- Email: [email protected]
Reversible aggregation of gold nanoparticles induced by pH dependent conformational transitions of a self-assembled polypeptide
Journal of Colloid and Interface Science 316 (2007) 977–983 www.elsevier.com/locate/jcis
Reversible aggregation of gold nanoparticles induced by pH dependent conformational transitions of a self-assembled polypeptide Jeung-Yeop Shim, Vinay K. Gupta ∗ Department of Chemical Engineering, University of South Florida, ENB 118, 4202 E Fowler Avenue, Tampa, FL 33620, USA Received 2 July 2007; accepted 8 August 2007 Available online 14 August 2007
Abstract Controlling the stable structures of metallic nanoparticles on mesoscopic and macroscopic length scales is of great interest in nanotechnology. Here, this task is accomplished using a synthetic biopolymer that is responsive to external stimuli and undergoes changes in secondary structure. Reversible aggregation of gold nanoparticles (GNP) is induced by pH dependent changes in a self-assembled monolayer of disulfide modified poly(L-glutamic acid) (SSPLGA) with Mw ∼ 27000. The disulfide anchoring group drives chemisorption onto the gold nanoparticles and leads to the formation of a self-assembled monolayer. Characterization of the modified GNP and its aggregation behavior is performed using dynamic light scattering (DLS), UV–vis and IR spectroscopy, and transmission electron microscopy (TEM). Experimental results show that decrease in pH near 5.5 leads to aggregation of the modified GNP. The change in aggregation behavior with pH occurs within minutes, is reversible, and happens within a narrow range of pH from about 4.5 to 5.5. Comparison with literature data on molar enthalpy of hydrogen bonding, specific optical rotation, and ionization for the helix–coil transition of PLGA indicates that the aggregation of the SSPLGA modified GNP corresponds to the transition in the secondary structure of the polyacid. © 2007 Elsevier Inc. All rights reserved. Keywords: Gold nanoparticles; Reversible aggregation; Self-assembly; Self-assembled polypeptide
1. Introduction The increasing ability to prepare a wide range of nanoparticles of various sizes, shapes, and materials has spurred the need for strategies to assemble these nanomaterials into constructs for functional devices such as sensors, photonic materials, and biomolecular electronics [1–8]. Traditional approaches that rely solely on top–down lithographic methods are of limited use when the nanostructures need to be created either in three-dimensions or on sub-100 nm scales. These approaches also rapidly become energy and materials intensive for twodimensional structures at nanometer length-scales. In this context, directed self-assembly for nanometer scale organization has been recognized as most promising for either enabling hybrid methods in combination with the established lithographic tools or providing affordable and simpler alternatives for nanoscale devices, sensors, and nano/micro fluidic systems * Corresponding author. Fax: +1 813 974 3651.
E-mail address: [email protected] (V.K. Gupta). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.08.021
[6,9–17]. However, the principles for creating, manipulating, and using directed self-assembling systems remain a scientific and technological challenge. In particular, strategies need to be discovered for encoding information that drives assembly of complex motifs, for incorporating molecular interactions that provide dimensional and positional control during assembly, and for including triggers to manipulate the assembled structures. Towards the objective of establishing principles for directed self-assembly of nanoparticles, several approaches have been pursued in recent years for controlling the aggregation of nanoparticles in solution. In these investigations, gold colloids have been a popular choice due to the ease of chemical modification of the gold surface and the fact that surface modification provides sites for covalent, electrostatic, or other inter-molecular interactions that subsequently drive aggregation or flocculation of the nanoparticles. Examples from literature include attachment of thiolated DNA oligomers that cause particle aggregation by undergoing hybridization with complementary molecules, use of ligands that contain termi-
978
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
nal carboxylic or amino groups for electrostatic and hydrogen bonding interactions or for further conjugation for biospecific interactions, and modification of the nanoparticles with an elastin biopolymer that induces interparticle hydrophobic interactions [18–26]. Among the diverse approaches that have been reported in the past few years, a majority of these studies have demonstrated irreversible aggregation and only a few examples of reversible aggregation exist. For the latter, primarily two methods have been used. One method relies on manipulating the colloidal aggregation by changing temperature to either cause thermal denaturation or induce a hydrophilic– hydrophobic phase transition of a grafted polymer [19,23,26, 27]. The second method relies on addition of a reagent that either causes a chemical cleavage of the linkage between particles or changes the pH and/or ionic strength of the solution [20, 21,28,29]. A related approach has also been reported in a system wherein streptavidin-mediated aggregation of biotinylated GNP can be reversed over a period of 80 days by addition of soluble biotin to block the streptavidin [24]. In studies where pH or ionic strength has been varied, the underlying goal has been disruption of hydrogen bonds or electrostatic interactions between modified gold nanoparticles. For example, Fuhrhop and coworkers [20] altered pH to influence hydrogen bonding between simple ω-functionalized alkanethiols such as mercaptoundecanoic acid on the nanoparticles while Fernig et al. [28] used both pH and ionic strength to change electrostatic interactions of a carboxylic acid moiety on the surface of gold nanoparticles. Current years have witnessed a growing focus on combination of synthetic proteins with gold nanoparticles for areas related to biosensors, biomimetics, and biomolecule-directed assembly [17,27,29–33]. Both designer peptides with carefully crafted sequences made by solid-phase peptide synthesis and one-pot materials made by solution-phase polymerization have been successfully used to modify gold nanoparticles surfaces. For example, Feldheim and coworkers [33] have used peptides sequences that mimic virus peptides to modify gold nanoparticles for targeting cells while Higashi et al. [34] have shown that a homopolymer such as poly(γ -benzyl-L-glutamate) can be used to functionalize gold nanoparticles. Towards the particular goal of using surface-tethered peptides to control nanoparticle assembly, recent studies have explored carefully designed peptide sequences for switching particle aggregation by exploiting protein folding and interactions between folded proteins [27–29]. In this context, the present study reports reversible aggregation of the gold nanoparticles that are modified by self-assembly of a simple homopolypeptide, namely, disulfide modified poly(L-glutamic acid) (SSPLGA) that is prepared by polymerization of the N -carboxy anyhydride (NCA). The SSPLGA spontaneously modifies the gold nanoparticle surface via chemisorption of the disulfide headgroup. Cycling the pH changes the aggregation state of the gold colloids in solution from dispersed to aggregated (Fig. 1) and this aggregation corresponds to a change in α-helical conformation of SSPLGA at pH <4 to a random coil conformation at pH >6. To our knowledge, this is the first demonstration where PLGA is employed
Fig. 1. Schematic illustration of the conformation transition from a helix to a random coil for the SSPLGA and the resulting aggregation of gold nanoparticles. The PLGA is modified with a disulfide group that anchors it to the nanoparticle surface.
