Photocatalytic Reactive Oxygen Species Formation by Semiconductor

May 25, 2016 - Semiconductor–metal hybrid nanoparticles manifest efficient light-induced spatial charge separation at the semiconductor–metal inte...
2 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Photocatalytic Reactive Oxygen Species Formation by Semiconductor−Metal Hybrid Nanoparticles. Toward Light-Induced Modulation of Biological Processes Nir Waiskopf,†,‡ Yuval Ben-Shahar,† Michael Galchenko,† Inbal Carmel,† Gilli Moshitzky,‡ Hermona Soreq,*,‡ and Uri Banin*,† †

Institute of Chemistry and the Center for Nanoscience and Nanotechnology ‡Department of Biological Chemistry and the Edmond and Lily Safra Center of Brain Science, The Hebrew University of Jerusalem, Safra Campus, Givat Ram, Jerusalem 91904, Israel S Supporting Information *

ABSTRACT: Semiconductor−metal hybrid nanoparticles manifest efficient light-induced spatial charge separation at the semiconductor−metal interface, as demonstrated by their use for hydrogen generation via water splitting. Here, we pioneer a study of their functionality as efficient photocatalysts for the formation of reactive oxygen species. We observed enhanced photocatalytic activity forming hydrogen peroxide, superoxide, and hydroxyl radicals upon light excitation, which was significantly larger than that of the semiconductor nanocrystals, attributed to the charge separation and the catalytic function of the metal tip. We used this photocatalytic functionality for modulating the enzymatic activity of horseradish peroxidase as a model system, demonstrating the potential use of hybrid nanoparticles as active agents for controlling biological processes through illumination. The capability to produce reactive oxygen species by illumination on-demand enhances the available peroxidase-based tools for research and opens the path for studying biological processes at high spatiotemporal resolution, laying the foundation for developing novel therapeutic approaches. KEYWORDS: Semiconductor−metal hybrid nanoparticles, photocatalysis, reactive oxygen species, peroxidase

S

they show profound advantages over organic dyes and fluorescent proteins; mainly in terms of wavelength tunability, light sensitivity, and photochemical stability.15 Moreover, they have been suggested also for photodynamic therapy and as light activated modulators of biological functions in neuronal stimulation applications, serving as alternatives to other approaches utilizing small molecule photoswitches, caged molecules enabling photorelease, and opto-genetics tools.16−21 Previous reports have shown that in the presence of oxygen, light excitation of semiconductor nanocrystals can result in the formation of ROS.22,23 The amount and type of ROS were found to depend on the composition and surface coating of the nanocrystals. Moreover, it was suggested that the produced ROS can be used to activate enzymes such as horseradish peroxidase (HRP).24,25 This would depend on reduction of molecular oxygen with two protons and two electrons transferred from the conduction band of the semiconductor nanocrystals (eq 1)

emiconductor−metal hybrid nanoparticles (HNPs) manifest efficient light-induced charge separation.1−3 Upon light excitation, the semiconductor component generates an electron−hole pair (exciton), followed by electron charge transfer to the metal domain due to energy band alignment of the semiconductor-metal nanojunction.4,5 This, combined with the known catalytic characteristics of metal nanocrystals, made these HNPs leading candidates as photocatalysts for energy and environmental applications.1 In particular, HNPs are being investigated as photocatalysts in water splitting for hydrogen generation under light irradiation as a means of directly converting solar energy to chemical energy stored in a fuel.6−8 Within these experiments, HNPs show efficient hydrogen generation, whereas semiconductor nanocrystals have negligible activity. Another photocatalytic application is the photoreduction of carbon dioxide, using solar energy for promoting the reduction of green-house gas contents along with alternative sustainable avenues for production of methane and other hydrocarbon fuels.9,10 Herein, we introduce for the first time the use of visible band gap semiconductor−metal HNPs as photocatalysts enabling the formation of reactive oxygen species (ROS), a key step toward establishing their function in modulating biological activity by light. Semiconductor nanocrystals have been investigated for their use in diverse bioimaging and sensing applications,11−14 where © XXXX American Chemical Society

