pubs.acs.org/Langmuir © 2010 American Chemical Society
Exploring the Surface Charge on Peptide-Gold Nanoparticle Conjugates by Force Spectroscopy Ariel R. Guerrero,†,‡ Leonardo Caballero,§ Alberto Adeva,^ Francisco Melo,*,§ and Marcelo J. Kogan*,‡ †
Materials and Surface Science Group, University of Windsor, 401 Sunset Avenue, Windsor, Ontario N9B 3P4, Canada, ‡Departamento de Quı´mica Farmacol ogica y Toxicol ogica, Facultad de Ciencias Quı´micas y Farmac euticas, Universidad de Chile, Olivos 1007, Independencia, Santiago, Chile, §Departamento de Fı´sica, Universidad de Santiago de Chile, Av. Ecuador 3493, Estaci on Central, Santiago, Chile, and ^Servicios Cientı´fico-T ecnicos, Universidad de Barcelona, Baldiri i Reixac 10-12, 08028 Barcelona, Spain Received September 7, 2009. Revised Manuscript Received May 21, 2010
The conformation and charge exposure of peptides attached to colloidal gold nanoparticles (AuNPs) are critical for both the colloidal stability and for the recognition of biological targets in biomedical applications such as diagnostics and therapy. We prepared conjugates of AuNPs and three isomer peptides capable of recognizing toxic aggregates of the amyloid beta protein (Aβ) involved in Alzheimer’s disease, namely, CLPFFD-CONH2 (i0), CDLPFF-CONH2 (i1), and CLPDFF-CONH2 (i2), where D is the amino acid aspartic acid that is negatively charged at pH = 7.4. We then studied the effect of peptide sequence on the charge exposure through force spectroscopy measurements. The peptide-AuNPs conjugates were fixed on glass surfaces, and their interactions with peptide-functionalized tips were determined. Our results show a higher density of surface charge in the conjugates of the isomers i0 and i2 and a lower density in i1, which is due to the higher degree of functionalization in the first two compared with the third. However, the charge per molecule of the peptide is higher for i1 with respect to i0 and i2, which could be related to the local conformation that the peptides adopt on the surface. The acid-base behavior of the peptide anchored to the AuNPs is different than expected in aqueous solutions of free peptides, which could be related to the low accessibility of the NH2-terminal group belonging to the cysteine that is located near the AuNPs surface. In contrast with other techniques, the fixation of the peptide-AuNPs conjugates to a surface allows for characterization of the local charge exposure of peptides anchored to AuNPs over a wide range of pH.
Introduction Nanoparticles (NPs) are currently under investigation for biomedical applications by a growing number of groups around the world. The small size of these particles makes them suitable for manipulation of subcellular structures. Various types of NPs have been explored for biomedical applications, and they have been widely employed in biological systems. NPs have been suggested for imaging, screening, and biosensing because of both their optical and electrical properties.1-3 Gold nanoparticles (AuNPs) are known to carry higher payloads of drugs than other vehicles and are therefore currently being used in gene and drug delivery as well as in cancer diagnostics and therapeutic applications.4,5 Furthermore, AuNPs have several features that make them wellsuited for biomedical applications including their straightforward synthesis, stability, and ability to easily incorporate secondary tags such as peptides targeted to specific cell types, which allows for selectivity. On the other hand, AuNPs have been used in photothermal therapy for the destruction of cancer cells or tumors. When irradiated with a focused laser in the near-infrared region (NIR) of suitable wavelength, aggregates of AuNPs, nanorods, or *Corresponding authors. E-mail:
[email protected] (F.M.).
[email protected] (M.J.K.);
(1) Kogan, M. J.; Olmedo, I.; Hosta, L.; Guerrero, A. R.; Cruz, L. J.; Albericio, F. Nanomedicine (London, U. K.) 2007, 2(3), 287–306. (2) Cai, W. B.; Cao, T.; Hong, H.; Sun, J. T. Nanotechnol., Sci. Appl. 2008, 1, 17–32. (3) Liong, M.; Lu, J.; Kovochich, M.; Xia, T.; Ruehm, S. G.; Nel, A. E.; Tamanoi, F.; Zink, J. I. ACS Nano 2008, 2(5), 889–896. (4) Hosta, L.; Pla-Roca, M.; Arbiol, J.; Lopez-Iglesias, C.; Samitier, J.; Cruz, L. J.; Kogan, M. J.; Albericio, F. Bioconjugate Chem. 2009, 20(1), 138–146. (5) Ghosh, P.; Han, G.; De, M.; Kim, C. K.; Rotello, V. M. Adv. Drug Delivery Rev. 2008, 60(11), 1307–1315.