to reversibly manipulate the state of aggregation of the gold nanoparticles. 2. Experimental 2.1. Materials Hydrogen tetrachloroaurate (HAuCl4 ·3H2 O) and trisodium citrate were purchased from Sigma-Aldrich (WI). A 33 wt% solution of hydrobromic acid in acetic acid (HBr/AcOH) and diethyl ether were purchased from Acros Organics (NJ). Water used in all synthesis was purified using an EasyPure™ UV system (Barnstead, IA). A 0.2 µm filter incorporated into this system removed particulate matter. 2.2. Synthesis of gold nanoparticles The synthesis methodology of gold nanoparticles is based on Frens’s method [35]. A solution HAuCl4 ·3H2 O (40 mg, 0.102 mmol) in deionized water (100 ml) was brought to a boil and citrate trisodium salt (100 mg) was added in one portion. The solution was refluxed for 30 min and then allowed to cool to room temperature. The resulting reddish solution of the GNP was stored in refrigerator for future use. 2.3. Preparation of SSPLGA A disulfide modified γ -benzyl-L-glutamate (SSPBLG) was converted to SSPLGA by debenzylation using HBr/AcOH. The preparation and characterization of SSPBLG has been reported elsewhere [36]. For the debenzylation, 100 mg of SSPBLG (Mn ∼ 22700, Mw ∼ 27000) was dispersed in warm AcOH (5 ml) and 3 ml of the HBr/AcOH solution was added dropwise. The solution was stirred at 50 ◦ C for 12 h. The resulting mixture was diluted with ether (30 ml) and sedimentation of crude
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
979
SSPLGA was allowed to occur overnight at room temperature. The solution was centrifuged for collection of SSPLGA, which was then washed with cold ether (30 ml × 5 times) and dried to give a SSPLGA (48 mg). The debenzylation was verified using IR spectroscopy of the bulk material in a KBr pellet. For control experiments, conversion of PBLG with no disulfide group was performed by the same procedure described above. 2.4. Self-assembly of SSPLGA on GNP A 5 ml aqueous solution of SSPLGA (1 mg) at pH 8.74 was mixed with the citrate-stabilized gold nanoparticles (1 ml). The solution was stirred for 2 h and then centrifuged to remove any free SSPLGA. The resulting precipitates were re-dispersed in fresh pH 8.6 solution and centrifuged. The above procedure was repeated four more times, to give the GNP-SSPLGA material. 2.5. Characterization UV–vis spectra were recorded using a V530 spectrophotometer (Jasco Inc., MD). Dynamic light scattering (DLS) measurements were performed on the colloidal solutions at 25 ◦ C on a Zetasizer Nano-S (Malvern Instrument, PA) equipped with red laser (633 nm). Backscattered light from the sample was analyzed at an angle of 173◦ from the incident light and data fitting was done using a multi-modal algorithm supplied by Malvern. The collected correlelograms were fitted to diffusion co-efficients and converted to hydrodynamic diameters. TEM measurements were performed on a FEI Morgagni 268D operated at an accelerating voltage of 120 kV. Samples for TEM analysis were prepared by placing drops of colloidal solution of GNP–SSPLGA on carbon-coated copper grids. The colloidal solution was allowed to stand for 10 min following which the extra solution was removed using a blotting paper. IR reflection absorption spectroscopy (IRRAS) was performed on a Nicolet Magna IR 860 spectrometer purged with dry air and equipped with polarization-modulation (PM) electronics and a liquid nitrogen-cooled MCT detector [36]. Typically, 2048 scans were accumulated at a resolution of 4 cm−1 for the PMIRRAS characterization. 3. Results and discussion A UV–vis spectrum (Fig. 2a) of the colloidal solution of citrate-stabilized GNP shows the surface plasmon absorption band due to the gold nanoparticles at 523 nm, which gives the colloidal solution a characteristic pink color. The absence of absorbance at wavelengths greater than 600 nm indicates a lack of coupling of plasmons of individual nanoparticles and is a signature of their well-dispersed state in solution [37,38]. Fig. 2a also reveals that the absorption peak shifts to 525 nm in the GNP–SSPLGA material, which is consistent with the red-shifts observed when an organic ligand is attached to the GNP surface and alters the local dielectric permittivity [38]. Templeton and coworkers [38] have shown that a shift in the visible absorption of the GNP can be used to estimate the thickness of a dielectric shell (tshell ). Using the equations proposed in their study, one
Fig. 2. (a) UV–vis absorption spectrum and (b) size distribution from DLS of the citrate stabilized GNP and the GNP–SSPLGA material. Sparse markers have been used for clarity in (a) with one marker for every 40 data points.
can write λ2shell − λ2core 2g = (εshell − εsolvent ), 3 λ2p
(1)
where λcore (∼523 nm) is the surface plasmon peak for the core GNP, λshell (∼525 nm) is the surface plasmon peak for the GNP–SSPLGA, and λp (∼132 nm) is the bulk metal plasmon wavelength. Using εsolvent ∼ 1.78 and εshell ∼ 2.25 for the dielectric constant for water and PLGA gives that the shell volume fraction g ∼ = 0.38. Fig. 2b shows the average hydrodynamic diameter based on scattered intensity distribution in DLS is about 37 nm (Dcore ) for the bare GNP and 41 nm for the GNP–SSPLGA, which indicates that the SSPLGA layer contributes a thickness of approximately 2 nm. Since the volume fraction of a shell, g, can be written [38] as (1 − g)1/3 =
Dcore /2 , Dcore /2 + tshell
(2)
980
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
Fig. 4. TEM images of the GNP–SSPLGA for samples from solution at (a) high pH (∼8.7) and (b) low pH (∼3.6).
Fig. 3. (a) UV–vis absorption spectra and (b) size of the GNP–SSPLGA as a function of the pH of the colloidal solution. Sparse markers have been used for clarity in (a) with one marker for every 30 data points.
the optical shift in Fig. 2a gives an estimated tshell ∼ = 3 nm, which is consistent with the DLS result. The thickness of the SSPLGA layer is also consistent with the degree of polymerization (∼104) of SSPLGA [36] and its estimated the Rg of ∼3 nm in a random coil state [39]. The colloidal solution of GNP–SSPLGA in high pH is stable and no sedimentation is observed even after twelve months. Lowering the solution pH by addition of 0.1 M HCl solution leads to visible changes in the color of the solution to blue/purple and the nanoparticles aggregate. Fig. 3a shows in detail the changes in the absorbance of the colloidal solution of the GNP–SSPLGA as a function of pH. It is observed that until a pH of about 5.5, the main change is the decrease in absorbance of the solution. While some of the decrease can be expected due to the decrease in concentration of the nanoparticles from the addition of aliquots of the acidic solution, the possibility of some aggregation and settling cannot be discounted. However,
these effects are neither visible by the naked eye nor observable in the optical spectrum. Further decrease in pH from 5.5 leads to the occurrence of a shoulder at long wavelengths in the optical spectrum and at pH values lower than 4.5 the dominant absorption is at wavelengths greater than 600 nm, which is a clear indication of the aggregation of the nanoparticles. DLS measurements of the average size of the GNP–SSPLGA as a function of pH are shown in Fig. 3b. These measurements clearly indicate that the aggregation occurs quite sharply in the range of pH between 4.5 and 5.5. The TEM images of the sample at high pH (>8) and low pH (<4) in Fig. 4 supports the physical picture deduced from the absorption and DLS measurements. The typical size of the GNP (∼23 nm) in Fig. 4a are smaller than the DLS measurement but a direct comparison between the two techniques is complicated by factors such as sample size, sensitivity to hydrodynamic conditions in DLS measurements, etc. [40,41]. The GNP–SSPLGA are fairly welldispersed as isolated particles in the sample deposited at high pH (Fig. 4a) and clustered into aggregates at the low pH conditions (Fig. 4b). All the changes shown in Figs. 3a and 3b occur rapidly within minutes of the change in pH of the solution. At low pH conditions and over a much longer time of several hours, the aggregated nanoparticles settle and the blue/purple colored colloidal solution changes to a clear solution and small aggregates are detectable with the naked eye. The reversibility of the aggregation process is demonstrated in Fig. 5. Changing the pH of the solution consecutively between high (∼8.75) and low (∼3.5) using addition of microliter quantities of 0.1 N HCl and 0.1 N NaOH easily cycles the
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
981
Fig. 5. Plot of wavelength corresponding to maximum absorption in the UV–vis spectrum as a function of successive changes in pH from high to low conditions for the colloidal solution of the GNP–SSPLGA. See text for discussion.
