 2009, 81, 3350–3355

Protein-Mediated Sandwich Strategy for Surface-Enhanced Raman Scattering: Application to Versatile Protein Detection Xiao X. Han,†,‡ Yasutaka Kitahama,‡ Tamitake Itoh,§ Chun X. Wang,| Bing Zhao,*,† and Yukihiro Ozaki*,‡ State Key Laboratory of Supramolecular Structure and Materials, Jilin University, Changchun 130012, P. R. China, Department of Chemistry, School of Science and Technology, Kwansei Gakuin University, Sanda, Hyogo 669-1337, Japan, Nano-Bioanalysis Team, Health Technology Research Center, National Institute of Advanced Industrial Science and Technology (AIST), 2217-14 Hayashi-cho, Takamatsu, Kagawa 761-0395, Japan, and State Key Laboratory on Integrated Optoelectronics, Jilin University, Changchun 130012, P. R. China For surface-enhanced Raman scattering (SERS)-based protein identification, immunoassay, and drug screening, metal sandwich substrates bridged by proteins have been created in the present study. The sandwich architectures are fabricated based on a layer-by-layer (LbL) technique. The first gold monolayer is prepared by the self-assembling of gold nanoparticles on a poly(diallyldimethylammonium chloride) (PDDA)-coated glass slide. The second gold or silver layer is produced by the interactions between proteins in the middle layer of the sandwich architecture and the metal nanoparticles. Highly reproducible surface-enhanced resonance Raman scattering (SERRS) and SERS spectra can be obtained by the present gold-protein-gold (Au/Au) and gold-protein-silver (Au/Ag) sandwiches, and we find that the latter yields about 7 times stronger SERRS than the former. Because of contributions from the two metal layers to the SERS, this sandwich strategy holds great potential in highly sensitive and reproducible protein detections.

indispensable to develop new detection methods for highthroughput protein analysis. More and more studies have proved the great potential of SERS in protein identification and detection of protein-ligand interactions.6-12 Colloidal gold, with the feature of biocompatibility and homogeneity, has been safely applied to biomedical research, for example, therapy for rheumatoid arthritis and as a drug carrier.13,14 As an excellent substrate of SERS, colloidal gold has recently been used for in vivo targeting of tumors15 and many gold nanoparticle probes have successfully been applied to SERS-based biosensors.8-10 Many SERS studies have revealed that SERS spectra from gold substrates are much better in stability than those from silver substrates,16 while the SERS enhancement ability of gold substrates is much weaker than that of silver ones.17 As an ultrasensitive analytical tool, SERS and surface-enhanced resonance Raman scattering (SERRS) have been used as readout methods for DNA and RNA detection,3 protein-DNA interactions,18 and immunoassays.8-10 All these belong to labeled methods, which detect receptors or ligands by SERS/SERRS of

Surface-enhanced Raman scattering (SERS)-based methods for biomolecules show great advantages over fluorescence-based ones in terms of photostability and spectral multiplexing.1-4 Intense enough SERS emissions could be recorded under favorable circumstances to detect single molecules, and the quest for highsensitivity biomolecular-sensing platforms have returned SERS to the front burner.5 With the development of proteomics, it is