12026 DOI: 10.1021/la1014237
nanoshells can kill bacteria and cancer cells.6,7 In a recent investigation, Kogan et al.8-10 used peptide-functionalized (i.e., conjugated) AuNPs with selectivity for the toxic aggregates of amyloid beta protein (Aβ) involved in Alzheimer’s disease, employing microwaves to disaggregate the toxic aggregates. These AuNPs were capped with the peptide CLPFFD-CONH2 (i0), which provides specificity for the Aβ toxic aggregates. Once the conjugates were bound to the toxic aggregates, a radio frequency was applied for several hours, leading to disassembly of the toxic aggregates by the local heating produced by the interaction of the microwaves with the AuNPs. From the point of view of recognition, delivery, and selectivity for the biological target (i.e., the toxic aggregates of Aβ in this last case) by the peptide-AuNPs, one of the main concerns is how peptides interact and pack on the AuNPs surface, which influences the peptide conformation and charge exposure.1,3,11-13 In a recent work,11 we studied (6) Zharov, V. P.; Mercer, K. E.; Galitovskaya, E. N.; Smeltzer, M. S. Biophys. J. 2006, 90(2), 619–627. (7) Zharov, V. P.; Galitovskaya, E. N.; Johnson, C.; Kelly, T. Laser Surg. Med. 2005, 37(3), 219–226. (8) Kogan, M. J.; Bastus, N. G.; Amigo, R.; Grillo-Bosch, D.; Araya, E.; Turiel, A.; Labarta, A.; Giralt, E.; Puntes, V. F. Nano Lett. 2006, 6(1), 110–115. (9) Araya, E.; Olmedo, I.; Bastus, N. G.; Guerrero, S.; Puntes, V. F.; Giralt, E.; Kogan, M. J.. Nanoscale Res. Lett. 2008, 3(11), 435–443. (10) Bastus, N. G.; Kogan, M. J.; Amigo, R.; Grillo-Bosch, D.; Araya, E.; Turiel, A.; Labarta, A.; Giralt, E.; Puntes, V. F. Mater. Sci. Eng., C 2007, 27(5-8), 1236–1240. (11) Olmedo, I.; Araya, E.; Sanz, F.; Medina, E.; Arbiol, J.; Toledo, P.; AlvarezLueje, A.; Giralt, E.; Kogan, M. J. Bioconjugate Chem. 2008, 19(6), 1154–1163. (12) Gratton, S. E. A.; Ropp, P. A.; Pohlhaus, P. D.; Luft, J. C.; Madden, V. J.; Napier, M. E.; DeSimone, J. M. Proc. Natl. Acad. Sci. U.S.A. 2008, 105(33), 11613–11618. (13) Nativo, P.; Prior, I. A.; Brust, M. ACS Nano 2008, 2(8), 1639–1644.
Published on Web 06/17/2010
Langmuir 2010, 26(14), 12026–12032
Guerrero et al.
Article
Figure 1. Schematic representation of the functionalized AuNPs with the peptides i0 and i1. The total number of peptide molecules per nanoparticles is higher in the first conjugate compared with the last. The goal of this paper was to determine by force spectroscopy the local charge exposure of the peptides anchored to the AuNPs.
the influence of the peptide sequence on the conjugation, functionalization, and affinity for toxic aggregates of Aβ. We explored the effect of changing the i0 sequence on the AuNP peptide conjugation, stability, and affinity to Aβ aggregates. The purpose of this study was to modify the distance between the AuNP surface and the position of the negatively charged hydrophilic residue at pH 7.4 (i.e., the D residue). For recognition purposes, it is important that the hydrophobic residues on the conjugate (in this case the L, F, and F residues) be exposed to Aβ. In our previous study, we chose two isomers of i0, the peptides CDLPFF-CONH2 (i1), and CLPDFF-CONH2 (i2). If the peptide adopts an extended conformation, the D residue will localize away from the gold surface in the case of AuNP-i0, while it will be near the AuNP surface in the case of AuNP-i1. From the point of view of molecular recognition, the level of local charge exposure and the conformation of the peptide anchored to the gold surface are important pieces of information that remain undetermined. The peptide sequence influences the degree of conjugation and stability of AuNP-peptide conjugates14 as well as their interactions with binding targets. For example, for AuNP-i0 and AuNPi2, a higher number of peptide molecules cover the surface than for AuNP-i0 (460 ( 30 vs 200 ( 5) (Figure 1) as we described previously for 13 nm diameter nanoparticles.11 Zeta potential measurements demonstrated that the total charges of AuNP-i0 and AuNP-i2 were higher compared with AuNP-i1 due to their higher degree of functionalization. However, because of the low stability of the particles at extreme pHs or ionic strengths, zeta potential measurements and other techniques such as ion exchange chromatography or classical titrations can determine neither the local charge nor the level of exposure of hydrophobic residues. These parameters are crucial for the nanoparticle’s ability to participate in biological recognition interactions. Atomic force microscopy (AFM) has proven to be a very useful tool for imaging nanoscale objects and is quickly becoming one of the most employed scanning probe microscopy (SPM) techniques. Force spectroscopy (FS), based on atomic force principles, has also been developed in the past decade to study, for instance, (14) 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(32), 10076–10084. (15) Bertz, M.; Rief, M. J. Mol. Biol. 2008, 378(2), 447–458. (16) Budich, C.; West, J.; Lampen, P.; Deckert, V. Anal. Bioanal. Chem. 2008, 390(5), 1253–1260. (17) Bura, E.; Klimov, D. K.; Barsegov, V. Biophys. J. 2008, 94(7), 2516–2528. (18) Cao, Y.; Yoo, T.; Zhuang, S. L.; Li, H. B. J. Mol. Biol. 2008, 378(5), 1132–1141. (19) Cao, Y.; Li, H. B. Nature Nanotechnol. 2008, 3(8), 512–516. (20) Chiovitti, A.; Heraud, P.; Dugdale, T. M.; Hodson, O. M.; Curtain, R. C. A.; Dagastine, R. R.; Wood, B. R.; Wetherbee, R. Soft Matter 2008, 4(4), 811–820. (21) Chtcheglova, L. A.; Atalar, F.; Ozbek, U.; Wildling, L.; Ebner, A.; Hinterdorfer, P. Pfl€ ug. Arch. Eur. J. Phys. 2008, 456(1), 247–254. (22) Dabrowska, A.; Lebed, K.; Kulik, A. J.; Forro, L.; Lekka, M. Acta Phys. Pol., A 2008, 113(2), 753–762. (23) Decker, B.; Kellermayer, M. S. Z. J. Mol. Biol. 2008, 377(2), 307–310.