O2 + 2e− + 2H+ → H 2O2 ,

E° = −0.682 V

(1)

Received: March 28, 2016 Revised: May 10, 2016

A

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters However, these systems show relatively limited efficiency for production of radicals and peroxides, resulting from inherently low efficiency of charge transfer due to strong competing decay pathways such as electron−hole recombination, along with oxidation and etching of the semiconductor nanocrystals surface.26 Radicals and peroxides are often considered as toxic entities that may induce significant damage and even result in cells death. However, their controlled production is required in numerous assays and biological functions.27,28 For example, HRP uses H2O2 as a substrate and is widely employed in standard assays for detection and quantification of proteins such as in enzyme-linked immunosorbent assay, immunohistochemistry, quantification after gel electrophoresis, and as signal amplifier in various sensors.17,29 Furthermore, diverse biological pathways, including metabolic and immune reactions include and depend on H2O2. Thyroid peroxidase (TPO) depends on H2O2 for the production of the thyroid hormones including thyroxine (T4) and triiodothyronine (T3). Interference with TPO activity may lead to the autoimmune disease Hashimoto’s thyroiditis resulting in fatigue, joint and muscle pain, depression, and more.30 In addition, potent antiparasite enzymes of the immune system, including myeloperoxidase, eosinophil peroxidase, and lacto-peroxidase use H2O2 in their fight against pathogens.31,32 The ability to induce such activities on demand by optical modulation has a potential value for biological research and biomedical applications. These ideas, combined with the interest to understand the redox activity of HNPs in aerobic conditions, have motivated the work reported herein. Using well developed metal-tipped semiconductor nanorod HNPs as model systems, we observed and characterized light-induced ROS formation through combination of spectroscopic assays, polarography, and electron paramagnetic resonance (EPR). Hydrogen peroxide, superoxide, and hydroxyl radicals were successfully detected, and the photocatalytic activity of the hybrid nanoparticles was found to be conspicuously improved over that of the semiconductor nanocrystals. This photocatalytic functionality was used for modulation of horseradish peroxidase enzymatic activity as a model system, demonstrating the potential use of hybrid nanoparticles as active agents to control biological processes through illumination. Our model HNP system consisted of metal-tipped seeded CdSe/CdS-Au nanorods, which were chosen thanks to their improved photocatalytic performance for hydrogen generation. The synthesis of the HNPs was carried out in consecutive steps. First, CdSe/CdS seeded nanorods (NRs) were synthesized by a previously described hot-injection method.21 Specifically, CdSe/CdS seeded NRs with a small CdSe seed (2.3 nm) leading to quasi type-II band alignment were prepared, as they offer improved charge separation compared to type I systems.33 Next, selective metal growth on the semiconductor nanorod apex was carried out (see Supporting Information) under dark conditions and at room temperature.26 Phase transfer with polyethylenimine (PEI) was used to replace the original hydrophobic ligands and to achieve well-dispersed HNPs in aqueous solutions34 (see Supporting Information). Figure 1a presents the transmission electron micrograph (TEM) of the HNPs (see Figure S2 for their size distribution histograms) showing their matchstick-like structure of 52 ± 4 nm × 4.4 ± 0.4 nm CdSe/CdS NRs with single gold tips on the apex (stronger contrast, typically 1.6 nm in diameter). The spectra of both types of NPs exhibit a sharp rise at 480 nm, attributed to