(6) Grabbe, E. S.; Buck, R. P. J. Am. Chem. Soc. 1989, 111, 8362–8366. (7) Drachev, V. P.; Thoreson, M. D.; Khaliullin, E. N.; Davisson, V. J.; Shalaev, V. M. J. Phys. Chem. B 2004, 108, 18046–18052. (8) Cao, Y. C.; Jin, R.; Nam, J. M.; Thaxton, C. S.; Mirkin, C. A. J. Am. Chem. Soc. 2003, 125, 14676–14677. (9) Xu, S. P.; Ji, X. H.; Xu, W. Q.; Li, X. L.; Wang, L. Y.; Bai, Y. B.; Zhao, B.; Ozaki, Y. Analyst 2004, 129, 63–68. (10) Driskell, J. D.; Uhlenkamp, J. M.; Lipert, R. J.; Porter, M. D. Anal. Chem. 2007, 79, 4141–4148. (11) Mulvaney, S. P.; Musick, M. D.; Keating, C. D.; Natan, M. J. Langmuir 2003, 19, 4784–4790. (12) (a) Han, X. X.; Jia, H. Y.; Wang, Y. F.; Lu, Z. C.; Wang, C. X.; Xu, W. Q.; Zhao, B.; Ozaki, Y. Anal. Chem. 2008, 80, 2799–2804. (b) Han, X. X.; Cai, L. J.; Guo, J.; Wang, C. X.; Ruan, W. D.; Han, W. Y.; Xu, W. Q.; Zhao, B.; Ozaki, Y. Anal. Chem. 2008, 80, 3020–3024. (13) Tsai, C. Y.; Shiau, A. L.; Chen, S. Y.; et al. Arthritis Rheum. 2007, 56, 544– 554. (14) Gibson, J. D.; Khanal, B. P.; Zubarev, E. R. J. Am. Chem. Soc. 2007, 129, 11653–11661. (15) Qian, X. M.; Peng, X. H.; Ansari1, D. O.; Yin-Goen, Q.; Chen, G. Z.; Shin, D. M.; Yang, L.; Young, A. N.; Wang, M. D.; Nie, S. M. Nat. Biotechnol. 2008, 26, 83–90. (16) Keating, C. D.; Kovaleski, K. M.; Natan, M. J. J. Phys. Chem. B 1998, 102, 9404–9413. (17) Zeman, E. J.; Schatz, G. C. J. Phys. Chem. 1987, 91, 634–643. (18) Bonham, A. J.; Braun, G.; Pavel, I.; Moskovits, M.; Reich, N. O. J. Am. Chem. Soc. 2007, 129, 14572–14573.

* To whom correspondence should be addressed. E-mail: [email protected] (B.Z.); [email protected] (Y.O.). † State Key Laboratory of Supramolecular Structure and Materials, Jilin University. ‡ Kwansei Gakuin University. § National Institute of Advanced Industrial Science and Technology (AIST). | State Key Laboratory on Integrated Optoelectronics, Jilin University. (1) Moskovits, M. Rev. Mod. Phys. 1985, 57, 783–825. (2) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. Rev. 1999, 99, 2957–2975. (3) Cao, Y. W. C.; Jin, R. C.; Mirkin, C. A. Science 2002, 297, 1536–1540. (4) Aroca, R. Surface-Enhanced Vibrational Spectroscopy; John Wiley & Sons Ltd.: Chichester, U.K., 2006; pp 141-176. (5) Kneipp, K., Moskovits, M., Kneipp, H. Eds.; Surface-Enhanced Raman ScatteringsPhysics and Applications; Springer: Heidelberg and Berlin, Germany; 2006.

3350

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

10.1021/ac802553a CCC: $40.75  2009 American Chemical Society Published on Web 04/10/2009

Raman dyes linked to gold nanoparticles. The protocol of silver staining conducted on gold nanoparticles is commonly used for Raman enhancement. However, the complex procedure of Au probe synthesis (i.e., ligand and dye linking, blocking) restricted its application to practical chip-based protein detection. In our previous studies, it was found that colloidal silver staining for proteins was able to produce significant SERS effect for the detections of label-free proteins and fluorescein isothiocyanate (FITC)-labeled immunoassays.12 Strong covalent and noncovalent interactions between silver nanoparticles and proteins can result in aggregation of silver nanoparticles, which make it possible to emerge the SERS effect. This protein-based method for formation of SERS-active substrates has great advantages of simplicity and high sensitivity for protein detection. As a more advanced system, we develop the novel type of SERRS-active sandwich substrates bridged by the target proteins for the detection of these proteins. The most significant advantage of the present method over other SERS-based methods for protein detections is that a self-assembling gold nanoparticle monolayer is used for both capturing proteins and forming the SERS-active substrate with the second metal layer, making the present sandwich substrate more accessible and sensitive than other gold based SERS methods and more reproducible than other silverbased SERS measurements. Moreover, besides labeled immunoassays, the present method can identify label-free proteins and protein-drug interactions without Raman dyes. With the use of a layer-by-layer (LbL) technique, gold-proteingold (Au/Au) and gold-protein-silver (Au/Ag) sandwiches for SERS have been fabricated by target proteins, gold, and silver nanoparticles. The first gold monolayer film is assembled by poly(diallyldimethylammonium chloride) (PDDA), and gold or silver nanoparticles of the second layer adsorb on proteins in the middle layer by protein-based colloidal metal staining. Enormous SERS “hot spots” may emerge among metal aggregates19 formed on the edges between the two metal layers and also between the junctions of metal nanoparticles in the second metal layer. Highquality SERRS/SERS spectra with both high sensitivity and reproducibility are obtained, which are crucial for an analytical method. With the use of these metal sandwich substrates, SERSbased versatile protein detections (i.e., cytochrome c identification, tetramethylrhodamine isothiocyanate (TRITC)-labeled immunoassays, and pipemidic acid/biotin screening) have been effectively carried out. About 7 times stronger SERRS spectra can be obtained from the Au/Ag sandwiches as compared to those from the Au/ Au sandwiches, and the detection limits of the present method for biotin is as low as 10 pg/mL. EXPERIMENTAL SECTION Biochemicals and Chemicals. Cytochrome c (C2037), avidin (A9275), human IgG (I4506), TRITC-antihuman IgG (Fc specific, T4530), Atto610-biotin (43292), pipemidic acid (P7903), and HAuCl4 · 3H2O were all purchased from Sigma-Aldrich Co., Ltd. and used without further purification. Poly(diallyldimethylammonium chloride) (PDDA) with medium molecular weight (200 000s350 000) was purchased from Aldrich Chemical Co., Inc. Bovine serum albumin (BSA), silver nitrate, trisodium (19) Jiang, J.; Bosnick, K.; Maillard, M.; Brus, L. J. Phys. Chem. B 2003, 107, 9964–9972.

