Article pubs.acs.org/est
Structural and Functional Effects of Cu Metalloprotein-Driven Silver Nanoparticle Dissolution Andrew J. Martinolich, Grace Park, Meagan Y. Nakamoto, Rachel E. Gate, and Korin E. Wheeler* Department of Chemistry & Biochemistry, Santa Clara University, Santa Clara, California 95053 United States S Supporting Information *
ABSTRACT: Interactions of a model Cu-metalloprotein, azurin, with 10−100 nm silver nanoparticles (NPs) were examined to elucidate the role of oxidative dissolution and protein interaction on the biological reactivity of NPs. Although minimal protein and NP structural changes were observed upon interaction, displacement of Cu(II) and formation of Ag(I) azurin species under aerobic conditions implicates Cu(II) azurin as a catalyst of NP oxidative dissolution. Consistent with NP oxidation potentials, largest concentrations of Ag(I) azurin species were recorded in reaction with 10 nm NPs (>50%). Apo-protein was also observed under anaerobic reaction with NPs of all sizes and upon aerobic reaction with larger NPs (>20 nm), where NP oxidation is slowed. Cu(II) azurin displacement upon reaction with NPs was significantly greater than when reacted with Ag(I)(aq) alone. Regardless of NP size, dialysis experiments show minimal reactivity between azurin and the Ag(I)(aq) species formed as a result of NP oxidative dissolution, indicating Cu displacement from azurin occurs at the NP surface. Mechanisms of azurin-silver NP interaction are proposed. Results demonstrate that NP interactions not only impact protein structure and function, but also NP reactivity, with implications for targeting, uptake, and cytotoxicity.
■
INTRODUCTION Abundantly used as a broad spectrum antimicrobial agent, silver nanoparticles (NPs) are some of the best studied NPs for their effects upon environmental and biological reactivity. Silver NPs represent a potential risk to a wide variety of organisms,1 including both aquatic species2,3 and essential microbial communities in soil and natural waters.4−6 The risks of silver NPs have been studied for a number of engineered properties such as particle size,4,7 shape,8 and surface coating.9 In complex environmental systems, however, the engineered properties of NPs readily change in response to environmental conditions, biological interactions, and NP exposure history. This includes changes in surface chemistry and silver NP−protein interactions. For example, silver NPs exposed to oxygen as they “age” have increased oxidative dissolution and are significantly more toxic than when freshly prepared.10 Recent reports of NP biomagnification within the food chain11,12 raise particular concerns about the long-term impact and risks of human exposure as NP reactivity evolves over time in environmental and biological systems. Changes in silver NP reactivity imbibed by these adventitious properties accumulate with biological and environmental exposure and are poorly understood.13 Surface oxidation and silver dissolution dominate silver NP reactivity (reaction 1).14 As a result, studies of toxicity and environmental impacts are complicated by the coexistence of multiple silver species in solution, including particulate zerovalent silver (Ag(s)), surface oxidized silver (Ag(I)(adsorbed), reaction 1a), and dissolved silver cations (Ag(I)(aq), reaction 1b). © XXXX American Chemical Society
2Ag(s) + 1/2O2 (aq) + 2H+ a
→ 2Ag(I)(adsorbed) + H 2O(l) b
⇌ 2Ag(I)(aq) + H 2O(l)
(1)
This oxidative dissolution mechanism also drives biological reactivity of silver NPs. Across species of organisms, lower levels of cellular reducing agents such as glutathione are measured after silver NP exposure with high levels of Ag(I)(aq).15 Numerous in vivo studies with silver NPs attribute bactericidal activity and toxicity to Ag(I)(aq).4,5,10 Silver NP antimicrobial activity, toxicity, and reactivity are not solely driven by dissolved Ag(I)(aq). Recent studies reveal that bactericidal effects arise not only from Ag(I)(aq) species, but also from the NPs themselves.5,16 For instance, colloidal silver NPs often demonstrate toxicity and bactericidal effects at significantly lower concentrations of silver than studies of Ag(I)(aq).17−19 The presence of O2(g) or common ligands can differentially affect the cellular activity of silver NPs vs Ag(I)(aq).6 This work underscores the need for detailed research to understand the reactivity at the biological interface of silver NPs and differentiate it from the reactivity of dissolved Ag(I)(aq). Despite the complexities of separating the biological reactivity of Ag(I)(aq) and silver NPs, reports on the Received: March 6, 2012 Revised: May 3, 2012 Accepted: May 7, 2012
A
dx.doi.org/10.1021/es300901h | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
