Quantification and Reactivity of Functional Groups in the Ligand Shell

May 6, 2009 - Lisa Maus, Joachim P. Spatz, and Roberto Fiammengo*. Department of New Materials and Biosystems, Max Planck Institute for Metals ...
0 downloads 0 Views 1MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

Quantification and Reactivity of Functional Groups in the Ligand Shell of PEGylated Gold Nanoparticles via a Fluorescence-Based Assay Lisa Maus, Joachim P. Spatz, and Roberto Fiammengo* Department of New Materials and Biosystems, Max Planck Institute for Metals Research, and Department of Biophysical Chemistry, University of Heidelberg, Heisenbergstrasse 3, D-70569 Stuttgart, Germany Received February 13, 2009. Revised Manuscript Received April 7, 2009 We present a fluorescence-based assay for the characterization of functionalized gold nanoparticles (AuNPs) capped with a self-assembled monolayer of mixed thiols derived from poly(ethylene glycol) (PEG). These water-soluble AuNPs carry primary amino groups at the solvent-exposed interface, which can be used for further conjugation of biologically active molecules. The reported assay allows quantification of the average number of functionalizable amino groups per particle (NNH2) with a relative uncertainty below or equal to (14% (95% confidence interval), thus providing essential information for the successive derivatization of the AuNPs. Here, a fluorescently labeled derivative of peptide-neurotoxin conantokin-G was coupled to the amino groups of the particle ligand shell via a flexible linker. We quantitatively determined the average number of peptides per particle (Npept) and the yield of the two-step conjugation strategy. AuNPs carrying 50-70 copies of the peptide were obtained. In addition, we have gained insights into the deterioration of the selfassembled monolayer due to thiol desorption with time. Under ordinary storage conditions in solution and at room temperature, a decrease in NNH2 between 48% and 75% could be observed at the end of the period of investigation (42-56 days). Slow desorption of the conjugated peptides upon storage was also observed and quantified (∼25% in 14 days).

Introduction Gold nanoparticles (AuNPs) are useful materials for the development of sensing systems and imaging techniques in biology, biotechnology, and nanomedicine.1-6 During the fabrication process, an important role is played by the functionalization of the AuNP surface, frequently achieved with thiol ligands. For instance, this step is used to impart water solubility to the particles,7-12 an essential feature for application of AuNPs in physiologically relevant media. In view of their peculiar properties, such as the nonspecific adsorption of proteins and the high dispersion stability, AuNPs capped with thiol ligands based on poly(ethylene glycol) (PEG) play a prominent role when aiming at *Corresponding author. E-mail: [email protected]; fax: +49 (0)711 6893612. (1) Murphy, C. J.; Gole, A. M.; Stone, J. W.; Sisco, P. N.; Alkilany, A. M.; Goldsmith, E. C.; Baxter, S. C. Acc. Chem. Res. 2008, 41, 1721–1730. (2) Baptista, P.; Pereira, E.; Eaton, P.; Doria, G.; Miranda, A.; Gomes, I.; Quaresma, P.; Franco, R. Anal. Bioanal. Chem. 2008, 391, 943–950. (3) Sperling, R. A.; Rivera Gil, P.; Zhang, F.; Zanella, M.; Parak, W. J. Chem. Soc. Rev. 2008, 37, 1896–1908. (4) Wilson, R. Chem. Soc. Rev. 2008, 37, 2028–2045. (5) Jain, P. K.; El-Sayed, I. H.; El-Sayed, M. A. Nano Today 2007, 2, 18–29. (6) Alivisatos, P. Nat. Biotechnol. 2004, 22, 47–52. (7) Uzun, O.; Hu, Y.; Verma, A.; Chen, S.; Centrone, A.; Stellacci, F. Chem. Commun. 2008, 196–198. (8) Gentilini, C.; Evangelista, F.; Rudolf, P.; Franchi, P.; Lucarini, M.; Pasquato, L. J. Am. Chem. Soc. 2008, 130, 15678–15682. (9) Pengo, P.; Polizzi, S.; Battagliarin, M.; Pasquato, L.; Scrimin, P. J. Mater. Chem. 2003, 13, 2471–2478. (10) Foos, E. E.; Snow, A. W.; Twigg, M. E.; Ancona, M. G. Chem. Mater. 2002, 14, 2401–2408. (11) Kanaras, A. G.; Kamounah, F. S.; Schaumburg, K.; Kiely, C. J.; Brust, M. Chem. Commun. 2002, 2294–2295. (12) Shon, Y.-S.; Wuelfing, W. P.; Murray, R. W. Langmuir 2001, 17, 1255– 1261. (13) Nativo, P.; Prior, I. A.; Brust, M. ACS Nano 2008, 2, 1639–1644. (14) Qian, X.; Peng, X.-H.; Ansari, D. O.; Yin-Goen, Q.; Chen, G. Z.; Shin, D. M.; Yang, L.; Young, A. N.; Wang, M. D.; Nie, S. Nat. Biotechnol. 2008, 26, 83–90. (15) Kim, D.; Park, S.; Lee, J. H.; Jeong, Y. Y.; Jon, S. J. Am. Chem. Soc. 2007, 129, 7661–7665. (16) Zhang, F.; Skoda, M. W. A.; Jacobs, R. M. J.; Zorn, S.; Martin, R. A.; Martin, C. M.; Clark, G. F.; Goerigk, G.; Schreiber, F. J. Phys. Chem. A 2007, 111, 12229–12237.