GNP–SSPLGA between dispersed and aggregated states. This phenomenon is evident from the wavelength of maximum absorption shown in Fig. 5. Correspondingly the color of colloidal dispersion also changes rapidly between red and blue. One interesting result observed in this experiment is that for complete re-dispersion of the aggregate, one has to wait considerably longer than minutes. This is evident in Fig. 5 where the maximum plasmon absorption wavelength does not recover to the beginning value (∼525 nm) after addition of NaOH in every cycle and slightly red-shifts to higher values. Control experiments show that the chemically attached SSPLGA is necessary for the process of aggregation and re-dispersion between particles. Both the bare citrate-stabilized gold nanoparticles and the gold nanoparticles treated with PLGA that has no disulfide group were subjected to same pH variation experiments. The nanoparticles aggregated irreversibly at low pH and could not be re-dispersed at high pH conditions. Therefore, these results indicate that the reversible aggregation of the GNP–SSPLGA system is dependent on the presence of a self-assembled layer of the polypeptide. To gain insight into the driving forces behind the aggregation and dispersion process of the modified GNP, it is necessary to consider the effect of changing pH on the polyacid (PLGA) [42,43]. It is well known that as the pH of the solution is changed from low to high pH, the carboxylic acid groups in the polyacid undergo de-protonation and the polymer becomes progressively negatively charged [42–44]. The isoelectric point of PLGA is near 3.2 [45]. The variation in pH also leads to a change in secondary structure of the polymer from an α-helical structure to a random coil and the breakage of the protein hydrogen bonds between –NH and –CO groups of the polymer [42]. To confirm the pH-dependent effects for SSPLGA self-assembled on gold surfaces we chose IR absorption spectroscopy, which is an accepted technique to probe structural changes peptides both in bulk samples and on surfaces [30,34,46–50]. GNP–SSPLGA were immobilized on
Fig. 6. External reflection IR spectra for a film of SSPLGA formed on gold coated substrates after immersion in solutions with high and low pH conditions as indicated. In each panel, the thick grey line is the experimentally measured spectrum and the fitted spectrum is the solid dark line. The deconvoluted Gaussian peaks for the fit are shown separately for clarity. The arrows indicate location of the peaks corresponding to carboxylic acid, amide I, carboxylate anion, and amide II stretches for PLGA (see text for discussion). The peak at ∼1455 cm−1 is due to methylene groups in the backbone.
gold films prepared by thermal evaporation on glass substrates and modified by a 0.01% ethanolic solution of poly(vinyl pyridine) (PVP) as per the procedure by Chumanov and coworkers [51]. The PM-IRRAS spectrum (supplementary material), although weak and exhibiting overlapping absorption from the PVP material, revealed the contribution of the SSPLGA. However, the sensitivity of this approach was not useful to probe pH-induced changed in the self-assembled monolayer of SSPLGA.
982
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
Fig. 7. Comparison of the DLS data on hydrodynamic diameter (DH ) the GNP–SSPLGA system with literature data [42] on the molar heat content of PLGA (−H 0 ), the specific optical rotation at 233 nm (−[α]233 ), and potentiometric titration for the degree of ionization (α) of PLGA as a function of pH.
In a fairly recent study, Mandal and Kraatz [30] have demonstrated that degree of surface curvature can have profound effects on the secondary structure of peptides self-assembled on nanoparticles with sizes below 20 nm. In our case, the GNP are large enough that surface curvature of the nanoparticles should have negligible influence on the secondary structure of the polymer and the SSPLGA monolayers should be similar to that on planar substrates. Therefore, evaporated gold substrates were directly immersed in a dilute aqueous solution (0.1 mg/ml) of SSPLGA for 3 days and the resulting SSPLGA conjugated gold film was then characterized using PM-IRRAS after exposure to high pH (∼8.4) and low pH (3.6–4.0) solutions (Figs. 6a–6c). The IR spectrum of SSPLGA, when exposed to low pH (Fig. 6a), is consistent with an α-helical structure based on the location of peaks at ∼1663 cm−1 and ∼1549 cm−1 , which can be assigned to amide I and amide II bands in the spectrum for surface tethered peptides [50]. These peaks are shifted from the positions observed in bulk KBr pellets due to optical effects from the IRRAS technique [30,52]. The shoulder at ∼1734 cm−1 is consistent with a protonated carboxylic acid side chain. The spectrum of SSPLGA after exposure to high pH (Fig. 6b) shows that the shoulder at 1734 cm−1 weakens due to ionization of the –COOH groups and the amide II band broadens due to an overlapping band near ∼1590 cm−1 from the carboxylate anion. The deconvoluted spectra show that amide I band shifts to ∼1668 cm−1 while the amide II shifts to 1544 cm−1 , which are in good agreement with the values indicated in literature for a helix to random coil transition [50]. These changes in the IR spectra are also reversible with pH as evident in Fig. 6c, which shows the reappearance of a strong shoulder at ∼1734 cm−1 , the amide I band at ∼1664 cm−1 , and the amide II band at ∼1550 cm−1 . In case of the GNP–SSPLGA system, we believe that the aggregation phenomenon is a macroscopic manifestation of the structural change of the polypeptide from coil to helix and the accompanying occurrence of hydrophobic interactions between
the PLGA modified GNP. This idea is supported by a comparison of the DLS data on the nanoparticle system with data on calorimetric and polarimetric characterization for PLGA (Fig. 7). Rialdi and Hermans [42] have reported an abrupt increase in the molar heat content of PLGA (−H 0 ) and the specific optical rotation at 233 nm (−[α]233 ) in the pH range between 4.5 and 5.5 due to formation of the peptide hydrogen bonding that accompanies helix conformation of PLGA. In contrast, their potentiometric titration results for the degree of ionization (α) of the PLGA only show a gradual change in the degree of ionization. Fig. 7 shows that the change in size of the GNP–SSPLGA system follows the enthalpy and optical rotation values, which supports the contention that the aggregation is a result of the structural change rather than purely a change in electrostatic interactions. The strategy demonstrated here wherein reversible aggregation of gold nanoparticles occurs with transformation of a polypeptide between helical and random coil conformations is of interest in the area of nanobiotechnology [16,17]. The use of a simple synthetic polypeptide such as PLGA demonstrates that materials with a designed sequence of residues may not always be necessary. Extension to other synthetic biopolymers that respond in a different range of pH values or to other stimuli or exploit other conformations (e.g., tertiary β-sheet) can lead to engineering novel constructs of isotropic nanoparticles, anisotropic nanorods, magnetic, and semi-conducting particles for functional devices in the photonics, sensors, and electronics areas. Some examples of applications include micro-analytical sensors wherein environmental changes can be sensed, transduced, and amplified simultaneously by exploiting the response of the biopolymer in tandem with the collective properties of the nanoparticles or in novel waveguide and optical devices based on plasmonics [53–55] wherein electromagnetic energy can be channeled through inter-connects of metallic nanoparticles with the polypeptides acting as effective conduits only in specific conformations.