Langmuir 2010, 26(14), 12026–12032
Figure 2. Sketch of the experimental configuration based on FS for charge measurements. The goal was to measure the electrostatic repulsion of charged gold nanoparticles at several pHs.
different processes in proteins,15-24 especially in the single-molecule mode using functionalized tips. Force spectroscopy has also allowed for the measurement of the nanomechanical properties of surfaces interacting with probes.25 The intermolecular forces acting at interfaces on a nanometer scale play a key role in a wide range of chemical, biological, and physical processes including chemical and physical absorption, wetting, wear, catalysis, adhesion, and cell recognition. The interactions involved, especially when measured in a liquid environment, can mainly be explained in terms of van der Waals forces, electrostatic Coulombic interactions, solvation forces, and hydrogen bonding.26-28 Interaction forces arising between the target substrate and a silica sphere or a silica-coated sphere mounted on an AFM cantilever that are measured as functions of the pH and ionic strength have been extensively used to examine the physical properties of different substrates and selfassembled monolayers (SAMs). These studies have yielded valuable information about the electrostatic properties of the surface pKa. The use of thiol chemistry and the strong covalent S-Au bonding allows for the formation of self-assembled monolayers with different functionalities on gold-coated cantilevers. Thus, the interaction forces between a chemically derivatized surface and the tip can be studied as a function of the chemical composition. Hence, this so-called “chemical force microscopy” operation method has been used to study the interactions between proteins and ligands. When the surface, the tip, or both surface and tip are functionalized with an ionizable group, the method has been referred to as “force titration”.29 In this work, we use atomic force techniques to assess the local electrical charge density and the exposure of the hydrophobic residues of isomer peptides conjugated to AuNPs at different pHs. The peptides are modified with an amide at the C-terminal (24) Denisov, N. N.; Chtcheglova, L. A.; Sekatskii, S. K.; Dieer, G. Colloids Surf., B 2008, 63(2), 282–286. (25) See for instance : Bushan, B., Ed.; Nanotribology and Nanomechanics: An Introduction; Springer: Heidelberg, 2005. (26) Israelachvili, J. Intermolecular and Surfaces Forces, 2nd ed.; Academic Press: London, 1991. (27) Butt, H. J. Biophys. J. 1992, 63(2), 578–582. (28) Butt, H. J. Biophys. J. 1991, 60(4), 1438–1444. (29) Garcia-Manyes, S.; Gorostiza, P.; Sanz, F. Anal. Chem. 2006, 78(1), 61–70.
DOI: 10.1021/la1014237
12027
Article
Guerrero et al.
position to avoid the contribution of the carboxyl group to the acid-base equilibrium. A modified tip is used to measure the interaction force between a microsphere and a surface that is nearly homogeneously covered by conjugates (Figure 2). Peptide AuNP conjugates are attached to a glass substrate using 3-mercaptopropylmethoxysilane (MPTMS) molecules that link to the AuNPs. The charge of the peptide-AuNP conjugates was investigated as a function of pH for all the conjugates independently of colloidal stability. The advantage of this method lies in the fact that peptide-AuNP conjugates cover a flat substrate, providing a well-defined electrical surface for interaction with the microsphere-cantilever gauge. In the case of AuNP-peptide dispersions, note that surface electrical charge measurements using classical zeta potential methods become less practical because peptide-AuNP conjugates have a strong tendency to agglomerate depending on the pH and ionic strength.