citrate, and all other chemicals were purchased from Wako Co., Ltd. Triply distilled water was used throughout the present study. Buffers. The phosphate buffered saline (PBS; 0.01 M, pH 7.2) used in this study contained 0.8% NaCl, 0.02% KH2PO4, 0.02% KCl, and 0.12% Na2HPO4 · 12H2O. A blocking buffer was prepared by dissolving BSA in the PBS buffer (containing 1% BSA). A washing buffer was prepared by adding Tween-20 to the PBS buffer (containing 0.05% Tween 20). Preparation of Gold and Silver Colloid. Colloidal gold was prepared as follows: 1 mL of a HAuCl4 solution (1%, w/v) was added to 99 mL of triply distilled water. After boiling, 1.8 mL of a trisodium citrate solution (1%, w/v) was added and then kept boiling for 15 min.20 The gold colloid thus obtained showed an absorption maximum at 519 nm with an average diameter of 20 nm. Colloidal silver was prepared by the aqueous reduction of silver nitrate (10-3 M, 200 mL) with trisodium citrate (1%, 4 mL) using the method of Lee and Meisel.21 The plasmon absorption maximum of the silver colloid we prepared was located at 415 nm with an average diameter of 40 nm. UV-vis spectra are recorded on a UV-vis-NIR scanning spectrophotometer (UV-3101PC, Shimadzu Corporation, Japan). Preparation of Gold Monolayer Film. Glass slides were cleaned by immersing them in a boiling solution prepared by mixing 30% H2O2 and 98% H2SO4 with a volume ratio of 3:7. After cooling, the slides were rinsed repeatedly with triply distilled water. The glass slides with hydroxyl surfaces were then immersed in a 0.5% PDDA solution for about 30 min and finally rinsed with triply distilled water. After that, the PDDAcoated slides were soaked in gold colloid for 12 h and rinsed by a PBS buffer. Protein Adsorptions. (1) Purified protein: Glass slides with the gold nanoparticle monolayer were soaked in a 5 mg/mL cytochrome c solution (PBS buffer, pH 7.2) for 1 h at 37 °C and then rinsed by a washing buffer for further use. (2) Immunocomplex: Human IgG (100 µg/mL in the 0.01 M PBS buffer) was immobilized on the gold nanoparticle-coated glass slides by immersing the slides in protein solutions for 2 h at 37 °C. The glass slides were soaked in a blocking buffer for 2 h at 37 °C after being rinsed three times with a washing buffer and then rinsed again three times with the washing buffer. The glass slides coated with human IgG were immersed in the solutions of its corresponding ligands TRITC-antihuman IgG (Fc specific) with different concentrations for 1 h at 37 °C. Then all the glass slides were rinsed three times by the washing buffer. (3) Protein-drug or protein-small molecule complex: Gold nanoparticle-coated glass slides were soaked in 1% BSA (w/w, in the 0.01 M PBS buffer) and avidin (40 µg/mL in the 0.01 M PBS buffer) solutions for 2 h, respectively. The glass slides with avidin were immersed in a blocking buffer for 2 h at 37 °C after washing with a washing buffer. Then, the glass slides with only BSA were soaked in pipemidic acid (4 mg/mL in 0.02 M KOH solution), and the ones with avidin were soaked in Atto610-biotin (PBS buffer, pH 7.2) solutions for 1 h at 37 °C. Finally, nonspecific-adsorbed small molecules were removed by the washing buffer. (20) Frens, G. Nat. Phys. Sci. 1973, 241, 20–22. (21) Lee, P. V.; Meisel, D. J. Phys. Chem. 1982, 86, 3391–3395.