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
4. Summary and conclusions In summary, a simple approach based on a self-assembled polypeptide on gold nanoparticles is reported for driving aggregation of the nanoparticles in solution. Reversible aggregation of gold nanoparticles is induced by pH dependent changes in a self-assembled monolayer of poly(L-glutamic acid). A disulfide anchoring group on the PLGA leads to spontaneous chemisorption onto the GNP surface and the formation of a self-assembled monolayer. Characterization of the modified GNP and its aggregation using dynamic light scattering, UV–vis and IR spectroscopy, and TEM shows that as the SSPLGA undergoes a change from a random coil to a helical conformation with decrease in pH, the modified GNP aggregates. The change in aggregation behavior with pH occurs rapidly on the time scale of minutes and is reversible. Acknowledgments This research was carried out with the support of ACS-PRF (Grant 38886-AC7) and the Department of Chemical Engineering at the University of South Florida. We thank Dr. Alveda Williams for the synthesis of the SSPBLG polymer used in this study. Supplementary material The online version of this article contains additional supplementary material. Please visit DOI: 10.1016/j.jcis.2007.08.021. References [1] G. Hodes, Adv. Mater. (Weinheim, Germany) 19 (5) (2007) 639–655. [2] A.H. Lu, E.L. Salabas, F. Schueth, Angew. Chem. Int. Ed. 46 (8) (2007) 1222–1244. [3] V.M. Rotello, Mater. Today (Oxford, UK) 4 (6) (2001) 24–29. [4] P. Mulvaney, Nanoscale Mater. Chem. (2001) 121–167. [5] M. Brust, C.J. Kiely, Colloids Surf. A Physicochem. Eng. Aspects 202 (2– 3) (2002) 175–186. [6] S. Mann, W. Shenton, M. Li, S. Connolly, D. Fitzmaurice, Adv. Mater. (Weinheim, Germany) 12 (2) (2000) 147–150. [7] A.N. Shipway, E. Katz, I. Willner, Chem. Phys. Chem. 1 (1) (2000) 18–52. [8] T. Sakai, P. Alexandridis, Chem. Mater. 18 (10) (2006) 2577–2583. [9] J. Dutta, H. Hofmann, Encyclopedia Nanosci. Nanotechnol. 9 (2004) 617– 640. [10] Z.M. Fresco, J.M.J. Frechet, J. Am. Chem. Soc. 127 (23) (2005) 8302– 8303. [11] H. Huang, J.N. Anker, K. Wang, R. Kopelman, J. Phys. Chem. B 110 (40) (2006) 19929–19934. [12] B.A. Parviz, D. Ryan, G.M. Whitesides, IEEE Trans. Adv. Pack. 26 (3) (2003) 233–241. [13] L. Samuelson, Mater. Today (Oxford, UK) 6 (10) (2003) 22–31. [14] M. Schmittel, V. Kalsani, Top. Curr. Chem. 245 (2005) 1–53. [15] W. Shenton, S.A. Davis, S. Mann, Adv. Mater. (Weinheim, Germany) 11 (6) (1999) 449–452. [16] J.J. Storhoff, C.A. Mirkin, Chem. Rev. (Washington, DC) 99 (7) (1999) 1849–1862. [17] C.M. Niemeyer, Angew. Chem. Int. Ed. 40 (22) (2001) 4128–4158. [18] R.P. Andres, J.D. Bielefeld, J.I. Henderson, D.B. Janes, V.R. Kolagunta, C.P. Kubiak, W.J. Mahoney, R.G. Osifchin, Science 273 (5282) (1996) 1690–1693.
983
[19] C.A. Mirkin, R.L. Letsinger, R.C. Mucic, J.J. Storhoff, Nature 382 (6592) (1996) 607–609. [20] G. Li, T. Wang, S. Bhosale, Y. Zhang, J.-H. Fuhrhop, Colloid Polym. Sci. 281 (11) (2003) 1099–1103. [21] C. Guarise, L. Pasquato, P. Scrimin, Langmuir 21 (12) (2005) 5537–5541. [22] A.N. Shipway, M. Lahav, R. Gabai, I. Willner, Langmuir 16 (23) (2000) 8789–8795. [23] N. Nath, A. Chilkoti, J. Am. Chem. Soc. 123 (34) (2001) 8197–8202. [24] K. Aslan, C.C. Luhrs, V.H. Perez-Luna, J. Phys. Chem. B 108 (40) (2004) 15631–15639. [25] Y. Yang, S. Matsubara, M. Nogami, J. Shi, W. Huang, Nanotechnology 17 (11) (2006) 2821–2827. [26] P. Hazarika, B. Ceyhan, C.M. Niemeyer, Angew. Chem. Int. Ed. 43 (47) (2004) 6469–6471. [27] M.G. Ryadnov, B. Ceyhan, C.M. Niemeyer, D.N. Woolfson, J. Am. Chem. Soc. 125 (31) (2003) 9388–9394. [28] R. Levy, N.T.K. Thanh, R.C. Doty, I. Hussain, R.J. Nichols, D.J. Schiffrin, M. Brust, D.G. Fernig, J. Am. Chem. Soc. 126 (32) (2004) 10076–10084. [29] D. Aili, K. Enander, J. Rydberg, I. Lundstroem, L. Baltzer, B. Liedberg, J. Am. Chem. Soc. 128 (7) (2006) 2194–2195. [30] H.S. Mandal, H.-B. Kraatz, J. Am. Chem. Soc. 129 (20) (2007) 6356– 6357. [31] M.-A. Morikawa, N. Kimizuka, Chem. Commun. (Cambridge, UK) (38) (2005) 4866–4868. [32] H. Nishikawa, T. Morita, J. Sugiyama, S. Kimura, J. Colloid Interface Sci. 280 (2) (2004) 506–510. [33] A.G. Tkachenko, H. Xie, D. Coleman, W. Glomm, J. Ryan, M.F. Anderson, S. Franzen, D.L. Feldheim, J. Am. Chem. Soc. 125 (16) (2003) 4700–4701. [34] N. Higashi, J. Kawahara, M. Niwa, J. Colloid Interface Sci. 288 (1) (2005) 83–87. [35] G. Frens, Nature (London), Phys. Sci. 241 (105) (1973) 20–22. [36] A.J. Williams, V.K. Gupta, J. Phys. Chem. B 105 (22) (2001) 5223– 5230. [37] U. Kreibig, Optical Properties of Metal Clusters, Springer, Berlin, 1995. [38] A.C. Templeton, J.J. Pietron, R.W. Murray, P. Mulvaney, J. Phys. Chem. B 104 (3) (2000) 564–570. [39] J.E. Kohn, I.S. Millett, J. Jacob, B. Zagrovic, T.M. Dillon, N. Cingel, R.S. Dothager, S. Seifert, P. Thiyagarajan, T.R. Sosnick, M.Z. Hasan, V.S. Pande, I. Ruczinski, S. Doniach, K.W. Plaxco, Proc. Natl. Acad. Sci. USA 101 (34) (2004) 12491–12496. [40] T. Ito, L. Sun, M.A. Bevan, R.M. Crooks, Langmuir 20 (16) (2004) 6940– 6945. [41] C.Y. Wang, W.E. Fu, H.L. Lin, G.S. Peng, Meas. Sci. Technol. 18 (2) (2007) 487–495. [42] G. Rialdi, J. Hermans Jr., J. Am. Chem. Soc. 88 (24) (1966) 5719–5720. [43] A. Warashina, A. Ikegami, Biopolymers 11 (3) (1972) 529–547. [44] D.T. Haynie, S. Balkundi, N. Palath, K. Chakravarthula, K. Dave, Langmuir 20 (11) (2004) 4540–4547. [45] S. Galier, H. Roux-de Balmann, J. Membr. Sci. 241 (1) (2004) 79–87. [46] T. Kawai, T. Komoto, N. Honda, Makromol. Chem. 182 (8) (1981) 2127– 2137. [47] T. Miyazawa, E.R. Blout, J. Am. Chem. Soc. 83 (1961) 712–719. [48] T. Abe, S. Hayashi, N. Higashi, M. Niwa, K. Kurihara, Colloids Surf. A Physicochem. Eng. Aspects 169 (1–3) (2000) 351–356. [49] K. Kurihara, T. Abe, N. Higashi, M. Niwa, Colloids Surf. A Physicochem. Eng. Aspects 103 (3) (1995) 265–272. [50] N.H. Lee, C.W. Frank, Langmuir 19 (4) (2003) 1295–1303. [51] S. Malynych, I. Luzinov, G. Chumanov, J. Phys. Chem. B 106 (2002) 1280–1285. [52] D.L. Allara, A. Baca, C.A. Pryde, Macromolecules 11 (6) (1978) 1215– 1220. [53] H.A. Atwater, S. Maier, A. Polman, J.A. Dionne, L. Sweatlock, MRS Bull. 30 (5) (2005) 385–389. [54] S.A. Maier, M.L. Brongersma, P.G. Kik, S. Meltzer, A.A.G. Requicha, H.A. Atwater, Adv. Mater. 13 (19) (2001) 1501–1505. [55] K.G. Thomas, S. Barazzouk, B.I. Ipe, S.T.S. Joseph, P.V. Kamat, J. Phys. Chem. B 108 (35) (2004) 13066–13068.