Experimental Section Synthesis of AuNPs. Citrate-coated AuNPs (15.1 ( 2.2 nm) were prepared by citrate reduction of HAuCl4 according to ref 30. An aqueous solution of HAuCl4 (100 mL, 1 mM) was refluxed for 5-10 min, and a warm (50-60 C) aqueous solution of sodium citrate (10 mL, 38.8 mM) was added quickly. Reflux was continued for another 30 min until a deep red solution was obtained. The solution was filtered through 0.45 μm Millipore syringe filters to remove any precipitate, the pH was adjusted to 7.4 using dilute NaOH solution, and the filtrate was stored at 4 C. AuNPs were observed by transmission electronic microscopy (TEM) using a JEOL JEM-1010 microscope (Supporting Information). The specimen was prepared by dropping AuNPs on Formvar carboncoated copper microgrids and letting them dry. Peptide Synthesis. CLPFFD-CONH2, CDLPFF-CONH2, and CLPDFF-CONH2 were synthesized following a fluorenylmethyloxycarbonyl (Fmoc) solid-phase synthesis strategy. Fmocprotected amino acids were purchased from Novabiochem (Laufelfingen, Switzerland) and PerSeptive Biosystems (Framingham, MA). O-(Benzotriazol-1-yl)-N,N,N0 ,N0 -tetramethyluronium tetrafluoroborate (TBTU), Fmoc-AM handle, and resin MBHA were also obtained from Novabiochem. Chemical reagents N,N0 diisopropylcarbodiimide (DIPCI), 1-hydroxy-1H-benzotriazole (HOBt), triethylsilane, and (dimethylamino)pyridine (DMAP) were from Fluka (Buchs, Switzerland). Manual synthesis included the following steps: (i) resin washing with DMF (5 30 s); (ii) Fmoc removal with 20% piperidine/DMF (1 1 min þ 2 7 min); (iii) washing with DMF (10 30 s); (iii) DMF washing (5 30 s). Cleavage of the peptide was carried out by acidolysis with trifluoroacetic acid (TFA) using triethylsilane and water as scavengers (94:3:3, v/v/v) for 60-90 min. TFA was removed with a N2 stream, and the oily residue was precipitated with dry tertbutyl ether. Crude peptide was recovered by centrifugation and decantation of the tert-butyl ether phase. The solid was redissolved in 10% acetic acid (AcH) and lyophilized. The peptide was analyzed by RP-HPLC [Waters 996 photodiode array detector (λ = 443 nm) equipped with a Waters 2695 separation module (Milford, MA), a symmetry column (C18, 5 μm, 4.6 150 mm), and Millennium software; flow rate: 1 mL/min, gradient: 5-100% B over 15 min (A) 0.045% TFA in H2O, and (B) 0.036% TFA in acetonitrile)]. The peptide was purified by semipreparative RP-HPLC [Waters 2487 dual absorbance detector equipped with a Waters 2700 sample manager, a Waters 600 controller, a Waters fraction collector, a Waters symmetry column (C18, 5 μm, 30 100 mm), and Millennium software]. The peptide was finally characterized by amino acid analysis using the AccQ.Tag method (Waters) in a Waters Delta 600 HPLC system equipped with a UV Waters 2487 detector and by (30) Turkevich, J.; Stevenson, P. C.; Hillier, J. Discuss. Faraday Soc. 1951, 11, 55–75.
12028 DOI: 10.1021/la1014237
MALDI-TOF with a Bruker model Biflex III. Amino acid analysis of CLPFFD-CONH2: Asp 1.0 (1), Pro 0.97 (1), Leu 1.0(1), Phe 2.03 (2), CDLPFF-CONH2: Asp 1.1 (1), Pro 1.0 (1), Leu 0.9 (1), Phe 2.0 (2), CLPDFF-CONH2: Asp 1.0 (1), Pro 1.0 (1), Leu 1.0(1), Phe 2.1 (2) and mass spectrometry MALDI-TOF: for CLPFFD-CONH2, CDLPFF-CONH2, and CLPDFF-CONH2 [M þ Hþ] 740 and [M þ Naþ] 762. Stock solutions of the peptides were prepared by dissolution in water. The pH was adjusted to 7.4 using concentrated NaOH, and the resulting 1.3 mM solution was filtered using 0.2 μm PVDF filters and kept as aliquots at -80 C. Amino acid analysis indicated a final concentration of 1.28 mM after filtering. Conjugation of Peptides to AuNP. CLPFFD-CONH2 (i0), CLPDFF-CONH2 (i1), and CDLPFF-NH2 (i2) capped AuNPs were prepared by mixing 5 nM AuNP and peptide stock solution in a volume ratio of 10 to 1. The conjugation was performed in the presence of excess peptide in order to ensure full conversion of the AuNP, thus ensuring homogeneous conjugation. The AuNPs were then purified, first with a 450 nm filter and then by dialysis (during 3 days in a membrane Spectra/Por MWCO 6-8000, dialyzed against 1.2 mM sodium citrate with six solution changes) to eliminate the excess peptide. UV-vis absorption spectra were recorded at room temperature with a Unicam UV/vis spectrophotometer (UV3). In order to check for the absence of nonconjugated peptide, 3 mL of the conjugated solution was centrifuged at 13 500 rpm for 30 min (AuNP peptide sediments), the supernatant was evaporated to dryness, and amino acid analysis was carried out. The absence of free peptide both in the supernatant and in the NP pellet (the conjugated AuNP) was also checked by HPLC ES-MS. The conjugates were exhaustively characterized by using UVvis spectrophotometry, electron energy loss spectra (EELS), X-ray photoelectron spectroscopy (XPS), amino acid analysis, gel electrophoresis, and CD according to ref 11. In the Supporting Information UV-vis results are given.