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3351

Figure 1. Schematic depiction of the proposed sandwich substrate and its applications in protein detections.

Atomic Force Microscope (AFM) Measurement. AFM images of the Au/Ag nanoparticles (for average size measurement), gold monolayer, and sandwich substrates were measured with a Digital Instruments Nanoscope IIIA by a multimode using Si cantilevers purchased from DI and Nanosensor Co. Ltd. Field emission scanning electron microscope (FE-SEM) images of the gold monolayer and sandwich substrates were measured with a JEOL JSM-6700F FE-SEM with an accelerating voltage of 3 kV. Colloidal Gold or Silver Staining. Gold monolayer films with adsorbed target proteins or protein-ligand complexes were immersed in gold or silver colloid prepared as above for 3 h at 37 °C and washed by triply distilled water three times. SERS Measurements. SERRS spectra were obtained by using the instrument reported elsewhere.22,23 The laser line was delivered through a 45° angle to the stage of an inverted Olympus (Tokyo, Japan) IX70 microscope using an objective lens (Olympus, 5×, NA 0.15). Magnification of 60× was employed for all the spectral measurements. The excitation laser used was a Kr ion laser (643R-AP-A01, Melles Griot, Tokyo) with the 568 nm line. The scattered signals were notch-filtered (Kaiser, Ann Arbor, MI) and directed to the spectrometer (Pro-275, Acton Research Corporation, Acton, MA) for spectra or to a Nikon (Tokyo, Japan) Coolpix5000 digital camera for optical dark-field microscope images. The laser power of 10 mW combined with an elliptical spot size produced a power density of 0.1 kW/cm-2. An Andor (Belfast, U.K.) DV434-FI charge coupled device (CCD) camera in the single-frame mode was used for collecting Raman spectra. The typical accumulation time for each SERS/SERRS measurement in this study was 10 s. RESULTS AND DISCUSSION Figure 1 illustrates the architecture of the present metal sandwiches and their applications in versatile protein detection. In this work, we developed a new way of immobilizing proteins on chips. Compared to a common Au substrate (generally fabricated by vacuum evaporation with an electron beam), it has two major advantages. On one hand, the first Au layer was prepared by electrostatic interaction between PDDA and negatively charged Au nanoparticles, which is a self-assembled process, so it is cheaper and more accessible. On the other hand, the size of Au nanoparticles used in this study is more homogeneous than (22) Itoh, T.; Hashimoto, K.; Ozaki, Y. Appl. Phys. Lett. 2003, 83, 5557–5559. (23) Itoh, T.; Hashimoto, K.; Ikehata, A.; Ozaki, Y. Chem. Phys. Lett. 2004, 389, 225–229.

3352

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 2. AFM images of (A) a gold nanoparticle monolayer, (B) an Au/Au sandwich, and (C) an Au/Ag sandwich (scan size, 2 µm; data scale, 80 nm) and (D) UV-vis spectra of a gold monolayer, Au/Au, and Au/Ag sandwiches.