Reversible aggregation of gold nanoparticles induced by pH dependent conformational transitions of a self-assembled polypeptide Jeung-Yeop Shim, Vinay K. Gupta ∗ Department of Chemical Engineering, University of South Florida, ENB 118, 4202 E Fowler Avenue, Tampa, FL 33620, USA Received 2 July 2007; accepted 8 August 2007 Available online 14 August 2007
Abstract Controlling the stable structures of metallic nanoparticles on mesoscopic and macroscopic length scales is of great interest in nanotechnology. Here, this task is accomplished using a synthetic biopolymer that is responsive to external stimuli and undergoes changes in secondary structure. Reversible aggregation of gold nanoparticles (GNP) is induced by pH dependent changes in a self-assembled monolayer of disulfide modified poly(L-glutamic acid) (SSPLGA) with Mw ∼ 27000. The disulfide anchoring group drives chemisorption onto the gold nanoparticles and leads to the formation of a self-assembled monolayer. Characterization of the modified GNP and its aggregation behavior is performed using dynamic light scattering (DLS), UV–vis and IR spectroscopy, and transmission electron microscopy (TEM). Experimental results show that decrease in pH near 5.5 leads to aggregation of the modified GNP. The change in aggregation behavior with pH occurs within minutes, is reversible, and happens within a narrow range of pH from about 4.5 to 5.5. Comparison with literature data on molar enthalpy of hydrogen bonding, specific optical rotation, and ionization for the helix–coil transition of PLGA indicates that the aggregation of the SSPLGA modified GNP corresponds to the transition in the secondary structure of the polyacid. © 2007 Elsevier Inc. All rights reserved. Keywords: Gold nanoparticles; Reversible aggregation; Self-assembly; Self-assembled polypeptide
1. Introduction The increasing ability to prepare a wide range of nanoparticles of various sizes, shapes, and materials has spurred the need for strategies to assemble these nanomaterials into constructs for functional devices such as sensors, photonic materials, and biomolecular electronics [1–8]. Traditional approaches that rely solely on top–down lithographic methods are of limited use when the nanostructures need to be created either in three-dimensions or on sub-100 nm scales. These approaches also rapidly become energy and materials intensive for twodimensional structures at nanometer length-scales. In this context, directed self-assembly for nanometer scale organization has been recognized as most promising for either enabling hybrid methods in combination with the established lithographic tools or providing affordable and simpler alternatives for nanoscale devices, sensors, and nano/micro fluidic systems * Corresponding author. Fax: +1 813 974 3651.
E-mail address: [email protected] (V.K. Gupta). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.08.021
[6,9–17]. However, the principles for creating, manipulating, and using directed self-assembling systems remain a scientific and technological challenge. In particular, strategies need to be discovered for encoding information that drives assembly of complex motifs, for incorporating molecular interactions that provide dimensional and positional control during assembly, and for including triggers to manipulate the assembled structures. Towards the objective of establishing principles for directed self-assembly of nanoparticles, several approaches have been pursued in recent years for controlling the aggregation of nanoparticles in solution. In these investigations, gold colloids have been a popular choice due to the ease of chemical modification of the gold surface and the fact that surface modification provides sites for covalent, electrostatic, or other inter-molecular interactions that subsequently drive aggregation or flocculation of the nanoparticles. Examples from literature include attachment of thiolated DNA oligomers that cause particle aggregation by undergoing hybridization with complementary molecules, use of ligands that contain termi-
978
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
nal carboxylic or amino groups for electrostatic and hydrogen bonding interactions or for further conjugation for biospecific interactions, and modification of the nanoparticles with an elastin biopolymer that induces interparticle hydrophobic interactions [18–26]. Among the diverse approaches that have been reported in the past few years, a majority of these studies have demonstrated irreversible aggregation and only a few examples of reversible aggregation exist. For the latter, primarily two methods have been used. One method relies on manipulating the colloidal aggregation by changing temperature to either cause thermal denaturation or induce a hydrophilic– hydrophobic phase transition of a grafted polymer [19,23,26, 27]. The second method relies on addition of a reagent that either causes a chemical cleavage of the linkage between particles or changes the pH and/or ionic strength of the solution [20, 21,28,29]. A related approach has also been reported in a system wherein streptavidin-mediated aggregation of biotinylated GNP can be reversed over a period of 80 days by addition of soluble biotin to block the streptavidin [24]. In studies where pH or ionic strength has been varied, the underlying goal has been disruption of hydrogen bonds or electrostatic interactions between modified gold nanoparticles. For example, Fuhrhop and coworkers [20] altered pH to influence hydrogen bonding between simple ω-functionalized alkanethiols such as mercaptoundecanoic acid on the nanoparticles while Fernig et al. [28] used both pH and ionic strength to change electrostatic interactions of a carboxylic acid moiety on the surface of gold nanoparticles. Current years have witnessed a growing focus on combination of synthetic proteins with gold nanoparticles for areas related to biosensors, biomimetics, and biomolecule-directed assembly [17,27,29–33]. Both designer peptides with carefully crafted sequences made by solid-phase peptide synthesis and one-pot materials made by solution-phase polymerization have been successfully used to modify gold nanoparticles surfaces. For example, Feldheim and coworkers [33] have used peptides sequences that mimic virus peptides to modify gold nanoparticles for targeting cells while Higashi et al. [34] have shown that a homopolymer such as poly(γ -benzyl-L-glutamate) can be used to functionalize gold nanoparticles. Towards the particular goal of using surface-tethered peptides to control nanoparticle assembly, recent studies have explored carefully designed peptide sequences for switching particle aggregation by exploiting protein folding and interactions between folded proteins [27–29]. In this context, the present study reports reversible aggregation of the gold nanoparticles that are modified by self-assembly of a simple homopolypeptide, namely, disulfide modified poly(L-glutamic acid) (SSPLGA) that is prepared by polymerization of the N -carboxy anyhydride (NCA). The SSPLGA spontaneously modifies the gold nanoparticle surface via chemisorption of the disulfide headgroup. Cycling the pH changes the aggregation state of the gold colloids in solution from dispersed to aggregated (Fig. 1) and this aggregation corresponds to a change in α-helical conformation of SSPLGA at pH <4 to a random coil conformation at pH >6. To our knowledge, this is the first demonstration where PLGA is employed
Fig. 1. Schematic illustration of the conformation transition from a helix to a random coil for the SSPLGA and the resulting aggregation of gold nanoparticles. The PLGA is modified with a disulfide group that anchors it to the nanoparticle surface.