Estimation of the Number of Peptide Molecules per AuNP. The amount of peptide molecule per NP was estimated by analysis of amino acids and absorption spectrophotometry. The concentration of AuNP in the solutions was obtained taking into account the molar coefficient of extinction of the 15 nm diameter AuNP (5.7 107 M-1 cm-1) and an analysis of amino acids of the pellet obtained after centrifugation of the conjugates at 13 500 rpm for 30 min (in such conditions the NP sediment). The number of peptide molecules per AuNP was obtained by dividing the number of peptide molecules per milliliter of solution by the number of particles per milliliter of solution. This ratio was obtained in triplicate in three independent synthesis and conjugation. The degree of conjugation follows the order AuNP-i0 > AuNP-i2 > AuNP-i1 (560 ( 30, 520 ( 10, and 200 ( 5 peptide molecules per AuNP). AFM Experiments. All AFM experiments were performed using a Nanoscope IIIa (Digital Instruments, Inc.) scanning probe microscopy instrument. Images were acquired at a 1 Hz scanning rate at 512 512 pixels of resolution using model A (1 1 μm) and model E (15 15 μm) scanners. Sample Preparation. Glass slides (12 12 mm) were rinsed in royal water (nitric acid/hydrochloric acid 1:3) to remove gold debris and later in piranha solution (hydrogen peroxide/sulfuric acid 1:4) to remove organic matter. After rinsing with piranha solution, the glass slides were thoroughly rinsed with Milli-Q water (18.2 MΩ 3 cm) and dried under a nitrogen stream. The conjugates were fixed to the glass slide using a protocol adapted from Bonanni and Cannistraro31 in which 3-mercaptopropylmethoxysilane (MPTMS; Aldrich, 95%) was dissolved in chloroform at a concentration of 0.03 M. A 30 μL droplet was added to the glass slide, and after 3 s, the excess MPTMS was removed by (31) Bonanni, B.; Cannistraro, S. AZojono ; Journal of Nanotechnology Online 2005, DOI: 10.2240/azojono0105; URL: http://www.azonano.com/details.asp?ArticleID=1436.
Langmuir 2010, 26(14), 12026–12032
Guerrero et al.
Article
washing with copious chloroform, and the slide was then dried under a nitrogen stream. The conjugate was then added by depositing a droplet of the colloid onto the slide. After 15 min, the excess colloid was removed by washing with Milli-Q water. A droplet of water was left on the slide for an additional 5 min, and then a final wash with Milli-Q water was completed by drying with nitrogen. Tapping-Mode Imaging. Tapping-mode-air AFM images are suitable for the characterization of the distribution of conjugate AuNPs on the glass slide. Measurements were performed in tapping mode using Veeco cantilevers (Model TESPA, antimony (n) doped Si) whose stiffness, resonance frequency, and tip radius were 40 N/m, 300 kHz, and 10 nm, respectively. Force Experiment. Gold-coated, modified cantilevers with gold-coated glass beads (5 μm diameter) at their tips (Novascan, model GS-PT-AU) were etched in a sputtering chamber for ∼10 s to clean the active surface of the sphere. The cantilevers were immediately functionalized with the peptide by dipping them into a 100 nM solution of the corresponding peptide. The cantilever stiffness was determined by using the thermal fluctuations method and was typically 0.1 N/m for the cantilevers used. The experimental principle used to measure the peptide-against-peptide AuNP electrostatic repulsion is sketched in Figure 2. The experiment was performed in a FC (Veeco) fluid cell under contact mode. To avoid errors in the force measurements introduced by cantilever stiffness uncertainty, all force measurements for a given peptide were taken with the same cantilever. In addition, at least 15 different sample locations were tested for each pH value. The experiments were done three independent times on different days with different samples. To make measurements over a range of pH values, the pH was potentiometrically adjusted by injecting dilute aqueous solutions of hydrochloric acid and sodium hydroxide into the fluid cell using a syringe pump. The pH was adjusted to 3, 4, 5, 6, 7, 8, 9, and 10 for FS measurements. Following the DLVO theory,26 the total interaction force measured on approach was assumed to be a linear summation of an attractive van der Waals component and a repulsive electrostatic double-layer component: FðDÞ ¼ Fdouble-layer ðDÞ þ FvdW ðDÞ
ð1Þ
The nonretarded van der Waals force due to the interaction of the microsphere and a monolayer of AuNP conjugates that is assumed to be perfectly flat is approximated by an inverse power law26 FðDÞ ¼ - A
R 6D2
ð2Þ
where D is the vertical (Z-axis) tip-sample distance, A is the Hamaker constant, and R is equal to the radius of the colloidal sphere located at the tip. The electrostatic force obeys the Poisson-Boltzmann equation28 that describes the interaction between the planar surface of the nanoparticles and the spherical probe tip when they have a constant surface charge per unit area in an electrolyte solution. Thus, the total interaction force is modeled as FðDÞ ¼
4πRσb σs - KD R e -A 2 Kε 6D
ð3Þ
where σb is the surface charge density of the spherical probe, σs is the effective surface density charge in the nanoparticle monolayer, ε is the relative permittivity of the medium (in this case, water), and the reciprocal of κ (κ-1) is the Debye screening length. Since the surface sample and the cantilever sphere are both peptideconjugated gold, we assume σs and σb to be equal. Thus, the effective electrostatic repulsion and the effective van der Waals attraction were measured by placing the AFM in ramp mode and fitting the experimental force curves to eq 3. A statistical analysis Langmuir 2010, 26(14), 12026–12032