7910 DOI: 10.1021/la900545t

the investigation of biological interfaces.13-17 Conjugation of appropriately chosen signaling molecules is then used to control the interaction of AuNPs with a desired target,18-20 affording functional AuNPs that specifically interact with proteins,21 nucleic acids,22 and/or other nanoparticles.23 These few examples clearly show that the ligand shell surrounding the gold core of a AuNP is of paramount importance for its functionality. Among the possible signaling molecules, peptides represent a valuable choice for preparing functional AuNPs.24-29 Conjugation of peptides to the ligand shell of AuNPs, and not directly to the Au surface through a thiol-Au bond, is an appealing strategy to preserve the biological activity of the peptides, although seldom employed.30-32 This strategy requires that (1) the AuNPs are (17) Zheng, M.; Li, Z.; Huang, X. Langmuir 2004, 20, 4226–4235. (18) De, M.; You, C.-C.; Srivastava, S.; Rotello, V. M. J. Am. Chem. Soc. 2007, 129, 10747–10753. (19) Takae, S.; Akiyama, Y.; Otsuka, H.; Nakamura, T.; Nagasaki, Y.; Kataoka, K. Biomacromolecules 2005, 6, 818–824. (20) Nam, J.-M.; Thaxton, C. S.; Mirkin, C. A. Science 2003, 301, 1884–1886. (21) Fischer, N. O.; McIntosh, C. M.; Simard, J. M.; Rotello, V. M. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5018–5023. (22) Rosi, N. L.; Giljohann, D. A.; Thaxton, C. S.; Lytton-Jean, A. K. R.; Han, M. S.; Mirkin, C. A. Science 2006, 312, 1027–1030. (23) Storhoff, J. J.; Elghanian, R.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc. 1998, 120, 1959–1964. (24) Patel, P. C.; Giljohann, D. A.; Seferos, D. S.; Mirkin, C. A. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17222–17226. (25) Olmedo, I.; Araya, E.; Sanz, F.; Medina, E.; Arbiol, J.; Toledo, P.; Alvarez-Lueje, A.; Giralt, E.; Kogan, M. J. Bioconjug. Chem. 2008, 19, 1154–1163. (26) Liu, Y.; Shipton, M. K.; Ryan, J.; Kaufman, E. D.; Franzen, S.; Feldheim, D. L. Anal. Chem. 2007, 79, 2221–2229. (27) Levy, R. ChemBioChem 2006, 7, 1141–1145. (28) Pengo, P.; Baltzer, L.; Pasquato, L.; Scrimin, P. Angew. Chem., Int. Ed. 2007, 46, 400–404. (29) Levy, R.; Thanh, N. T. K.; Doty, R. C.; Hussain, I.; Nichols, R. J.; Schiffrin, D. J.; Brust, M.; Fernig, D. G. J. Am. Chem. Soc. 2004, 126, 10076– 10084. (30) De la Fuente, J. M.; Berry, C. C.; Riehle, M. O.; Curtis, A. S. G. Langmuir 2006, 22, 3286–3293. (31) De la Fuente, J. M.; Berry, C. C. Bioconjug. Chem. 2005, 16, 1176–1180. (32) Templeton, A. C.; Hostetler, M. J.; Warmoth, E. K.; Chen, S.; Hartshorn, C. M.; Krishnamurthy, V. M.; Forbes, M. D. E.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 4845–4849.

Published on Web 05/06/2009

Langmuir 2009, 25(14), 7910–7917

Maus et al.

capped by a mixed ligand shell displaying some functionalizable groups, and (2) an efficient coupling reaction for attachment of the peptide to these reactive groups is established. A quantitative characterization of the ligand shell, aiming to determine the number of reactive groups and the number of conjugated signaling molecules, is therefore expected to be of great interest, and it may provide a better understanding of the interaction of functionalized AuNPs with their environment at the molecular level. The characterization of mixed self-assembled monolayers of thiols on AuNPs and the study of the reactivity of the passivation shell of AuNPs have been frequently carried out using 1H NMR.8,33-37 However, despite the usefulness of this technique, it is unsuitable for the analysis of diluted samples (in the low nanomolar range), such as in the case of the watersoluble, relatively large AuNPs obtained from citrate synthesis. Fluorescence-based assays have a much higher sensitivity and have been, for instance, employed by Mirkin and co-workers for the determination of the surface coverage of thiol-capped oligonucleotides bound to 13-250 nm AuNPs.38,39 However, the thiol ligands employed in these and other studies24,40 were fluorescently labeled. Our aim is instead to develop a method that gives an estimate of how many (bio)molecules can be coupled to PEGylated AuNPs once a specific coupling chemistry is selected (vide infra). This information is indeed very significant for the preparation of (bio)functionalized AuNPs, especially when coupling of the signaling molecules is carried out on AuNPs protected by a mixed monolayer whose composition can be varied according to necessity or desire. In order to address this problem, the assay we have developed combines fluorescence quantification with a labeling step. The feasibility of fluorescence labeling of primary amino groups at the solvent-exposed interface of nanoparticles has been shown for 2.2 nm diameter AuNPs surrounded by a silica shell41 and for a 4.0 nm diameter FePt nanoparticle capped with a PEG-based thiol monolayer.42 However, these two systems are substantially different from the AuNPs used in this study, and no fluorescence labeling has been reported to date for AuNPs capped with thiol ligands based on PEG. In this article, we present the setting up of an assay for the quantitative determination of the number of reactive (functionalizable) amino groups in the ligand shell of AuNPs protected with a mixed monolayer of carboxy- and amino-terminated PEGthiols. The N-hydroxysuccinimidyl (NHS)-ester of fluorescein has been used as a labeling compound, so that the assay gives an estimation of the number of molecules that can be conjugated to the AuNPs using NHS-ester chemistry, a reliable and commonly used reaction in many bioconjugation strategies. It is important to (33) Pengo, P.; Polizzi, S.; Pasquato, L.; Scrimin, P. J. Am. Chem. Soc. 2005, 127, 1616–1617. (34) Zheng, M.; Huang, X. J. Am. Chem. Soc. 2004, 126, 12047–12054. (35) Templeton, A. C.; Wuelfing, W. P.; Murray, R. W. Acc. Chem. Res. 2000, 33, 27–36. (36) Hostetler, M. J.; Templeton, A. C.; Murray, R. W. Langmuir 1999, 15, 3782–3789. (37) Templeton, A. C.; Hostetler, M. J.; Kraft, C. T.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 1906–1911. (38) Hurst, S. J.; Lytton-Jean, A. K. R.; Mirkin, C. A. Anal. Chem. 2006, 78, 8313–8318. (39) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A.; Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535–5541. (40) Hong, R.; Fernandez, J. M.; Nakade, H.; Arvizo, R.; Emrick, T.; Rotello, V. M. Chem. Commun. 2006, 2347–2349. (41) Schroedter, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41, 3218–3221. (42) Latham, A. H.; Williams, M. E. Langmuir 2006, 22, 4319–4326. (43) The acidity of the carboxylic and amino groups in the self-assembled monolayer covering the AuNPs is expected to be different from the bulk materials, and this might in turn influence the reactivity of the amino groups. For studies on planar substrates see(a) Aureau, D.; Ozanam, F.; Allongue, P.; Chazalviel, J. N. Langmuir 2008, 24, 9440–9448. (b) Burris, S. C.; Zhou, Y.; Maupin, W. A.; Ebelhar, A. J.; Daugherty, M. W. J. Phys. Chem. C 2008, 112, 6811–6815.