to reversibly manipulate the state of aggregation of the gold nanoparticles. 2. Experimental 2.1. Materials Hydrogen tetrachloroaurate (HAuCl4 ·3H2 O) and trisodium citrate were purchased from Sigma-Aldrich (WI). A 33 wt% solution of hydrobromic acid in acetic acid (HBr/AcOH) and diethyl ether were purchased from Acros Organics (NJ). Water used in all synthesis was purified using an EasyPure™ UV system (Barnstead, IA). A 0.2 µm filter incorporated into this system removed particulate matter. 2.2. Synthesis of gold nanoparticles The synthesis methodology of gold nanoparticles is based on Frens’s method [35]. A solution HAuCl4 ·3H2 O (40 mg, 0.102 mmol) in deionized water (100 ml) was brought to a boil and citrate trisodium salt (100 mg) was added in one portion. The solution was refluxed for 30 min and then allowed to cool to room temperature. The resulting reddish solution of the GNP was stored in refrigerator for future use. 2.3. Preparation of SSPLGA A disulfide modified γ -benzyl-L-glutamate (SSPBLG) was converted to SSPLGA by debenzylation using HBr/AcOH. The preparation and characterization of SSPBLG has been reported elsewhere [36]. For the debenzylation, 100 mg of SSPBLG (Mn ∼ 22700, Mw ∼ 27000) was dispersed in warm AcOH (5 ml) and 3 ml of the HBr/AcOH solution was added dropwise. The solution was stirred at 50 ◦ C for 12 h. The resulting mixture was diluted with ether (30 ml) and sedimentation of crude
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
979
SSPLGA was allowed to occur overnight at room temperature. The solution was centrifuged for collection of SSPLGA, which was then washed with cold ether (30 ml × 5 times) and dried to give a SSPLGA (48 mg). The debenzylation was verified using IR spectroscopy of the bulk material in a KBr pellet. For control experiments, conversion of PBLG with no disulfide group was performed by the same procedure described above. 2.4. Self-assembly of SSPLGA on GNP A 5 ml aqueous solution of SSPLGA (1 mg) at pH 8.74 was mixed with the citrate-stabilized gold nanoparticles (1 ml). The solution was stirred for 2 h and then centrifuged to remove any free SSPLGA. The resulting precipitates were re-dispersed in fresh pH 8.6 solution and centrifuged. The above procedure was repeated four more times, to give the GNP-SSPLGA material. 2.5. Characterization UV–vis spectra were recorded using a V530 spectrophotometer (Jasco Inc., MD). Dynamic light scattering (DLS) measurements were performed on the colloidal solutions at 25 ◦ C on a Zetasizer Nano-S (Malvern Instrument, PA) equipped with red laser (633 nm). Backscattered light from the sample was analyzed at an angle of 173◦ from the incident light and data fitting was done using a multi-modal algorithm supplied by Malvern. The collected correlelograms were fitted to diffusion co-efficients and converted to hydrodynamic diameters. TEM measurements were performed on a FEI Morgagni 268D operated at an accelerating voltage of 120 kV. Samples for TEM analysis were prepared by placing drops of colloidal solution of GNP–SSPLGA on carbon-coated copper grids. The colloidal solution was allowed to stand for 10 min following which the extra solution was removed using a blotting paper. IR reflection absorption spectroscopy (IRRAS) was performed on a Nicolet Magna IR 860 spectrometer purged with dry air and equipped with polarization-modulation (PM) electronics and a liquid nitrogen-cooled MCT detector [36]. Typically, 2048 scans were accumulated at a resolution of 4 cm−1 for the PMIRRAS characterization. 3. Results and discussion A UV–vis spectrum (Fig. 2a) of the colloidal solution of citrate-stabilized GNP shows the surface plasmon absorption band due to the gold nanoparticles at 523 nm, which gives the colloidal solution a characteristic pink color. The absence of absorbance at wavelengths greater than 600 nm indicates a lack of coupling of plasmons of individual nanoparticles and is a signature of their well-dispersed state in solution [37,38]. Fig. 2a also reveals that the absorption peak shifts to 525 nm in the GNP–SSPLGA material, which is consistent with the red-shifts observed when an organic ligand is attached to the GNP surface and alters the local dielectric permittivity [38]. Templeton and coworkers [38] have shown that a shift in the visible absorption of the GNP can be used to estimate the thickness of a dielectric shell (tshell ). Using the equations proposed in their study, one
Fig. 2. (a) UV–vis absorption spectrum and (b) size distribution from DLS of the citrate stabilized GNP and the GNP–SSPLGA material. Sparse markers have been used for clarity in (a) with one marker for every 40 data points.
can write λ2shell − λ2core 2g = (εshell − εsolvent ), 3 λ2p
(1)
where λcore (∼523 nm) is the surface plasmon peak for the core GNP, λshell (∼525 nm) is the surface plasmon peak for the GNP–SSPLGA, and λp (∼132 nm) is the bulk metal plasmon wavelength. Using εsolvent ∼ 1.78 and εshell ∼ 2.25 for the dielectric constant for water and PLGA gives that the shell volume fraction g ∼ = 0.38. Fig. 2b shows the average hydrodynamic diameter based on scattered intensity distribution in DLS is about 37 nm (Dcore ) for the bare GNP and 41 nm for the GNP–SSPLGA, which indicates that the SSPLGA layer contributes a thickness of approximately 2 nm. Since the volume fraction of a shell, g, can be written [38] as (1 − g)1/3 =
Dcore /2 , Dcore /2 + tshell
(2)
980
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
Fig. 4. TEM images of the GNP–SSPLGA for samples from solution at (a) high pH (∼8.7) and (b) low pH (∼3.6).
Fig. 3. (a) UV–vis absorption spectra and (b) size of the GNP–SSPLGA as a function of the pH of the colloidal solution. Sparse markers have been used for clarity in (a) with one marker for every 30 data points.
the optical shift in Fig. 2a gives an estimated tshell ∼ = 3 nm, which is consistent with the DLS result. The thickness of the SSPLGA layer is also consistent with the degree of polymerization (∼104) of SSPLGA [36] and its estimated the Rg of ∼3 nm in a random coil state [39]. The colloidal solution of GNP–SSPLGA in high pH is stable and no sedimentation is observed even after twelve months. Lowering the solution pH by addition of 0.1 M HCl solution leads to visible changes in the color of the solution to blue/purple and the nanoparticles aggregate. Fig. 3a shows in detail the changes in the absorbance of the colloidal solution of the GNP–SSPLGA as a function of pH. It is observed that until a pH of about 5.5, the main change is the decrease in absorbance of the solution. While some of the decrease can be expected due to the decrease in concentration of the nanoparticles from the addition of aliquots of the acidic solution, the possibility of some aggregation and settling cannot be discounted. However,
these effects are neither visible by the naked eye nor observable in the optical spectrum. Further decrease in pH from 5.5 leads to the occurrence of a shoulder at long wavelengths in the optical spectrum and at pH values lower than 4.5 the dominant absorption is at wavelengths greater than 600 nm, which is a clear indication of the aggregation of the nanoparticles. DLS measurements of the average size of the GNP–SSPLGA as a function of pH are shown in Fig. 3b. These measurements clearly indicate that the aggregation occurs quite sharply in the range of pH between 4.5 and 5.5. The TEM images of the sample at high pH (>8) and low pH (<4) in Fig. 4 supports the physical picture deduced from the absorption and DLS measurements. The typical size of the GNP (∼23 nm) in Fig. 4a are smaller than the DLS measurement but a direct comparison between the two techniques is complicated by factors such as sample size, sensitivity to hydrodynamic conditions in DLS measurements, etc. [40,41]. The GNP–SSPLGA are fairly welldispersed as isolated particles in the sample deposited at high pH (Fig. 4a) and clustered into aggregates at the low pH conditions (Fig. 4b). All the changes shown in Figs. 3a and 3b occur rapidly within minutes of the change in pH of the solution. At low pH conditions and over a much longer time of several hours, the aggregated nanoparticles settle and the blue/purple colored colloidal solution changes to a clear solution and small aggregates are detectable with the naked eye. The reversibility of the aggregation process is demonstrated in Fig. 5. Changing the pH of the solution consecutively between high (∼8.75) and low (∼3.5) using addition of microliter quantities of 0.1 N HCl and 0.1 N NaOH easily cycles the
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
981
Fig. 5. Plot of wavelength corresponding to maximum absorption in the UV–vis spectrum as a function of successive changes in pH from high to low conditions for the colloidal solution of the GNP–SSPLGA. See text for discussion.