Figure 1. Optical and structural characterization of HNPs along with their HRP photocatalytic activation. (a) Absorption spectra of CdSe/ CdS NRs (blue) and CdSe/CdS-Au HNPs (black), showing similar features. The contribution of the small Au tips is mainly manifested in the small increase of the absorption tail toward the red. Inset- TEM image presenting matchstick-like structure of 52 ± 4 nm × 4.4 ± 0.4 nm CdSe/CdS nanorods with small gold tips on their apex (stronger contrast, 1.6 nm in diameter). (b) Staircase behavior of HRP stimulation while turning off and on the excitation source, indicated by black and blue arrows, respectively. (Inset) Detection of HRP product by absorbance measurements with its characteristic peak around 495 nm, growing with increased irradiation times of the HNP and HRP solution.

the CdS rod absorption onset (Figure 1a). Note that the band gap of the seeded rod is at ∼560 nm (Figure S1). Upon growth of the small Au tip, an increased absorption tail toward the red is observed.35 Photoinduced modulation of peroxidases activity by HNPs was examined using an assay for HRP activity that catalyzes the production of quinoneimine dye in the presence of H2O2, 4aminoantipyrine (4-AAP), and phenol (Figure S3a). Excitation of the HNPs with a 405 nm 20 mW/cm2 laser resulted in the formation of a new absorption peak at 495 nm, which is a signature of the formation of HRP products, that significantly increases with irradiation time (Figure 1b, inset). In addition, turning the illumination source on and off (blue and black arrows, respectively in Figure 1b) resulted in a staircase behavior, where the absorption increases only while the light is on, proving a photoswitching of HRP activity by excitation of the HNPs. Furthermore, control experiments without oxygen, light or HNPs did not cause significant absorption changes, confirming the central inter-related role of these three elements in product formation. Without HRP, some product formation is also detected (Figure S3). Hence, to represent the net contribution of the HNPs to the effective enzyme action, this B

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

smaller molar extinction coefficient of the CdSe QDs relative to the NRs and HNPs, the actual concentrations of the QDs are significantly higher and yet HNPs reveal the higher efficiency. Importantly, light excitation of 5 nm gold nanoparticles did not show any product formation, supporting the hypothesis that the higher product formation results from the synergistic lightinduced charge separation phenomenon across the semiconductor−metal nanojunction. Similar behavior was also observed when comparing CdS NRs and CdS−Au HNPs (Figures S2 and S4). This further substantiated the advantage of the use of HNPs over CdS QDs reported in previous literature.25 The inherent small capacities of the nanoparticles without hole acceptor and SOD to produce H2O2 showed a similar trend, HNPs > NRs > CdSe QDs (Figure 2b and Figure S4). The addition of SOD, which is used by cells to transform superoxide radicals to H2O2 following reaction 2, resulted in about 7-fold increase in product formation in comparison to the native capacities (Figure 2b). The enhancement of the signal by SOD provides direct evidence for the formation of superoxide radicals by the HNPs, which then contribute to the overall H2O2 generation.

signal was subtracted from the kinetic data analyzed and presented. We first examined and compared the structural effect of the nanoparticles on the efficiency of product formation, using solutions of CdSe quantum dots (QDs), CdSe/CdS seeded NRs, and CdSe/CdS-Au HNPs with similar optical density at the excitation wavelength of 405 nm. In this manner, a total apparent efficiency comparison is conducted between the different types of nanosystems. The capacity of each type of NPs to stimulate HRP activity was examined with and without addition of hole acceptor and SOD. Figure 2a shows that HNPs

Cu+(SOD) + O−2 + 2H+ → Cu+2(SOD) + H 2O2

(2)