Langmuir 2009, 25(14), 7910–7917

Article

stress that this assay does not give the total average number of amino groups per particle, but, more interestingly, how many of these can be reacted under a defined set of conditions.43 Taking advantage of this information, we have exploited a heterobifunctional linker and NHS chemistry to couple a fluorescently labeled derivative of the peptide-neurotoxin conantokin-G (5) to the ligand shell of these AuNPs. The number of conjugated peptides per particle has been determined similarly to the number of reactive amino groups. In both cases (determination of the number of reactive amino groups and of coupled peptides per particle), small AuNP sample volumes are sufficient for the quantification: we typically used 100-200 μL of a 1-5 nM AuNP solution for analysis.

Results and Discussion Preparation of AuNPs Protected with a Mixed Monolayer of PEG-Thiols. AuNPs having an average diameter between 29 and 41 nm and a narrow size distribution (relative standard deviation σ ∼ 10%) were prepared by seeded-growth of citrate-capped nanoparticles44-46 using an improved experimental procedure.47 Passivation of the AuNP surface, i.e., formation of a self-assembled monolayer of ligands, was achieved by treating the AuNPs with solutions containing carboxyl- and aminoterminated PEG-based thiols (1-4, Chart 1) in large excess (>1  106 ligands per particle). For the preparation of AuNPs with a mixed monolayer ligand shell, solutions containing mixtures of PEG3000 thiols 1 and 2 or alkyl-PEG600 thiols 3 and 4 were used, where the molar fraction of the amino-terminated ligand was varied between xNH2 = 0.09 and 0.33. The AuNPs therefore display an excess of carboxylate-terminated PEG molecules on their surface, are negatively charged, and do indeed migrate toward the positive electrode in gel electrophoresis (Figure S-7, Supporting Information). These passivated AuNPs are very resistant to irreversible aggregation and can be purified to remove excess thiols using a combination of ultrafiltration (membrane cutoff 100 kDa) followed by gel filtration on prepacked Sephadex G-25 columns. The recovery of purified AuNPs is generally high, between 60 and 95%. Assay for Determination of the Number of Reactive Amino Groups Per Particle (NNH2). Thiols 2 and 4 were used to introduce primary amino groups at the solvent exposed interface of the AuNP ligand shell. These groups can easily be reacted via a number of strategies developed for the preparation of protein conjugates.48 ω-Functionalized thiols have already been used to prepare monolayer-protected AuNPs carrying reactive functional groups in the ligand shell35,49 that have then been further derivatized using alkylation37 and acylation32,34,42,50,51 reactions. Yet, much to our surprise, ω-amino thiols52 have not been used for this purpose and are practically unexplored for the passivation of AuNPs.53,54 Our interest, after having established (44) Cao, L.; Zhu, T.; Liu, Z. J. Colloid Interface Sci. 2006, 293, 69–76. (45) Brown, K. R.; Walter, D. G.; Natan, M. J. Chem. Mater. 2000, 12, 306–313. (46) Brown, K. R.; Natan, M. J. Langmuir 1998, 14, 726–728. (47) Maus, L.; Spatz, J. P.; Fiammengo, R. Manuscript in preparation. (48) Hermanson, G. T., Ed. Bioconjugate Techniques; Academic Press: New York, 1995. (49) Tan, H.; Zhan, T.; Fan, W. Y. J. Phys. Chem. B 2006, 110, 21690–21693. (50) Roux, S.; Garcia, B.; Bridot, J.-L.; Salome, M.; Marquette, C.; Lemelle, L.; Gillet, P.; Blum, L.; Perriat, P.; Tillement, O. Langmuir 2005, 21, 2526–2536. (51) Templeton, A. C.; Cliffel, D. E.; Murray, R. W. J. Am. Chem. Soc. 1999, 121, 7081–7089. (52) Tshikhudo, T. R.; Wang, Z.; Brust, M. Mater. Sci. Technol. 2004, 20, 980– 984. (53) Rucareanu, S.; Gandubert, V. J.; Lennox, R. B. Chem. Mater. 2006, 18, 4674–4680. (54) Lin, S.-Y.; Tsai, Y.-T.; Chen, C.-C.; Lin, C.-M.; Chen, C.-h. J. Phys. Chem. B 2004, 108, 2134–2139.

DOI: 10.1021/la900545t

7911

Article Chart 1. ω-Functionalized PEG3000 Thiols (1 and 2) and Alkyl-PEG600 Thiols (3 and 4)a

a

Average ethylene glycol units: 1 and 2, n = 68; 3 and 4, m = 13.

the preparation of AuNPs capped with a mixed monolayer containing ω-amino thiols, was therefore to quantify how many amino groups per particle were actually functionalizable. Acylation of the amino groups with NHS ester derivatives was chosen as the model reaction because of the widespread use of these derivatives in many bioconjugation strategies. Experimental conditions such as temperature, AuNP concentration, and cosolvents were chosen to be representative of those that could be used during conjugation reactions involving peptides (vide infra). The assay strategy is illustrated in Scheme 1. Directly after preparation, AuNPs were reacted with in situ-prepared 5(6)-carboxyfluorescein NHS-ester. Excess dye was removed by ultrafiltration and gel filtration, and, at this point, the AuNP sample was split into two portions. One portion was treated with dithiothreitol in phosphate buffer (DTT/PB), which effectively stripped off the ligand shell from the AuNPs surface.38,55,56 The fluorescence signal of the released ligands was measured, and the concentration of labeled amino groups was obtained by comparison to a calibration curve. The second portion was analyzed using atomic adsorption (AA) or inductively coupled plasma optical emission spectroscopy (ICP-OES), affording the gold content in the solution. The concentration of AuNPs was calculated from the amount of gold and the average diameter of the gold core measured by electron microscopy (scanning (SEM) or transmission (TEM)).29 Since the AuNPs used for this study are spherical and have a narrow size distribution, these calculations allow accurate determination of AuNP concentration.57 Finally, the average number of functionalizable amino groups per particle (NNH2) was calculated by dividing the concentration of labeled amino groups by the AuNPs concentration. The complete assay was, in most cases, performed in duplicate to assess its repeatability. In parallel, control assays excluding 1-(3-diethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) and NHS in the in NHS-ester formation step were run and resulted in a residual fluorescence signal practically indistinguishable from the blank. This confirms that fluorescein must have been covalently bound to the passivation layer of the AuNPs via amide bond formation and shows that nonspecific adsorption of fluorophore on the AuNPs surface did not occur. From repeated measurements (fluorescence and ICP-OES or AA) and repeated assays, the uncertainty associated with the determination of NNH2 was estimated to be e14% (at a 95% confidence interval; see Supporting Information), a very satisfactory value for an indirect determination. The results of 37 assays performed on 21 different AuNP samples are presented in Table 1. Results (55) Thaxton, C. S.; Hill, H. D.; Georganopoulou, D. G.; Stoeva, S. I.; Mirkin, C. A. Anal. Chem. 2005, 77, 8174–8178. (56) Letsinger, R. L.; Elghanian, R.; Viswanadham, G.; Mirkin, C. A. Bioconjug. Chem. 2000, 11, 289–291. (57) We have used the same method of determining AuNP concentration for seeded-growth reactions. The excellent agreement between the actual particle size after growth and the calculated size based on the amount of added Au confirms that the number of AuNPs in solution is accurately estimated. Details of these experiments will be reported elsewhere.