Figure 3. Tapping mode height AFM images of the conjugates peptide-AuNPs on a glass slide fixed with MPTMS: (a) AuNP-i0, (b) AuNP-i1, and (c) AuNP-i2. White scale bars represent 300 nm. of all experiments was performed to obtain the averages and standard deviations of all results. Note again that formula 3 is valid for flat substrates. To take into account the effect of local surface curvature due to AuNP coverage on the glass piece, a geometrical factor can be incorporated, as discussed below. In addition, it is important to notice that eq 3 remains valid when the Debye length is larger than the typical surface roughness. In our case, this constraint is relatively well-fulfilled for high pH, where the Debye length is about 30 nm. However, it is observed that for pH smaller than 4, the Debye length is on the order of surface roughness, and the results produced by eq 3 can be considered as only rough estimates of charge distribution. Thus, at low pH, the colloidal probe is no longer suitable to detect electrostatic repulsion on AuNPs. To overcome this difficulty and to provide more insight about the peptide charges at low pH, we have performed an additional experiment in which the AuNps are replaced by a layer of Au that is deposited onto glass slides. The gold sample was prepared by evaporation in a vacuum chamber at 7.5 10-6 Torr on glass with 2 nm of Cr (99.996%, Alfa Aesar) and 30 nm of gold (99.99%, Aldrich). The width was measured and controlled with a quartz balance. The pressure of the chamber was recovered by injection of argon or nitrogen to avoid contamination of the gold surface. Peptide solutions (100 nM) were deposited by dipping the gold surface on the peptide solution. The 100 nM peptide solution concentration was chosen to avoid the multilayer peptide deposit on the surface observed at higher peptide concentrations. After chemisorption to the surface, it was washed with Milli-Q water to eliminate excess of peptide. Finally, the surface was dried with nitrogen.
Results AFM Imaging. Figure 3 shows tapping-mode AFM images of the peptide-AuNP conjugates fixed to a glass slide in air. The AuNPs were fixed on the substrate by interaction with MTPMS, which contains a thiol group allowing for the formation of a stable S-Au bond. The images show little aggregation of the AuNPs. They present a relatively flat surface of a roughness that is compatible with the presence of AuNPs. Further analysis of height profiles indicates that the typical distance between the two nearest objects is consistent with the AuNP diameter. In addition, the vertical roughness of the surface (measured by the mean quadratic dispersion) does not exceed 4 nm in any case, indicating that AuNPs produce a relatively homogeneous monolayer on the surface at the scale of the size of the colloidal probe. FS Experiment. Figure 4 shows typical force plots in an aqueous medium for the approach of the functionalized microsphere to the three different conjugated AuNPs covering the sample surface. In all cases, the electrostatic repulsion appears gradually with an increase in pH. All curves show the presence of the van der Waals attractive force, which is responsible for the abrupt change in the force that takes place right before the sample makes contact with the sphere (see Figure 4 and Figure S4, Supporting Information). On the other hand, retracting curves were obtained for each conjugate at different pHs. However, there were no particular trends to distinguish each conjugate (Supporting Information Figure S5). DOI: 10.1021/la1014237
12029
Article
Guerrero et al.
Figure 4. Typical curves of the interaction between the functionalized microsphere and the conjugate peptide-AuNPs that homogeneously cover the sample at several pH values.
Figure 5. (a) Charge surface density for AuNP peptide conjugates obtained with the isomers i0, i1, and i2 as a function of pH. (b) Charge per molecule obtained with the isomers i0, i1, and i2 as a function of pH.
A simple model of the cantilever moving in a nonlinear force field shows that the “jump-into-contact” distance, Djump_contact, can be used to find an estimate of the Hamaker constant.32 Notice that the Djump_contact results from the mechanical instability of the cantilever and corresponds to the point at which the gradient of the force exceeds the cantilever spring constant (dF/dD > k). Thus, the Hamaker constant is determined from the sample-tip distance at the “jump into contact”, using the spring constant of the cantilever k and the tip radius of curvature R, according to31 Djump
contact
¼
AR 3k
1=3
This formula is valid for sufficiently low or screened electrostatic charge, which, in our case, occurs at relatively low pH depending on the considered peptide. To estimate the Hamaker constant, the jump to contact distance has been averaged over 10 data samples taken at different locations for each pH value. The typical value of the Hamaker constant is about 40 10-20 J, which is of the order of the value for naked gold. This indicates that the van der Waals interaction is dominated by the metallic substrate instead of the peptide layer.33 Figure 5 show the charge densities obtained from the fit of the electrical force for i0, i1, and i2 peptide-AuNPs as a function of pH. Charges values are obtained from 10 data samples for each pH. The experiments show increasing surface charge values for all peptides as the pH increases. Experiments with different samples showed similar results. To fit eq 3 to the data, the Hamaker constant for all pH values was (32) Vandiver, J.; Dean, D.; Patel, N.; Botelho, C.; Best, S.; Santos, J. D.; Lopes, M. A.; Bonfield, W.; Ortiz, C. J. Biomed. Mater. Res., Part A 2006, 78(2), 352–63. (33) See for instance: Butt, H. J.; Capella, B.; Kappl, M. Surf. Sci. Rep. 2005, 59 (1-6), 1–152.