that of the common Au substrate,24 and thus, it enables more homogeneous properties in SERS spectra and makes easier analysis of the results. The gold nanoparticles we prepared in the present study are negatively charged, while PDDA is well-known to be strong cationic. Thus, through the LbL technique, PDDA followed by the gold nanoparticles are assembled on hydroxyl-coated glass slides in turn by electrostatic interactions. The sandwich structures are achieved by strong interactions (i.e., hydrophobic, electrostatic, and covalent interactions) between proteins assembled in the middle layer and the two layers of metal nanoparticles. Amino acids residues with high hydrophobicity (e.g., tyrosine and tryptophan residues) bind to metal surfaces by hydrophobic binding.25 Our results demonstrate that all target proteins with different PI can be sandwiched between the two metal layers, which indicate that besides electrostatic interactions the other two binding modes cannot be ignored. The density of gold nanoparticles on the PDDA-coated glass slides increased with the soaking time, and a saturated monolayer could form after about 6 h soaking because of electrostatic repulsion among gold nanoparticles.26 Figure 2A-C show AFM images of a gold nanoparticle monolayer, Au/Au and Au/Ag sandwiches, respectively. The 20 nm gold nanoparticles in the monolayer are too small to be clearly observed on the same scan scale of 2 µm as that for parts B and C of Figure 2. In contrast, a large number of metal aggregates are observed in Figure 2B,C, which indicates the protein-mediated adsorption of gold and silver (24) Vongsvivut, J.; Itoh, T.; Ikehata, A.; Ekgasit, S.; Ozaki, Y. ScienceAsia 2006, 32, 261–269. (25) Carney, J.; Braven, H.; Seal, J.; Whitworth, E. Present and Future Applications of Gold in Rapid Assays. IVD Technol. 2006, 11, 41-51. (26) Grabar, K. C.; Smith, P. C.; Musick, M. D.; Davis, J. A.; Walter, D. G.; Jackson, M. A.; Guthrie, A. P.; Natan, M. J. J. Am. Chem. Soc. 1996, 118, 1148–1153.

nanoparticles. SI-Figure 1 in the Supporting Information shows detailed morphology of the three substrates by field emission scanning electron microscope (FE-SEM). Ultraviolet-visible (UV-vis) spectra of the gold nanoparticle monolayer, Au/Au, and Au/Ag sandwich films are shown in Figure 2D. After assembed on the glass slides, the maximum absorption of the gold colloid shows a red shift by about 30 nm (from 519 to 550 nm), corresponding to dipole-dipole electromagnetic (EM) interactions between the gold nanoparticles.27 As can be seen in Figure 2D, the formation of the Au/Au sandwich architecture results in the enhancement of the absorption peak by 4 times and the further red shift of the maximum absorption to 583 nm, which may arise from the absorption mode of gold aggregates between adjacent nanoparticles. For the Au/Ag sandwich film, after the adsorption of silver nanoparticles, one can find a band at 398 nm due to dipole absorption mode of silver nanoparticles and a broadband at around 580 nm due to EM interactions of upper Ag/Ag nanoparticles or Au/Ag aggregates between the two metal layers.28-30 It is very likely that the observed red-shifts of the plasmon resonance peaks result from metal nanoparticle aggregation. Electromagnetic enhancement of SERS is optimized using a laser line whose frequency is close to that of the plasmon resonance which arises from the aggregation.31 Thus, we chose a 568 nm excitation line for all the SERS measurements. As can be seen from SERS/SERRS spectra shown in parts A, B, and C of Figure 3, we detected cytochrome c (cyt c), protein-ligand interactions between human immunoglobin (IgG) and TRITC-antihuman IgG and between albumin and pipemidic acid (PA) from the proposed Au/Ag sandwich substrates. SERRS spectra of cyt c (excitation into the R region of heme)32 and SERS spectra of PA (absorption maximum located at 332 nm) on Au/ Au sandwich films were too weak to be detected. Notice that we could obtain strong SERS/SERRS signals of cyt c and PA from the Au/Ag sandwiches (Figure 3), which indicated that the adsorption of Ag nanoparticles greatly contributes to the high SERS enhancement. Moreover, we could detect SERRS of TRITC and Atto610 from Au/Au as well as Au/Ag sandwich films (Figure 4). Accordingly, it is very likely that the differences in the SERS signal enhancement of different molecules on the Au/Au sandwich substrates may arise from the differences in SERS activities and the resonance Raman effect as well (SI-Table 1 in the Supporting Information). To verify the large enhancement from the absorption of the second silver layer, we compared SERRS spectra from Au/Ag sandwich films with those from Au/Au sandwich films. It is noted in Figure 4 that SERRS intensities of TRITC and Atto610 labeled complexes from the Au/Ag sandwiches are about 7 times stronger (27) Haynes, C. L.; McFarland, A. D.; Zhao, L.; Van Duyne, R. P.; Schatz, G. C.; Gunnarsson, L.; Prikulis, J.; Kasemo, B.; Kall, M. J. Phys. Chem. B 2003, 107, 7337–7342. (28) Li, X. L.; Xu, W. Q.; Zhang, J. H.; Jia, H. Y.; Yang, B.; Zhao, B.; Li, B. F.; Ozaki, Y. Langmuir 2004, 20, 1298–1304. (29) Zhou, Q.; Li, X.; Fan, Q.; Zhang, X.; Zheng, J. Angew. Chem. 2006, 118, 4074–4077. (30) Pieczonka, N. P. W.; Goulet, P. J. G.; Aroca, R. F. J. Am. Chem. Soc. 2006, 128, 12626–12627. (31) Wang, H.; Levin, C. S.; Halas, N. J. J. Am. Chem. Soc. 2005, 127, 14992– 14993. (32) Hu, S. Z.; Morris, I. K.; Singh, J. P.; Smith, K. M.; Spiro, T. G. J. Am. Chem. Soc. 1993, 115, 12446–12458.