7912 DOI: 10.1021/la900545t

Maus et al.

are separated according to the set of thiol ligands that have been used to passivate the AuNP surface (PEG3000-based thiols 1 and 2 or alkyl-PEG600 thiols 3 and 4). For each batch number, different letters identify distinct AuNP samples obtained from independent passivation reactions, but starting from the same batch of seeded-growth particles (Scheme S-1). Therefore, in these cases, the assay could be used to assess the variability introduced during AuNPs preparation, limited to the mixed monolayer formation step (passivation reaction). In seven out of eight cases, the uncertainty intervals around the NNH2 values partially overlap, indicating no significant difference between the values (batches 2a,b, 3a,b, 5a-c, 6a,b, 7a,b, 9a,b, and 11a,b). In the remaining case (batch 1a,b), the uncertainty intervals are almost contiguous. Taken together, these findings indicate that, starting from a particular batch of AuNPs obtained by seeded-growth, it is possible to prepare mixed monolayer-protected AuNPs carrying a specified number NNH2 of reactive amino groups on their surface with a relative uncertainty of (14% (95% confidence interval). Furthermore, the assay allows monitoring of the variability of completely independently prepared AuNPs batches, i.e., AuNPs prepared by separate seeded-growth reactions followed by passivation (Scheme S-1). The results obtained from independent AuNP batches with almost overlapping size distributions (Figure 1b) showed a larger variability (batches 2, 3, and 4-6, Figure 1a) compared to the situation where only the passivation reaction was independently performed. In fact, even if the conditions for preparation of AuNPs were kept as constant as possible, NNH2 varied up to a factor of 3 (batches 4-6). Although in this study we only dealt with a relatively narrow range of AuNP sizes (15 - 41 nm), it was expected that an increase in AuNP surface area would lead to an increase in the average number of functionalizable amino groups. A plot of the 21 NNH2 values in Table 1 as a function of the respective particle surface area (Figure S-6) shows only a moderate correlation between the two variables (R = 0.67). Interestingly, there is no significant difference in NNH2 between AuNPs protected with PEG3000 thiols 1 and 2 or alkyl-PEG600 thiols 3 and 4 (Figure S-6). The observed scattering in the data suggests, as already described, that the seeded-growth process produces slightly different batches of AuNPs even when analysis of the AuNPs by electron microscopy shows practically overlapping size distributions. This leads to the formation of the mixed monolayer ligand shells with a certain degree of variability from one batch to another. Interestingly, we are now in the position to spot these differences with a simple and sensitive fluorescence-based assay that can be performed routinely at the end of particle preparation. Finally, using both PEG3000 thiols 1 and 2 and alkyl-PEG600 thiols 3 and 4, we have prepared 10 different AuNPs samples by varying the molar fraction of the ω-amino functionalized thiol xNH2 between 0.09 and 0.33 during the passivation reaction (Figure 2). The results show a linear increase in NNH2 irrespective of the chosen thiol ligand system, demonstrate the possibility of preparing AuNPs carrying a higher number of functionalizable groups, and confirm the suitability of the reported assay for the quantification of the number of reactive amino groups per AuNP. Additionally, the total number of reactive amino groups NNH2 can be linearly extrapolated to estimate NNH2 for hypothetical AuNPs passivated exclusively with ω-amino functionalized thiol ligands 2 or 4 (these particles cannot be prepared because they irreversibly aggregate during the passivation reaction). The extrapolation affords ∼2800 NNH2 for 2 and ∼2700 for 4 and confirms that there is no significant difference in NNH2 between AuNPs passivated with PEG3000 thiols 1 and 2 and those with alkyl-PEG600 thiols 3 and 4, as already concluded from the data in Figure S-6. Langmuir 2009, 25(14), 7910–7917

Maus et al.

Article Scheme 1. Design of the Fluorescence-Based Assay for the Measurement of NNH2

Table 1. Average Number of Functionalizable Amino Groups Per Particle (NNH2)a batch

L (nm)

NNH2

PEG3000 (1 + 2) 1a 1b 2a 2b 3a 3b 4 5a 5b 5c 6a 6b 7a 7b 8

41 ( 4 35 ( 4 34 ( 4 32 ( 3 32 ( 3 31 ( 3 29 ( 3 15 ( 1

613 ( 86 477 ( 41 221 ( 31 281 ( 39 319 ( 45 294 ( 41 136 ( 19 388 ( 54 385 ( 32 365 ( 51 233 ( 33b 259 ( 36b 238 ( 33 209 ( 29 165 ( 23b

alkyl-PEG600 (3 + 4) 223 ( 34 211 ( 30 32 ( 3 202 ( 28 31 ( 3 220 ( 30b 180 ( 12b 30 ( 3 139 ( 19 a Assays were performed in duplicate unless otherwise noted. For each batch number, different letters identify AuNP samples obtained from independent passivation reactions, but starting from the same batch of seeded-growth particles. Different batch numbers correspond to completely independently prepared AuNPs. Uncertainty intervals for NNH2 values are 95% confidence intervals. Uncertainty intervals on the diameter L correspond to the width parameter (2  σ) of the Gaussian distribution fitting the particle size distribution. Data labeled PEG3000 are for monolayer-protected AuNPs using thiols 1 and 2, and data are labeled alkyl-PEG600 when using thiols 3 and 4, xNH2 = 0.14 in all cases. b Assay performed only once. 9a 9b 10 11a 11b 12

35 ( 5

Moreover, if the estimated NNH2 is assumed to be equal to the total number of ω-amino functionalized ligands per particle, a ligand footprint area of ∼1 nm2 can be calculated, which is less than 3 times larger (lower ligand density) than the 0.35 nm2 (58) Wuelfing, W. P.; Gross, S. M.; Miles, D. T.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 12696–12697.