Characterization of Silver NPs. BioPure, citrate coated spherical silver NPs were purchased from Nanocomposix (La Jolla, CA). Monodispersity of the silver NPs was confirmed by Nanocomposix via dynamic light scattering size distribution (DLS) and transmission electron microscopy (TEM) images, indicating relatively stable silver NP solutions with similar morphologies and size distribution. Particle sizes and hydrodynamic radii were determined inhouse by DLS using a ZetaPlus Particle Sizing instrument (Brookhaven Instruments Corp, NY). Temperature was held at 25 °C; the scattering angle was 90°. Average measurements are reported. Prior to TEM (Hitachi H-9500 at 300 kV), azurin and silver NPs were reacted for 6 h as described below, then azurin-NP solutions were centrifuged to remove excess protein. Sample plates were prepared by depositing 4 μL of the homogeneous solution (silver NPs, or silver NP- azurin mixture) on a 400-mesh copper grid (SPI Supplies, PA) coated with Formvar, then air-dried at room temperature. Silver NP-Azurin Interaction Experiments. Silver NP reactivity with proteins was compared across NPs sizes to assess the effects of surface curvature and reduction potential on reactivity. The available NP surface area was kept constant across NP sizes (rather than maintaining consistent total silver concentration or number of particles) to exclude surface area and available binding sites as an experimental variable. Surface area of the silver NPs was calculated assuming each NP is a sphere. As outlined by Mattoussi et al.,36 the theoretical number of proteins that each spherical particle can accommodate on its surface (Nazurin) can then be derived from steric considerations using the following expression: Nazurin = 0.65 (Rcomplex3 − RNP3)/(Razurin3), where RNP is the silver NP radius; Razurin is estimated at approximately 3 nm ; and Rcomplex is the radius of the silver NP plus the bound protein (RNP + 2Razurin). Cu(II) azurin (100 μM) was reacted with silver NPs at a constant 1.414 × 1014 nm2 surface area. NP concentrations (calculated by particle molarity, not silver concentrations) were 3.74 nM 10 nm NPs, 0.955 nM 20 nm NPs, 0.416 nM 30 nm NPs, 0.233 nM 40 nm NPs, 0.156 nM 50 nm NPs, 0.104 nM 60 nm NPs, and 0.0374 nM 100 nm NPs. Supernatant control solutions, which were prepared by simply centrifuging NPs out of solution, were also used to test reactivity of residual reactants. In O2(g) free experiments, protein and NP solutions were purged with Ar(g) prior to mixing. Kinetics of silver NP azurin reactivity were monitored by UV−vis spectroscopy. Cu(II)-azurin, a type-1 blue Cu protein, has a strong ligand-to-metal charge transfer (LMCT) band assigned to the S(Cys-π)→Cu(II) dx2 − y2 transition (λmax = 630 nm). By monitoring the LMCT band, Cu(II) ligation was followed in a region of the spectrum separate from the plasmon resonance band of 10 nm resulted in increased concentrations of apo-azurin and lessened concentrations of Ag(I) species. Role of O2(g) in Protein Catalyzed Oxidative Dissolution of Silver NPs. Upon removal of oxygen from the reaction, the contributions of Ag(I) and Ag(0) in silver NP reactivity can be compared upon reaction with Cu(II) azurin (Figures 1C and 2C). ESI-MS spectra of Cu(II) azurin reacting with silver NPs under Ar(g) shows that anaerobic reactions increase formation of apo-azurin and decrease formation of Ag(I) azurin species for small NPs (30 nm (Figure 2b), consistent with the observation that apo-azurin formation increases significantly when reacted with NPs in the absence of O2(g) (Figure 2c). In confirmation of LMCT band results with control NP
indeed dependent upon the size of silver NP in reaction with the smallest silver NPs, causing rapid and significant disruption to the native bound Cu(II) in azurin. The most dramatic effect from 10 nm silver NPs resulted in 48% hypochromicity after 6 h. Silver NPs >30 nm demonstrated a significantly smaller effect with a maximum 25% hypochromicity. This is consistent with NP reduction potentials, which decrease as silver NP size decreases.15,38 The minimal impact of the supernatant control solutions suggests the LMCT band decrease is not a result of contaminants. Metalation of azurin as a result of Ag(I) or silver NP interaction was evaluated with ESI-MS (Figure 2, vide infra for anaerobic results shown in Figure 2C). ESI- MS spectra taken after 6 h of reactivity revealed the formation of both apo- and Ag(I) azurin species. Due to the more negative reduction potential of smaller NPs, the concentration of Ag(I) azurin was highest in reaction with 10 nm silver NPs (Figure 2B). Ag(I) azurin species included azurin bound to 2Ag(I), 1Ag(I), and 1Ag(I)1Cu (Supporting Information). These data reveal a second metal binding site on azurin, consistent with previous reports of azurin in reaction with soft metal cations.36,37 The second metal binding site is believed to be at one or more of azurin’s surface methionines. For the smallest particles, 10 nm, the percentage of 2Ag(I) azurin species formed was greatest, while the distribution between Ag(I) and 2Ag(I) for the larger particles was roughly even. With little surface oxidation, NPs D
dx.doi.org/10.1021/es300901h | Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Environmental Science & Technology
Article
supernatant solutions,