12030 DOI: 10.1021/la1014237
fixed to the average value obtained with the jump to contact method that was calculated at a pH for which electrostatic repulsion was neglected (Figure S6, Supporting Information). Despite the dispersion on the Hamaker constant, because at relatively large sample-tip distances the van der Waals attraction is rather small compared with electrical repulsion, the resulting fitted values for charges densities are not significantly affected by uncertainties in the Hamaker constant. As a trend, the total charge values for the i0 and i2 conjugates are larger than those of the i1 conjugates, which is due to the higher degree of functionalization in the first two. However, at pHs higher than 7, the charge per peptide molecule on the surface is higher for i1 than for the other conjugates (Figure 5b). Figure S7 (Supporting Information) depicts the electrical charges as functions of the pH exhibited by peptides deposited onto a glass slide.
Discussion The total charge values for the i0 and i2 conjugates are larger than those exhibited by i1 conjugates, which is due to the higher degree of functionalization in the first two. The degree of functionalization of the AuNPs was determined both by the hydrolysis of the three peptide AuNP conjugates and by their absorbance. The results indicated that the number of peptides on AuNPs is similar for conjugates i0 and i2 (about 560 and 520 molecules/NP, respectively), whereas there are ∼50% fewer in the case of i1 (about 200 molecules/NP). It is important to mention that the density of functionalization of AuNPs is approximately 0.3-0.8 peptide molecules/nm2 (calculated by considering the surface of a sphere of 15 nm diameter (706 nm2) and the number of peptide molecules per nanoparticle), which is on the order of a density expected for a peptide SAM on a gold surface.34 (34) Arikuma, Y.; Takeda, K.; Morita, T.; Ohmae, M.; Kimura, S. J. Phys. Chem. B 2009, 113, 6256–6266.
Langmuir 2010, 26(14), 12026–12032
Guerrero et al.
The pH plays an important role in the charge exposure of the AuNP-peptide conjugates, as it directly affects the conjugation and stability of the colloidal particles and the biological recognition of the toxic aggregates of Aβ.11 Surprisingly, the charge-versus-pH behavior of the peptides attached to the AuNP surface is different from what would be expected from free peptides. In the conjugates, the charge was not detectable below pH ∼ 4, but in the free peptides the isoelectric points are around 6 (obtained using the software Insight II). Thus, the charge is around zero between pH 5 and pH 7. This would indicate that the amine group that belongs to the cysteine in the conjugate does not contribute to the acid-base behavior, whereas the carboxylic group belonging to the lateral chain of the aspartic group (pKa ∼ 3.9) likely governs this behavior. As was demonstrated in the self-assembled monolayer of mercaptotetradecanoic acid, the altered surface acidity could be attributable to interactions of the molecules with the gold surface.35 In the case of peptides, the low exposure of the amine group that belongs to the N-terminal cysteine (which is bound directly to the gold that forms the monolayer on the surface) means that it is not accessible to ions (in our case, hydroxyls and protons). Thus, this group does not likely participate in the equilibrium acid-base. The monolayer on the surface can provide an excellent barrier for hydrated ionic probe penetration, as was demonstrated for a 2-mercapto3-n-octylthiophene SAM.36 On the other hand, in SAMs on nanoparticle surfaces, it was demonstrated that ions do not enter the monolayer without an electrostatic driving force between the core and bulk solution.37 In addition, by considering the average number of peptides per AuNP and the geometrical factor previously discussed, the charge per molecule of peptide on the AuNP-peptide conjugates can be obtained (Figure 5b). A tiny fraction of the carboxyl groups are charged, which could be attributed to a low exposure of these groups to the solvent due to the formation of the SAM that avoid the lateral carboxyl groups of the aspartic acid to the solvent reducing the ion permeability. In the case of i1, the number of charges per peptide molecule is higher with respect to the other conjugates. We hypothesize that in this case molecules are deposited horizontally on the gold surface due to the interaction of the phenylalanine residues with the gold surface, as was suggested previously.11 Thus, the chain molecules are fixed on the surface, exposing the aspartic groups to the aqueous medium (Figure 6). In contrast, the i0 and i2 peptide chain chemisorbed molecules are extended orthogonally with respect to the surface. This allows a certain degree of flexibility of the chains that may cause intermolecular interactions, which could mask the carboxylic groups belonging to the D residues and thus mask their exposure to the medium (Figure 6). On the other hand, regarding the potential of the experimental method to be applied elsewhere, it is important to emphasize that even for the same degree of AuNP coverage on the glass surface, the electrical force measurements presented here provide only relative values of the charge of AuNP conjugates. For absolute measurements of the superficial charge on AuNP conjugates, a precise study of the double-layer structure in the intermediate region, defined by the glass plate and the lower boundary of nanoparticles layer, is necessary. Qualitatively, two limiting situations can be encountered depending on the Debye length. For a Debye length that is smaller than the diameter of the AuNPs, electrolyte net charges that are present in the intermedi(35) Leopold, M. C.; Black, J. A.; Bowden, E. F. Langmuir 2002, 18, 978–980. (36) Peng, Z.; Dong, S. Langmuir 2001, 17, 4904–4909. (37) Laaksonen, T.; Pelliniemi, O.; Quinn, B. M. J. Am. Chem. Soc. 2006, 128, 14341–14346.