Langmuir 2009, 25(14), 7910–7917

calculated by Murray and co-workers for much smaller, 2.8 nm diameter, AuNPs capped with PEG5000 thiols.58 Since it is wellknown that small AuNPs display a higher number of thiol ligands per surface area because of their high surface curvature,59 the calculated ligand footprint area appears to be consistent with the literature and provides a strong indication that indeed the measured NNH2 is probably close to the total number of ω-amino functionalized ligands. Effect of Aging on the Number of Reactive Amino Groups Per Particle. Even if thiols are extensively employed as ligands for the stabilization of AuNPs, literature data point out that the quality of the self-assembled monolayer deteriorates with time as a result of ligand desorption. Oxidation of the sulfur atom attached to the gold surface seems to be the major process responsible for this phenomenon.60-63 Macroscopically, the colloidal AuNP solutions prepared for this study were stable for months. In particular, their optical properties did not significantly vary, even after 7 months (Figure S-1), and no change was observed by electron microscopy (SEM, Figure S-2). Nevertheless, we set out to determine if, and to what extent, the number of reactive amino groups at the AuNPs surface (NNH2) changes over time as a consequence of particle aging. After passivation and purification, AuNPs were stored at room temperature in conventional darkened disposable plasticware.64 At later times (between 1 and 56 days) samples of these solutions were taken, and the number of amino groups was determined. Measurements were performed on 10 AuNPs batches (five passivated with PEG3000 thiols 1 and 2 and five with alkyl-PEG600 thiols 3 and 4), and in all cases a significant decrease in NNH2 over time was observed. In particular, for 7 out of 10 batches, the decrease was between 48% and 75% at the end of the period of investigation (42-56 days). (59) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Science 2007, 318, 430–433. (60) Herdt, A. R.; Drawz, S. M.; Kang, Y.; Taton, T. A. Colloids Surf., B 2006, 51, 130–139. (61) Schoenfisch, M. H.; Pemberton, J. E. J. Am. Chem. Soc. 1998, 120, 4502– 4513. (62) Scott, J. R.; Baker, L. S.; Everett, W. R.; Wilkins, C. L.; Fritsch, I. Anal. Chem. 1997, 69, 2636–2639. (63) Schlenoff, J. B.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528–12536. (64) For best long-term storage of functionalized AuNPs, protective atmosphere (Ar or N2) and low temperature are recommended. However, these conditions were chosen to investigate nanoparticle robustness under daily use conditions commonly encountered in more biologically oriented laboratories or even in medicine.

DOI: 10.1021/la900545t

7913

Article

Figure 1. (a) The average number of functionalizable amino groups per AuNP (NNH2) for five completely independent AuNP batches, is found to be quite variable. Particles can be grouped into two size classes with practically overlapping size distributions (], 0, O and b, 2). Data points with the same symbols are AuNPs obtained from independent passivation reactions, but starting from the same batch of seeded-growth particles. Error bars are 95% confidence intervals. (b) Size distribution for the five independent AuNPs batches.

Figure 2. The average number of functionalizable amino groups per AuNP (NNH2) increases with the molar fraction of ω-amino functionalized thiols (xNH2) used during the passivation reaction (monolayer formation). (9) PEG3000 thiols 1, 2; (O) alkylPEG600 thiols 3, 4. AuNP diameter L = 30 ( 3 nm.

For the remaining three batches, the decrease was between 24% and 33%, (Table S-1). While a decrease was clearly observable in all cases and was most prominent after the first 14 days, the magnitude of this phenomenon seemed to vary from sample to sample, possibly reflecting initial differences between independent batches, and/or small variations during storage of the individual samples (e.g., how often samples were opened, vortexed, pipetted). AuNP batches obtained from independent passivation reactions, but from a common seeded-growth, show good reproducibility (Figure 3). A clear distinct behavior between AuNPs passivated with PEG3000 thiols 1 and 2 or with alkyl-PEG600 thiols 3 and 4 was observed only in one experiment (Figure 3a), indicating that differences related to the nature of the passivation layer are probably comparable to or smaller than the actual variability caused by sample history. To rule out the possibility that NNH2 decreased over time as a result of factors other than thiol desorption, freshly passivated AuNP batches were reacted with fluorescein-NHS ester directly after their preparation, purified, and stored as before at room temperature and in the dark. At later times, samples were taken, purified again by 7914 DOI: 10.1021/la900545t

Maus et al.