Figure 3. SERS spectra of (A) cyt c, (B) TRITC-labeled immunocomplex (90 µg/mL TRITC-antihuman IgG), and (C) PA-BSA complex from the Au/Ag sandwiches.

Figure 4. SERRS spectra of TRITC-labeled immunocomplex (90 µg/mL TRITC-antihuman IgG) (A) and (B) Atto610-labeled complex (10 µg/mL Atto610-avidin) from the Au/Au and Au/Ag sandwiches, respectively.

than those from the Au/Au sandwiches. The enhanced SERRS spectra from the Au/Ag sandwiches also indicate 7 times increase Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3353

Figure 5. SERRS spectra of TRITC-labeled immunocomplex (90 µg/mL TRITC-antihuman IgG) at five different spots of one sample (A) from protein-mediated Ag aggregates and (B) from the present Au/Ag sandwich.

in detection sensitivity for antihuman IgG and biotin compared to those from the Au/Au sandwiches. We consider that the SERRS enhancement from these sandwiches originates mainly from EM enhancement which involves enhancements in the local EM field due to coupling of plasmon resonance with incident light.1,5 SERRS spectra of TRITC and Atto610 labeled immunocomplexes were nearly undetectable from the gold monolayer, while we could obtain obvious SERRS signals of the two labeled immunocomplexes from the Au/Au sandwiches, which are almost the same as those from gold aggregates (produced by PDDA-bridged LbL technique) in spectral shape and relative intensity. These are indicative of an EM enhancement from the adjacent gold nanoparticles by sandwiched target molecules. Furthermore, we obtained 7 times larger SERRS signals from the Au/Ag sandwiches than those from the Au/Au ones, and no new or larger enhanced band was observed from the former compared to those from the latter. Accordingly, the larger SERRS enhancement from the Au/Ag sandwiches is due to the larger EM enhancement of the silver spheroidal particles than the gold ones17 and as well as the enhanced EM field between the gold and silver naoparticles as a result of the broad SPR of the Au/Ag aggregates.33 SERS bands of analytes from Ag aggregates are susceptible to change because of different orientations on metal surfaces or conformational changes due to intermolecular interactions.34 In our previous study based on protein-mediated silver aggregates,35 we frequently observed different SERS/SERRS spectra with changing wavenumbers from one target molecule, which made it troublesome to choose reliable SERS spectra for target proteins (Figure 5A). The changed wavenumbers probably arise from other species besides the target molecules in the present system that give rise to the observed spectra. In addition, it seems that the (33) Moskovits, M.; Jeong, D. H. Chem. Phys. Lett. 2004, 397, 91–95. (34) MacDonald, I. D. G.; Smith, W. E. Langmuir 1996, 12, 706–713. (35) Han, X. X.; Kitahama, Y.; Tanaka, Y.; Guo, J.; Xu, W. Q.; Zhao, B.; Ozaki, Y. Anal. Chem. 2008, 80, 6567–6572.