Ethanol, although not biologically relevant and not the best sacrificial hole acceptor, was selected for our study, due to minimal interference to the used characterization assays. The addition of ethanol resulted in faster kinetics for both the HNPs and NRs systems (Figure 2c). Notably, although the addition of hole acceptor affected both systems, its effect on the HNPs efficiency was found to be much more pronounced. Moreover, while in experiments without ethanol the activity was reduced with the irradiation time until a plateau was seen (Figure S3), the addition of ethanol yielded longer sustained activity and stability. These observations are compatible with the notion that sacrificial hole acceptors enhance the photocatalytic performance and prolong the stability of HNPs by minimizing the probability for electron−hole recombination and photooxidation of their semiconductor component.36,37 To investigate the mechanism of the ROS formation by HNPs, we studied the kinetics of soluble molecular oxygen consumption under illumination. Figure 3a presents normalized kinetic measurements of oxygen consumption in a closed cell as measured by polarography at isothermal conditions. HNPs-PEI presented a considerably higher rate of oxygen consumption, 1.7 μmol L−1 sec−1, compared to the rate of bare NRs-PEI, 0.8 μmol L−1 sec−1. This rate is consistent with and similar to the rate of HRP product formation (0.6 μmol L−1 sec−1), suggesting that the formation of H2O2 and superoxide are the dominant processes. The formation of additional possible radicals was measured either via fluorescence assay or directly by electron paramagnetic resonance spectroscopy (EPR). Figure 3b shows kinetic measurements of hydroxyl radicals production by HNPs and NRs by a fluorescence assay. Nanoparticles were excited in the presence of the hydroxyl radical indicator, terephthalic acid (TPA), which reacts to produce 2-hydroxyterephthalic acid (HTA, see Supporting Information for further details). The latter has a distinctive emission peak at 425 nm upon 310 nm excitation (inset). Superior hydroxyl radical formation was observed when stimulating the HNPs in comparison to the bare NRs (Figure 3b). The observed higher formation of hydroxyl radicals could originate from either the excited holes in the semiconductor valence band22 or by decomposition of H2O2.

Figure 2. Comparative study of product formation by HRP activation using HNPs, NRs, and QD seeds in different conditions. (a) Light stimulation of hybrids (black) in the presence of SOD and ethanol results in 3- and 15-fold higher product formation after 500 sec, than when stimulating bare nanorods (blue) or CdSe seeds (red), respectively. (b,c) Comparative bar charts for the HRP activity in different conditions, plotting product formation after 500 sec (note the different scale in b,c). Addition of SOD to the solution increases the efficiency of all of the tested nanoparticle systems and the efficiency is further enhanced upon adding the hole acceptor (HS), which is ethanol in this case. Notably, the hole acceptor and SOD effects on the efficiency of the HNPs system was much more pronounced compared to the NRs and QDs.

excitation in the presence of hole acceptor and superoxide dismutase (SOD) results in ∼3 times faster kinetics and final product concentration as compared to the excitation of bare NRs, showing a clear superiority for the use of HNPs for applications requiring H2O2 (black versus blue curve). In comparison, CdSe quantum dots (red curve) showed a rather limited capacity to produce H2O2, due to alternative charge recombination processes in the absence of effective electron− hole charge separation in the small volume QD. Moreover, given the same optical density used in the experiments and the C

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 3. HNPs photocatalytic ROS formation mechanism. (a) Kinetic measurements of molecular oxygen consumption by HNPs and NRs coated with PEI upon light illumination using polarography. (b) Kinetic measurements of hydroxyl radical formation using the fluorescence TPA assay (inset). Note the significantly faster and more efficient hydroxyl radicals production upon stimulation of HNPs (black) in comparison to NRs (blue). Addition of ethanol as a hole scavenger to the hybrids solution (orange) prevented this reaction, yielding similar signals to those of control TPA alone (purple). (c) EPR measurements following excitation of HNPs (upper panel) and NRs (lower panel), and corresponding fits (red). The results show near-solely (∼90%) superoxide production for the bare NRs, whereas the HNPs signal reveals additional peaks which are attributed to significant hydroxyl radicals production. (d) Time-resolved measurements for the superoxide radical signal at static magnetic field of 3448G (star mark in panel c) show higher superoxide signal buildup for NRs over HNPs, followed by a plateau when light is on (blue arrows), resulting from balanced formation and destruction of the radicals. Fast decay of the radical signal is seen upon turning off the excitation source (black arrows).