ultrafiltration and gel filtration to remove any desorbed fluorescently labeled PEG-derivative, and analyzed (Scheme S-2). A significant decrease in NNH2, comparable to that previously observed, was also measured for these batches (-55% for AuNPs passivated with 1 and 2 and -40% for 3 and 4; Figure 3c) and supports the hypothesis that the observed decrease is indeed caused by ligand desorption. A detailed investigation of all the factors influencing this phenomenon is beyond the scope of this research. However, the collected data clearly show that aging produces considerable changes of the AuNPs passivation layer and give an idea of the time scale of this phenomenon under common laboratory conditions even when no significant change was detected by UVvisible (UV-vis) or SEM analysis. Coupling of Peptides to AuNPs. Peptide-functionalized AuNPs have been studied for targeting particular cell types, or cellular organelles.13,31 At least two strategies are conceivable for the preparation of peptide-functionalized AuNPs: the first makes use of thiol-terminated peptides that are directly immobilized on the gold surface through the strong S-Au interaction, while the second involves conjugation of the peptide to the ligand shell. The first strategy is attractive because of its simplicity,27,65 but it suffers from some limitations such as the often impaired solubility of the functionalized AuNPs. These problems can only be overcome sometimes, e.g., by the judicious choice of a combination of peptides.66 To the best of our knowledge, only one group has reported AuNP functionalization with peptides following the second strategy; however, no indication about the number of functionalizable groups per particle or about the number of coupled peptides was given.30,31 Templeton et al. reported a multistep synthetic strategy that allowed them to grow a tripeptide on the surface of organic soluble AuNPs.32 Franzen and Feldheim used a mixed strategy where peptides were first coupled to bovine serum albumin (BSA), and the purified conjugates were successively used to functionalize the AuNPs.67-70 The AuNPs with a mixed monolayer ligand shell containing primary amino groups described in this study possess very good water solubility, stability toward aggregation, and are therefore ideally suited to being further functionalized with peptides by conjugation to the ligand shell. Coupling of a fluorescently labeled peptide to the AuNPs should allow determination of the average number of peptides per particle with an assay analogous to that developed to measure the average number of reactive amino groups per particle (Scheme 1). To test this idea, tetramethylrhodamine (TAMRA) fluorescently labeled peptide 5 was prepared and coupled to AuNPs by means of the heterobifunctional cross-linker 6 (Scheme 2). The amino acid sequence of synthetic peptide 5 corresponds to that of the naturally occurring peptide-toxin conantokin-G (con-G), a component of the venom of the predatory marine snail Conus geographus.71 Peptide 5 carries an additional cystein at the C-terminus, which is used for conjugation to the AuNPs via Michael addition to the mealeimido group, and the side chain of lysine-15 is TAMRA labeled. This peptide (65) Wang, Z.; Levy, R.; Fernig, D. G.; Brust, M. Bioconjug. Chem. 2005, 16, 497–500. (66) Sun, L.; Liu, D.; Wang, Z. Langmuir 2008, 24, 10293–10297. (67) Ryan, J. A.; Overton, K. W.; Speight, M. E.; Oldenburg, C. N.; Loo, L.; Robarge, W.; Franzen, S.; Feldheim, D. L. Anal. Chem. 2007, 79, 9150–9159. (68) Tkachenko, A. G.; Xie, H.; Liu, Y.; Coleman, D.; Ryan, J.; Glomm, W. R.; Shipton, M. K.; Franzen, S.; Feldheim, D. L. Bioconjug. Chem. 2004, 15, 482–490. (69) Xie, H.; Tkachenko, A. G.; Glomm, W. R.; Ryan, J. A.; Brennaman, M. K.; Papanikolas, J. M.; Franzen, S.; Feldheim, D. L. Anal. Chem. 2003, 75, 5797–5805. (70) Tkachenko, A. G.; Xie, H.; Coleman, D.; Glomm, W.; Ryan, J.; Anderson, M. F.; Franzen, S.; Feldheim, D. L. J. Am. Chem. Soc. 2003, 125, 4700–4701. (71) McIntosh, J. M.; Olivera, B. M.; Cruz, L. J.; Gray, W. R. J. Biol. Chem. 1984, 259, 14343–14346.

Langmuir 2009, 25(14), 7910–7917

Maus et al.

Article

Figure 3. The average number of functionalizable amino groups per AuNP (NNH2) decreases significantly as a function of time. Circles + dashed line: AuNPs passivated with PEG3000 thiols 1 and 2; triangles + solid line: AuNPs passivated with alkyl-PEG600 thiols 3 and 4; filled and open symbols of the same shape indicate AuNPs batches prepared by independent passivation reactions, but starting from the same batch of seeded-growth particles. At each given time, a sample of AuNPs with diameter (a) L = 35 ( 4 nm and (b) L = 31 ( 3 nm. was withdrawn from a stock solution and analyzed. (c) Labeling with fluorescein at 0 days (Scheme S-2). At each given time point, samples of the fluorescently labeled AuNPs (L = 30 ( 3 nm) were withdrawn, purified, and analyzed. Error bars are 95% confidence intervals.

still possesses the potent inhibitory activity on N-methyl-Daspartate receptors72 of its natural counterpart. Coupling of the peptide was carried out as a two-step procedure: the more labile NHS-ester functionality was reacted with the amino groups at the AuNPs surface, followed by coupling of the thiol-terminated peptide to the linker-functionalized particles (Scheme 2). In between these two steps, the AuNPs were separated from excess uncoupled linker by ultrafiltration and gel filtration. At the end, peptide-AuNP conjugates were purified to remove uncoupled peptide in the same way. This process was monitored by measuring the fluorescence intensity of the solutions in equilibrium with the AuNPs until background levels of fluorescence were reached. According to the calibration curve, this corresponds to a concentration of uncoupled peptide in solution below 0.37 nM or, in other words, less than one molecule per AuNP was free in solution after purification. Two aliquots of the purified peptide-AuNP conjugate sample were withdrawn, the first for determining the concentration of coupled peptides by fluorescence intensity (after DDT treatment) and the second for measuring the gold concentration via AA or ICP-OES, analogously to the assay sketched in Scheme 1. Table 2 reports the average number of peptide copies per particle (Npept)73 for two batches of AuNPs passivated with alkylPEG600 thiols 3 and 4.74 Coupling of peptide 5 was carried out directly after preparation of the AuNPs (1 day, samples 1a and 2a), so that the amount of amino groups per particle could be as high as possible. For batches 1a and 2a, NNH2 was also measured (Table 1 batches 10 and 12, respectively) in order to determine the yield of the coupling procedure, which proved to be quite satisfactory (26% and 50%) considering the two-step procedure involving a heterobifunctional linker. In addition, since peptide 5 is Cys-terminated, experiments were carried out in parallel, omitting linker 6 during the first step of the coupling procedure to investigate direct binding of 5 to the Au surface of the particles. The results show that direct binding is negligible compared to conjugation via the linker (Npept_C vs Npept batches 1a-c and 2a), suggesting that the peptides were not able to penetrate the passivation layer and place-exchange the alkylPEG600 thiols. This result is certainly dependent on the chemical (72) Mena, E. E.; Gullak, M. F.; Pagnozzi, M. J.; Richter, K. E.; Rivier, J.; Cruz, L. J.; Olivera, B. M. Neurosci. Lett. 1990, 118, 241–244. (73) See Table S-2 for a comparison of these data with the average number of peptide per particle reported in other works on peptide-AuNP conjugates. (74) AuNPs passivated with alkyl-PEG600 thiols 3 and 4 were chosen for illustrating peptide conjugation because of their reduced unspecific surface adsorption during purification by ultrafiltration and gel filtration.