GNP–SSPLGA between dispersed and aggregated states. This phenomenon is evident from the wavelength of maximum absorption shown in Fig. 5. Correspondingly the color of colloidal dispersion also changes rapidly between red and blue. One interesting result observed in this experiment is that for complete re-dispersion of the aggregate, one has to wait considerably longer than minutes. This is evident in Fig. 5 where the maximum plasmon absorption wavelength does not recover to the beginning value (∼525 nm) after addition of NaOH in every cycle and slightly red-shifts to higher values. Control experiments show that the chemically attached SSPLGA is necessary for the process of aggregation and re-dispersion between particles. Both the bare citrate-stabilized gold nanoparticles and the gold nanoparticles treated with PLGA that has no disulfide group were subjected to same pH variation experiments. The nanoparticles aggregated irreversibly at low pH and could not be re-dispersed at high pH conditions. Therefore, these results indicate that the reversible aggregation of the GNP–SSPLGA system is dependent on the presence of a self-assembled layer of the polypeptide. To gain insight into the driving forces behind the aggregation and dispersion process of the modified GNP, it is necessary to consider the effect of changing pH on the polyacid (PLGA) [42,43]. It is well known that as the pH of the solution is changed from low to high pH, the carboxylic acid groups in the polyacid undergo de-protonation and the polymer becomes progressively negatively charged [42–44]. The isoelectric point of PLGA is near 3.2 [45]. The variation in pH also leads to a change in secondary structure of the polymer from an α-helical structure to a random coil and the breakage of the protein hydrogen bonds between –NH and –CO groups of the polymer [42]. To confirm the pH-dependent effects for SSPLGA self-assembled on gold surfaces we chose IR absorption spectroscopy, which is an accepted technique to probe structural changes peptides both in bulk samples and on surfaces [30,34,46–50]. GNP–SSPLGA were immobilized on
Fig. 6. External reflection IR spectra for a film of SSPLGA formed on gold coated substrates after immersion in solutions with high and low pH conditions as indicated. In each panel, the thick grey line is the experimentally measured spectrum and the fitted spectrum is the solid dark line. The deconvoluted Gaussian peaks for the fit are shown separately for clarity. The arrows indicate location of the peaks corresponding to carboxylic acid, amide I, carboxylate anion, and amide II stretches for PLGA (see text for discussion). The peak at ∼1455 cm−1 is due to methylene groups in the backbone.
gold films prepared by thermal evaporation on glass substrates and modified by a 0.01% ethanolic solution of poly(vinyl pyridine) (PVP) as per the procedure by Chumanov and coworkers [51]. The PM-IRRAS spectrum (supplementary material), although weak and exhibiting overlapping absorption from the PVP material, revealed the contribution of the SSPLGA. However, the sensitivity of this approach was not useful to probe pH-induced changed in the self-assembled monolayer of SSPLGA.
982
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
Fig. 7. Comparison of the DLS data on hydrodynamic diameter (DH ) the GNP–SSPLGA system with literature data [42] on the molar heat content of PLGA (−H 0 ), the specific optical rotation at 233 nm (−[α]233 ), and potentiometric titration for the degree of ionization (α) of PLGA as a function of pH.
In a fairly recent study, Mandal and Kraatz [30] have demonstrated that degree of surface curvature can have profound effects on the secondary structure of peptides self-assembled on nanoparticles with sizes below 20 nm. In our case, the GNP are large enough that surface curvature of the nanoparticles should have negligible influence on the secondary structure of the polymer and the SSPLGA monolayers should be similar to that on planar substrates. Therefore, evaporated gold substrates were directly immersed in a dilute aqueous solution (0.1 mg/ml) of SSPLGA for 3 days and the resulting SSPLGA conjugated gold film was then characterized using PM-IRRAS after exposure to high pH (∼8.4) and low pH (3.6–4.0) solutions (Figs. 6a–6c). The IR spectrum of SSPLGA, when exposed to low pH (Fig. 6a), is consistent with an α-helical structure based on the location of peaks at ∼1663 cm−1 and ∼1549 cm−1 , which can be assigned to amide I and amide II bands in the spectrum for surface tethered peptides [50]. These peaks are shifted from the positions observed in bulk KBr pellets due to optical effects from the IRRAS technique [30,52]. The shoulder at ∼1734 cm−1 is consistent with a protonated carboxylic acid side chain. The spectrum of SSPLGA after exposure to high pH (Fig. 6b) shows that the shoulder at 1734 cm−1 weakens due to ionization of the –COOH groups and the amide II band broadens due to an overlapping band near ∼1590 cm−1 from the carboxylate anion. The deconvoluted spectra show that amide I band shifts to ∼1668 cm−1 while the amide II shifts to 1544 cm−1 , which are in good agreement with the values indicated in literature for a helix to random coil transition [50]. These changes in the IR spectra are also reversible with pH as evident in Fig. 6c, which shows the reappearance of a strong shoulder at ∼1734 cm−1 , the amide I band at ∼1664 cm−1 , and the amide II band at ∼1550 cm−1 . In case of the GNP–SSPLGA system, we believe that the aggregation phenomenon is a macroscopic manifestation of the structural change of the polypeptide from coil to helix and the accompanying occurrence of hydrophobic interactions between
the PLGA modified GNP. This idea is supported by a comparison of the DLS data on the nanoparticle system with data on calorimetric and polarimetric characterization for PLGA (Fig. 7). Rialdi and Hermans [42] have reported an abrupt increase in the molar heat content of PLGA (−H 0 ) and the specific optical rotation at 233 nm (−[α]233 ) in the pH range between 4.5 and 5.5 due to formation of the peptide hydrogen bonding that accompanies helix conformation of PLGA. In contrast, their potentiometric titration results for the degree of ionization (α) of the PLGA only show a gradual change in the degree of ionization. Fig. 7 shows that the change in size of the GNP–SSPLGA system follows the enthalpy and optical rotation values, which supports the contention that the aggregation is a result of the structural change rather than purely a change in electrostatic interactions. The strategy demonstrated here wherein reversible aggregation of gold nanoparticles occurs with transformation of a polypeptide between helical and random coil conformations is of interest in the area of nanobiotechnology [16,17]. The use of a simple synthetic polypeptide such as PLGA demonstrates that materials with a designed sequence of residues may not always be necessary. Extension to other synthetic biopolymers that respond in a different range of pH values or to other stimuli or exploit other conformations (e.g., tertiary β-sheet) can lead to engineering novel constructs of isotropic nanoparticles, anisotropic nanorods, magnetic, and semi-conducting particles for functional devices in the photonics, sensors, and electronics areas. Some examples of applications include micro-analytical sensors wherein environmental changes can be sensed, transduced, and amplified simultaneously by exploiting the response of the biopolymer in tandem with the collective properties of the nanoparticles or in novel waveguide and optical devices based on plasmonics [53–55] wherein electromagnetic energy can be channeled through inter-connects of metallic nanoparticles with the polypeptides acting as effective conduits only in specific conformations.