3354

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 6. Concentration-dependent SERRS spectra of Atto610labeled bition-avidin complex from the Au/Ag sandwiches. (The inset plots concentration-dependent SERRS intensities at 1637 cm-1. Each data point represents an average of 5-7 measurements, and each error bar indicates the standard deviation).

irreproducible silver substrate is also an important reason because we can observe SERS/SERRS spectra of one target molecule with no obvious change of wavenumbers at different concentrations from the present sandwich substrates. In contrast, the SERRS spectral reproducibility is greatly enhanced as is shown in Figure 5B. We can observe reproducible SERRS spectra from both the Au/Ag and Au/Au sandwiches. It is very likely that in the proposed sandwich architecture, the first gold monolayer plays a crucial role in the enhanced spectral reproducibility. The target labels that are adsorbed on the gold monolayer can retain their stable conformation even after the adsorption of the silver nanoparticles,16 which results in the much higher reproducibility and stability of the SERS spectra from the Au/Ag sandwich substrates as compared to those from the Ag aggregates. Major SERS/SERRS bands of cyt c, PA, and labels used in this study were assigned (SI-Table 2 in the Supporting Information). Furthermore, we compared SERRS spectra of TRITC-immunocomplexes with the same antigen and antibody concentrations from protein-mediated Au aggregates, Au/Au sandwich, Ag aggregates,35 and an Au/Ag sandwich, respectively (Figure 5 and SI-Figure 2 in the Supporting Information). Much stronger SERRS signals can be observed from the proposed Au/Au sandwich than those from the Au aggregates. The reproducibility of SERRS spectra is significantly improved by the proposed Au/Ag sandwich compared to that from the Ag aggregates. Therefore, both the bottom Au and the upper Ag or Au nanoparticles contributed to the SERRS signals observed in this study. Figure 6 exhibits concentration-dependent SERRS spectra of the Atto610-labeled biotin-avidin complex from the Au/Ag sandwiches. It is noted that SERRS intensities of the label significantly increase with the increase of target analyte (Atto610-biotin). The inset of Figure 6 plots SERRS intensities of the Atto610-complex versus the concentrations of the Atto610-biotin. A peak at 1637 cm-1 of Atto610 was chosen to plot these calibration curves, and the detection limits of the Au/Ag sandwiches for Atto610-biotin is as low as 10 pg/mL. Furthermore, in contrast to many Ag-based substrates whose

SERS activity could easily be lost,36 we found that the present metal sandwiches can remain stable in SERS/SERRS intensities for at least 3 weeks. CONCLUSION In summary, based on the combined LbL and colloidal metal staining techniques, the target protein or protein-ligand complexes are sandwiched between the metal nanoparticles where large SERS enhancement emerges. SERS signals of some probes that are very weak from the Au/Au sandwiches can be largely enhanced by the proposed Au/Ag sandwich architectures, and 7 times stronger SERRS spectra are found from the Au/Ag sandwiches as compared to those from the Au/Au sandwiches. Moreover, because of the stable adsorption on the first gold monolayer, the target molecules between two metal layers show much better SERS/SERRS spectra in reproducibility than those from silver aggregates. All the results presented in this paper indicate that the proposed sandwich strategy holds great potential in SERS-based structural and functional proteomic studies. ACKNOWLEDGMENT This study was supported by the National Natural Science Foundation (Grants 20573041, 20773044, 20873050) of P. R. China, (36) Norrod, K. L.; Sudnik, L. M.; Rousell, D.; Rowlen, K. L. Appl. Spectrosc. 1997, 51, 994–1001.

by the Program for Changjiang Scholars and Innovative Research Team in University (Grant IRT0422), Program for New Century Excellent Talents in University, the 111 project (Grant B06009), and the Development Program of the Science and Technology of Jilin Province (Grant 20060902-02). This work was also supported by KAKENHI (Grant-in-Aid for Scientific Research) on Priority Area “Strong Photon-Molecule Coupling Fields (Grant No. 470, 20043032)” from the Ministry of Education, Culture, Sports, Science and Technology of Japan. We also thank the China Scholarship Council for their funding support. SUPPORTING INFORMATION AVAILABLE Different enhancement of target molecules from the Au/Au and Au/Ag sandwiches, major SERS/SERRS band assignments of the target molecules, FE-SEM images of gold monolayer, Au/ Au sandwich and Au/Ag sandwich substrates, SERRS spectra of TRITC-immunocomplexes with the same antigen and antibody concentrations from the Au aggregates and Au/Au sandwich, respectively. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review December 3, 2008. Accepted March 31, 2009. AC802553A

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3355