Langmuir 2009, 25(14), 7910–7917

nature of the peptides and the passivation layer75 and to the experimental conditions used during coupling, but it is worth noting that, under commonly used coupling conditions, such as being kept overnight in aqueous solution containing 22 μM of 5 (ratio peptides/AuNPs g5000) at pH 7 and 0 °C, this kind of selectivity can be reached. Effect of Aging on the Number of Peptides Per Particle. It was expected that the number of coupled peptides would decrease over time due to thiol desorption from the AuNPs surface, similarly to the decrease in functionalizable amino groups. Therefore, 5 was coupled to the particles shortly after their preparation (day 1), and the purified conjugates were stored at room temperature in conventional disposable plasticware in the dark. Samples of this stock solution were taken at increasing time intervals, purified to remove desorbed peptide, and analyzed (batches 2b-d). After 42 days, 35% of the conjugated peptides were lost. As a control (Scheme S-2), one AuNPs batch was stored after preparation at room temperature in conventional disposable plasticware, and the coupling of peptide 5 was repeated on withdrawn samples at increasing time intervals (batches 1b-c). The number of peptides per particle decreased from one sample to the next and amounted to 30% after 42 days. The two results agree quite well with each other and with the data previously obtained for the decrease of NNH2 in time (e.g., -40% after 42 days for similar particles labeled with fluorescein, Figure 3c). This further supports that the observed phenomenon corresponds to desorption of alkyl-PEG600 thiols (either underivatized or ω-derivatized with fluorescent molecules).

Conclusions Many applications in the new fields of bionanotechnology and nanomedicine rely on the use of functionalized AuNPs. In order to understand the interaction between these nanometer-sized objects with biologically relevant interfaces, it is important to precisely characterize them. Here we have described a fluorescence-based assay for the characterization of mixed self-assembled monolayers containing functionalizable amino groups on the AuNPs surface. Starting from a particular batch of AuNPs obtained by seeded-growth, it is possible to prepare mixed monolayer-protected AuNPs carrying a specified number NNH2 of reactive amino groups on their surface with a relative uncertainty of (14% (95% confidence interval). The same idea has been used for the determination of the average number of (75) The highly negatively charged, 18-amino acid-long peptide 5 most probably has very unfavorable interaction with the apolar region (C11-chains) of the passivation layer, which prevents its diffusion.

DOI: 10.1021/la900545t

7915

Article

Maus et al. Scheme 2. Functionalization of AuNPs with Fluorescently Labeled Peptide 5 via Hetero-bifunctional Crosslinker 6

Table 2. Average Number of Copies of Peptide 5 per Particle (Npept)a batch

time (d)

Npept

Npept_Cb

yieldc (%)

a 1 53 ( 4 3 26 b 14 41 ( 3 1 c 42 37 ( 3 5 a 1 69 ( 6 0 50 2d,f b 14 50 ( 3 c 28 51 ( 10 d 42 45 ( 11 a AuNPs were passivated with alkyl-PEG600 thiols 3 and 4 (XNH2 = 0.14). b Number of peptides per particle when linker 6 was omitted from the coupling reaction. c Coupling yield calculated as Npept/NNH2. d AuNPs batches 1 and 2 correspond to batch 10 (L = 32 ( 3 nm), and batch 12 in Table 1 (L = 30 ( 3 nm), respectively. e At the indicated day, peptide coupling was carried out followed by analysis. f At day 1, peptide was coupled to the entire AuNP batch. Samples were withdrawn, purified, and analyzed on the indicated day. 1d,e

fluorescently labeled peptide 5 that can be conjugated to these amino groups. In addition, we have showed and quantified a significant desorption of thiol ligands from the AuNP surface caused by aging, even when the macroscopic appearance of the samples seemed unaffected. We expect that the implementation of this or analogous fluorescence-based assays in the routine preparation of functionalized AuNPs will allow a precise characterization of the particles, and consequently a much better understanding of the results obtained in studies where AuNPs are used. For instance, the quantification of the number of biologically active peptides on the AuNP surface at a certain time is certainly of primary importance for the study of epitopereceptor interaction using AuNPs as tools. Currently, we are aiming to use AuNPs-peptide 5 conjugates to investigate the inhibition of N-methyl-D-aspartate receptors in hippocampal neurons.

prepared according to literature procedures.76,77 Peptide 5 (sequence: GEγγLQγNQγLIRγK(TAMRA)SNC-NH2, γ = γ-carboxyglutamate) was purchased in a custom synthesis from Peptide Specialty Laboratories (PSL, Heidelberg, Germany). All employed glassware was cleaned with aqua regia (HCl (37%)/ HNO3 (65%) 3:1). Ultrapure deionized water (TKA GenPure water purification system, 18.2 MΩ cm) was used for all solution preparations. AuNP synthesis,78,79 seeded-growth,44,45 and passivation was carried out according to optimization of known literature procedures, and a detailed description will be reported elsewhere. Purification of AuNPs was performed by ultrafiltration using Amicon Ultra-4 Centrifugal Filter Units - 100 kDa from Millipore and gel filtration using NAP-10 columns from GE Healthcare. UV-vis spectra and fluorescence measurements were carried out using a TECAN Infinite M200 plate reader. The concentration of gold was determined via ICP-OES (Spectro, CIROS CCD) or AA (Perkin-Elmer, PE 5000). AuNP diameter was measured via SEM (Zeiss, Ultra 55 equipped with a Gemini gun, 5 kV) or TEM (Philips CM 200, 200 kV).

Determination of the Number of Reactive Amino Groups Per Particle (NNH2). Data reported in Table 1 and Figures 1, 2,

Materials and Instruments. 5(6)-Carboxyfluorescein was purchased from Novabiochem; dimethylformamide (DMF)peptide, synthesis quality, was obtained from Fluka; EDC and NHS were purchased from Acros Organics; heterobifunctional linker 6 was from Pierce; DTT was from Serva; and PEG3000 thiols 1 and 2 were from RAPP Polymere GmbH :: (Tubingen, Germany). Alkyl-PEG600 thiols 3 and 4 were