J.-Y. Shim, V.K. Gupta / Journal of Colloid and Interface Science 316 (2007) 977–983
4. Summary and conclusions In summary, a simple approach based on a self-assembled polypeptide on gold nanoparticles is reported for driving aggregation of the nanoparticles in solution. Reversible aggregation of gold nanoparticles is induced by pH dependent changes in a self-assembled monolayer of poly(L-glutamic acid). A disulfide anchoring group on the PLGA leads to spontaneous chemisorption onto the GNP surface and the formation of a self-assembled monolayer. Characterization of the modified GNP and its aggregation using dynamic light scattering, UV–vis and IR spectroscopy, and TEM shows that as the SSPLGA undergoes a change from a random coil to a helical conformation with decrease in pH, the modified GNP aggregates. The change in aggregation behavior with pH occurs rapidly on the time scale of minutes and is reversible. Acknowledgments This research was carried out with the support of ACS-PRF (Grant 38886-AC7) and the Department of Chemical Engineering at the University of South Florida. We thank Dr. Alveda Williams for the synthesis of the SSPBLG polymer used in this study. Supplementary material The online version of this article contains additional supplementary material. Please visit DOI: 10.1016/j.jcis.2007.08.021. References [1] G. Hodes, Adv. Mater. (Weinheim, Germany) 19 (5) (2007) 639–655. [2] A.H. Lu, E.L. Salabas, F. Schueth, Angew. Chem. Int. Ed. 46 (8) (2007) 1222–1244. [3] V.M. Rotello, Mater. Today (Oxford, UK) 4 (6) (2001) 24–29. [4] P. Mulvaney, Nanoscale Mater. Chem. (2001) 121–167. [5] M. Brust, C.J. Kiely, Colloids Surf. A Physicochem. Eng. Aspects 202 (2– 3) (2002) 175–186. [6] S. Mann, W. Shenton, M. Li, S. Connolly, D. Fitzmaurice, Adv. Mater. (Weinheim, Germany) 12 (2) (2000) 147–150. [7] A.N. Shipway, E. Katz, I. Willner, Chem. Phys. Chem. 1 (1) (2000) 18–52. [8] T. Sakai, P. Alexandridis, Chem. Mater. 18 (10) (2006) 2577–2583. [9] J. Dutta, H. Hofmann, Encyclopedia Nanosci. Nanotechnol. 9 (2004) 617– 640. [10] Z.M. Fresco, J.M.J. Frechet, J. Am. Chem. Soc. 127 (23) (2005) 8302– 8303. [11] H. Huang, J.N. Anker, K. Wang, R. Kopelman, J. Phys. Chem. B 110 (40) (2006) 19929–19934. [12] B.A. Parviz, D. Ryan, G.M. Whitesides, IEEE Trans. Adv. Pack. 26 (3) (2003) 233–241. [13] L. Samuelson, Mater. Today (Oxford, UK) 6 (10) (2003) 22–31. [14] M. Schmittel, V. Kalsani, Top. Curr. Chem. 245 (2005) 1–53. [15] W. Shenton, S.A. Davis, S. Mann, Adv. Mater. (Weinheim, Germany) 11 (6) (1999) 449–452. [16] J.J. Storhoff, C.A. Mirkin, Chem. Rev. (Washington, DC) 99 (7) (1999) 1849–1862. [17] C.M. Niemeyer, Angew. Chem. Int. Ed. 40 (22) (2001) 4128–4158. [18] R.P. Andres, J.D. Bielefeld, J.I. Henderson, D.B. Janes, V.R. Kolagunta, C.P. Kubiak, W.J. Mahoney, R.G. Osifchin, Science 273 (5282) (1996) 1690–1693.
983
[19] C.A. Mirkin, R.L. Letsinger, R.C. Mucic, J.J. Storhoff, Nature 382 (6592) (1996) 607–609. [20] G. Li, T. Wang, S. Bhosale, Y. Zhang, J.-H. Fuhrhop, Colloid Polym. Sci. 281 (11) (2003) 1099–1103. [21] C. Guarise, L. Pasquato, P. Scrimin, Langmuir 21 (12) (2005) 5537–5541. [22] A.N. Shipway, M. Lahav, R. Gabai, I. Willner, Langmuir 16 (23) (2000) 8789–8795. [23] N. Nath, A. Chilkoti, J. Am. Chem. Soc. 123 (34) (2001) 8197–8202. [24] K. Aslan, C.C. Luhrs, V.H. Perez-Luna, J. Phys. Chem. B 108 (40) (2004) 15631–15639. [25] Y. Yang, S. Matsubara, M. Nogami, J. Shi, W. Huang, Nanotechnology 17 (11) (2006) 2821–2827. [26] P. Hazarika, B. Ceyhan, C.M. Niemeyer, Angew. Chem. Int. Ed. 43 (47) (2004) 6469–6471. [27] M.G. Ryadnov, B. Ceyhan, C.M. Niemeyer, D.N. Woolfson, J. Am. Chem. Soc. 125 (31) (2003) 9388–9394. [28] R. Levy, N.T.K. Thanh, R.C. Doty, I. Hussain, R.J. Nichols, D.J. Schiffrin, M. Brust, D.G. Fernig, J. Am. Chem. Soc. 126 (32) (2004) 10076–10084. [29] D. Aili, K. Enander, J. Rydberg, I. Lundstroem, L. Baltzer, B. Liedberg, J. Am. Chem. Soc. 128 (7) (2006) 2194–2195. [30] H.S. Mandal, H.-B. Kraatz, J. Am. Chem. Soc. 129 (20) (2007) 6356– 6357. [31] M.-A. Morikawa, N. Kimizuka, Chem. Commun. (Cambridge, UK) (38) (2005) 4866–4868. [32] H. Nishikawa, T. Morita, J. Sugiyama, S. Kimura, J. Colloid Interface Sci. 280 (2) (2004) 506–510. [33] A.G. Tkachenko, H. Xie, D. Coleman, W. Glomm, J. Ryan, M.F. Anderson, S. Franzen, D.L. Feldheim, J. Am. Chem. Soc. 125 (16) (2003) 4700–4701. [34] N. Higashi, J. Kawahara, M. Niwa, J. Colloid Interface Sci. 288 (1) (2005) 83–87. [35] G. Frens, Nature (London), Phys. Sci. 241 (105) (1973) 20–22. [36] A.J. Williams, V.K. Gupta, J. Phys. Chem. B 105 (22) (2001) 5223– 5230. [37] U. Kreibig, Optical Properties of Metal Clusters, Springer, Berlin, 1995. [38] A.C. Templeton, J.J. Pietron, R.W. Murray, P. Mulvaney, J. Phys. Chem. B 104 (3) (2000) 564–570. [39] J.E. Kohn, I.S. Millett, J. Jacob, B. Zagrovic, T.M. Dillon, N. Cingel, R.S. Dothager, S. Seifert, P. Thiyagarajan, T.R. Sosnick, M.Z. Hasan, V.S. Pande, I. Ruczinski, S. Doniach, K.W. Plaxco, Proc. Natl. Acad. Sci. USA 101 (34) (2004) 12491–12496. [40] T. Ito, L. Sun, M.A. Bevan, R.M. Crooks, Langmuir 20 (16) (2004) 6940– 6945. [41] C.Y. Wang, W.E. Fu, H.L. Lin, G.S. Peng, Meas. Sci. Technol. 18 (2) (2007) 487–495. [42] G. Rialdi, J. Hermans Jr., J. Am. Chem. Soc. 88 (24) (1966) 5719–5720. [43] A. Warashina, A. Ikegami, Biopolymers 11 (3) (1972) 529–547. [44] D.T. Haynie, S. Balkundi, N. Palath, K. Chakravarthula, K. Dave, Langmuir 20 (11) (2004) 4540–4547. [45] S. Galier, H. Roux-de Balmann, J. Membr. Sci. 241 (1) (2004) 79–87. [46] T. Kawai, T. Komoto, N. Honda, Makromol. Chem. 182 (8) (1981) 2127– 2137. [47] T. Miyazawa, E.R. Blout, J. Am. Chem. Soc. 83 (1961) 712–719. [48] T. Abe, S. Hayashi, N. Higashi, M. Niwa, K. Kurihara, Colloids Surf. A Physicochem. Eng. Aspects 169 (1–3) (2000) 351–356. [49] K. Kurihara, T. Abe, N. Higashi, M. Niwa, Colloids Surf. A Physicochem. Eng. Aspects 103 (3) (1995) 265–272. [50] N.H. Lee, C.W. Frank, Langmuir 19 (4) (2003) 1295–1303. [51] S. Malynych, I. Luzinov, G. Chumanov, J. Phys. Chem. B 106 (2002) 1280–1285. [52] D.L. Allara, A. Baca, C.A. Pryde, Macromolecules 11 (6) (1978) 1215– 1220. [53] H.A. Atwater, S. Maier, A. Polman, J.A. Dionne, L. Sweatlock, MRS Bull. 30 (5) (2005) 385–389. [54] S.A. Maier, M.L. Brongersma, P.G. Kik, S. Meltzer, A.A.G. Requicha, H.A. Atwater, Adv. Mater. 13 (19) (2001) 1501–1505. [55] K.G. Thomas, S. Barazzouk, B.I. Ipe, S.T.S. Joseph, P.V. Kamat, J. Phys. Chem. B 108 (35) (2004) 13066–13068.