Langmuir 2010, 26(14), 12026–12032
Article
Figure 6. Hypothetical model of the disposition of peptide molecules in the conjugated peptide-AuNPs. In part A, peptide molecules are oriented orthogonally with respect to the gold surface. The chain molecules are flexible and their mutual interactions (symbolized by the red arrows) contribute to mask the exposure of carboxylic groups belonging to the aspartic acids. In part B, peptide molecules are horizontally adsorbed on the gold surface. The adsorption fixes the molecules so that their carboxylic acids are exposed to water. The green arrow indicates the exposure of carboxylic groups. Peptide molecules are in an extended conformation represented in a ball model drawn with the GaussView 3.09 (Gaussian, Inc.) software. Oxygen, nitrogen, and sulfur atoms are represented in red, blue, and yellow, respectively.
ate region are sufficient to equilibrate peptide surface charges. Under this condition, the colloidal probe would sense the surface charge located on the upper surface of the AuNP layer. In the opposite limit, i.e., a Debye length that is larger than the diameter of the AuNPs, the number of net charges in this region is small compared with the surface peptide charge. Then, the colloidal probe senses the total charge of the AuNP’s conjugate. In addition, for quantitative interpretation our method requires knowledge of the effective surface area for charge distribution. This geometrical factor can be deduced from an analysis of the images in Figure 3, which show a high surface packing and small vertical variation of nanoparticle positions. For the purposes of this investigation, we estimate the effective surface area by considering two-dimensional hexagonal close packing of AuNps. This results in an excess surface area of π/31/2 with respect to a flat surface. Thus, to obtain the average charge density on the AuNPs, the charge density obtained from eq 1 must be corrected by the factor 31/4/(2π)1/2.38 The factor of 2 comes from the fact that the entire surface of the AuNPs contributes to the sensed force, as we have considered the regime of large Debye length. Studies of charge density (Figure S7 in the Supporting Information) were performed by replacing the AuNP monolayer by a nearly flat gold layer deposited onto a glass slide. These produced equivalent results for peptide charges at high pH (large Debye length), thus providing additional support for the contribution of the bottom half of AuNPs to the total charge. These studies also show that charges can be detected on peptides by the colloidal probe even at low pH when the Debye length is relatively small if substrate roughness is decreased (Figure S7).
Conclusions This investigation has provided a simple alternative method of characterizing surface electrical interactions on peptide-AuNP conjugates, and it has characterized the influence of charge exposure at extreme pH conditions. Linkage of AuNP conjugates to a flat surface through a relatively strong link prevents conjugate agglomeration resulting from van der Waals attractive forces. When the surface roughness achieved by nanoparticle coverage is much smaller than the Debye length, this method (38) This is a phenomenological factor which considers homogeneous charge distribution on nanoparticles when these are in the monolayer configuration. A detailed description of double layer in the lower bottom of AuNps is beyond the scope of this paper.
DOI: 10.1021/la1014237
12031
Article
Guerrero et al.
should allow for assessment of the charges of molecules anchored to metallic nanoparticles. Furthermore, in this paper we learn that the acid-base behavior of the studied peptide conjugates bound to a surface of an AuNP is different than their expected behavior when free in solution. This information should prove useful for the design and optimization of peptide-AuNP conjugates with improved molecular recognition abilities for potential applications in drug delivery and treatment of degenerative diseases.
Note Added after ASAP Publication. This article was published ASAP on June 17, 2010. The Acknowledgment section has been modified. The correct version was published on July 13, 2010.
Acknowledgment. We are very grateful to Professor Christine Ortiz for introducing us to the force spectroscopy methods and to Dr. Claudia Ya~nez for the fruitful discussion related with
Supporting Information Available: Figures S1-S7. This material is available free of charge via the Internet at http:// pubs.acs.org.
12032 DOI: 10.1021/la1014237
monolayers and solvent accessibility. We thank the funding from projects FONDECYT, 1090143, FONDAP 11980002, AECID and ACT-95.
Langmuir 2010, 26(14), 12026–12032