and S-6 are obtained by carrying out the assay directly after AuNP preparation. A solution of 5(6)-carboxyfluorescein NHS-ester was freshly prepared by dissolving 5(6)-carboxyfluorescein (5.60 mg, 14.9 μmol) in 200 μL of DMF followed by addition of NHS (2.11 mg, 18.7 μmol, 1.2 equiv) and EDC (3.60 mg, 18.3 μmol, 1.2 equiv). The reaction mixture was shaken for 2 h at 4 °C. In a typical experiment, a 100-200 μL aliquot of an ∼1-5 nM solution of passivated AuNPs in 40 mM NaHCO3 was diluted to 1 mL with 40 mM NaHCO3 and 20 μL of the freshly prepared solution of 5(6)-carboxyfluorescein NHS-ester (74 mM) were added. The reaction mixture was shaken for 2 h at 4 °C. In parallel, a 74 mM solution of 5(6)-carboxyfluorescein in DMF was added to a separate aliquot of passivated AuNPs, and the mixture was also shaken for 2 h at 4 °C (as a negative control for the detection of nonspecifically adsorbed fluorophore). After this time, AuNPs were purified, and the excess uncoupled 5(6)carboxyfluorescein was removed by ultrafiltration (washed with 2  4 mL of 50 mM N2B4O7 3 10 H2O, pH ∼ 9.4; 2  4 mL 2:8 EtOH /50 mM N2B4O7 3 10 H2O, pH ∼ 9.4; 2  4 mL 40 mM NaHCO3, pH 8.4) followed by gel filtration, eluting 40 mM NaHCO3. Finally, the eluate was concentrated via ultrafiltration and taken up in 350 μL of 40 mM NaHCO3. The intensity of the fluorescence signal for each filtrate was measured and reached background levels at the end of the purification, ensuring that all uncoupled 5(6)-carboxyfluorescein had been removed. A 50150 μL aliquot of the purified AuNP solution was withdrawn and

(76) Chirakul, P.; Perez-Luna, V. H.; Owen, H.; Lopez, G. P.; Hampton, P. D. Langmuir 2002, 18, 4324–4330. (77) Lee, J. K.; Kim, Y.-G.; Chi, Y. S.; Yun, W. S.; Choi, I. S. J. Phys. Chem. B 2004, 108, 7665–7673.

(78) Pong, B.-K.; Elim, H. I.; Chong, J.-X.; Ji, W.; Trout, B. L.; Lee, J.-Y. J. Phys. Chem. C 2007, 111, 6281–6287. (79) Grabar, K. C.; Freeman, R. G.; Hommer, M. B.; Natan, M. J. Anal. Chem. 1995, 67, 735–743.

Experimental Section

7916 DOI: 10.1021/la900545t

Langmuir 2009, 25(14), 7910–7917

Maus et al. treated with 0.5 mL of concd HCl and 0.25 mL of concd HNO3 to dissolve the AuNPs, diluted to 10 mL with H2O, and the Au concentration was determined either by ICP-OES or AA (measurements repeated three times). A second aliquot of the purified AuNP solution (100-200 μL) was added to one volume of DTTPB buffer (1 M DTT in 0.18 M sodium phosphate buffer pH 7.9), well-sealed, and shaken overnight at 40 °C. After this time, the sample was placed in ultrasonic bath for 15 min and then centrifuged (5 min, 13 000 rpm). Four aliquots of the supernatant were diluted 10 times with 40 mM NaHCO3, and the obtained solutions were pipetted in a 96-well plate. The intensity of the fluorescence signal was measured with a plate reader (excitation 480 nm, emission 520 nm). The concentration of fluorescently labeled amino groups was determined by comparison with a calibration curve (Figure S-3). NNH2 was calculated by dividing the concentration of fluorescently labeled amino groups by the Au concentration. Coupling of Peptide 5 to AuNPs. In a typical experiment, a 200 μL aliquot of a ∼1-5 nM solution of AuNPs in 40 mM NaHCO3 was placed in a ultrafiltration device, and the electrolyte solution was exchanged to 100 mM sodium phosphate buffer, pH 7.0. The residue was taken up in 650 μL of 100 mM sodium phosphate buffer, pH 7.0, and 20 μL of a 50 mM solution of linker 6 in DMF was added. The reaction mixture was shaken for 2 h at 4 °C. After this time, the AuNPs were purified from excess crosslinker via utrafiltration (washed with 2  4 mL 2:8 EtOH /20 mM phosphate buffer pH 7.0 and 2  4 mL H2O) followed by gel filtration eluting with water. The eluate was concentrated via ultrafiltration and taken up in 350 μL 40 mM NaHCO3. Peptide 5 (20 μL, 360 μM in 40 mM NaHCO3) was added to the purified AuNPs and the reaction mixture was shaken overnight at 4 °C. Excess uncoupled peptide was removed by ultrafiltration (washed 1  4 mL H2O, 2  4 mL 2:8 EtOH /40 mM NaHCO3 pH 8.4, 3  4 mL 40 mM NaHCO3 pH 8.4) followed by gel filtration eluting 40 mM NaHCO3. Finally, the eluate was concentrated via ultrafiltration and taken up in 350 μL of 40 mM NaHCO3. The intensity of the fluorescence signal for each filtrate was measured

Langmuir 2009, 25(14), 7910–7917

Article and reached background levels at the end of the purification ensuring that all uncoupled peptide had been removed. As a control experiment, to investigate direct binding of peptide 5 to the AuNPs, an aliquot of passivated AuNPs was treated as described before, however omitting linker 6 during AuNP functionalization (20 μL of DMF was instead used).

Determination of the Number of Coupled Peptides per Particle (Npept). 100-200 μL of peptide-functionalized AuNPs were added of one volume of DTT-PB buffer and further treated as reported above for the determination of NNH2. Four aliquots of the supernatant were diluted 10 times with 40 mM NaHCO3, and the obtained solutions were pipetted in a 96-well plate. The intensity of the fluorescence signal was measured with a plate reader (excitation 546 nm, emission 586 nm). The concentration of peptide was determined by comparison with a calibration curve (Figure S-4). Npept was calculated by dividing the concentration of peptide by the Au concentration measured by ICP-OES or AA.

Acknowledgment. The authors thank Mr. Albrecht Meyer and Mr. Gerhard Werner (ZWE-Analytische Chemie, MPI for Metals Research, Stuttgart) for the ICP-OES and AA measurements and Dr. Claudia Pacholski, Dr. Richard Segar, and Ms. Carina Frauer for proof-reading the manuscript and for helpful discussion. L.M. acknowledges the Studienstiftung des Deutschen Volkes (Germany) for a fellowship. The Max Planck Society is acknowledged for financial support. Supporting Information Available: UV-vis spectra and SEM micrographs for AuNPs, calibration curves for the fluorescence measurements, estimation of the uncertainty affecting NNH2, graphical representation of NNH2 as a function of the AuNP surface area, NNH2 values as a function of particle aging, and gel electrophoresis of passivated AuNPs. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la900545t

7917