Table 1. Area Quantification Results for the Formation of Different ROS Following Visible Light Excitation of NPs Coated by PEI with and without Ethanola

a

nanoparticle type

ethanol

•OH

•OOH

CdSe/CdS NRs CdSe/CdS NRs CdSe/CdS-Au HNPs CdSe/CdS-Au HNPs

− + − +

3%

94% 94% 54% 50%

31% 15%

SO3•−

CH3C•HOH 3%

12% 16%

16%

correlation 0.94 0.95 0.93 0.9

All samples showed 3% of a nonspecific signal.

coupling of aN = 14.1G, aH = 11.3G, and aH = 1.25G that supports the observation of significant superoxide production triggered by light excitation. However, in the NR system this signal accounted to more than 90% of the total signal, whereas its contribution in the case of the HNPs was only ∼50%. In the hybrid system, we found a second strong signal of DMPO−OH with a hyperfine coupling of aN = aH = 15G, suggesting that the HNPs also produce significant amounts of hydroxyl radicals. This observation is compatible with the observations of the fluorescence assay for hydroxyl radical formation and was confirmed by the addition of ethanol, which as noted above can trap hydroxyl radicals. Ethanol addition yielded a third radical specie with hyperfine coupling of aN = 15.8G and aH = 22.8G, which indicates the presence of CH3C*HOH, an ethanolderived radical adduct (Table 1 and Figure S5). Interestingly, in the HNP systems a weak signal with hyperfine coupling of aN = 14.4G and aH = 15.9G was also observed. This can be assigned to sulfite radical anions,38 which are typically generated by the

This is again in line with more efficient charge separation in the HNPs and further demonstrates that the hybrids can be applied to achieve efficient and controlled production of hydroxyl radicals. For both systems, the addition of ethanol suppressed HTA production (orange squares). This may result from dual functioning of ethanol, both as a hole acceptor and as a hydroxyl radical scavenger. Hydroxyl radicals may contribute to unspecific effects on biological systems and can also rapidly attack the superoxide radicals. We surmised that this might somewhat limit the effect of added SOD in the case of HRP activation, and hypothesized that adding a sacrificial hole acceptor would increase both the efficiency and the specificity of the system. EPR measurements on both HNPs and NRs in phosphate buffered saline, using 5,5dimethyl-1-pyrroline N-oxide (DMPO) as the radical trapping agent (black/blue) and the following theoretical simulations (red) are presented in Figure 3c (further experimental and model details are summarized in Tables 1 and S1). Both systems showed a strong signal of DMPO-OOH with hyperfine D

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 4. Summary scheme showing the different pathways for ROS formation after HNP light activation, and their use for HRP modulation. Excitation of the semiconductor rods results in charge separation followed by reduction of molecular oxygen by the excited electrons. This results in direct formation of H2O2 that can be used as a substrate for HRP or in formation of superoxide that can be converted to H2O2 with the aid of SOD. In parallel, the holes can be used to produce hydroxyl radicals or could be scavenged by hole acceptors.

(PSMA),42 which is an amphiphilic polymer that coats the original hydrophobic ligands (see Supporting Information). The surface coating was found to exert significant effects over HRP activation (Figure 5a). The lowest production was observed for HNPs coated with PSMA (orange). In comparison, HNPs coated with PEI showed the highest production efficiency (black) with 5-fold higher product formation than thiolate ligands, such as GSH and MPA (red and blue). Similarly, the rate and overall decrease in oxygen concentration was significantly lower for nanoparticles coated with GSH (Figure S6a). Still, the GSH-coated systems showed similar trends of increased functionality of HNPs versus NRs regarding oxygen consumption, HRP activation, and hydroxyl radical formation (Figure S6). The surface coating effect was reminiscent of but distinct from its effects on the photocatalytic activity for hydrogen production.26 Whereas PSMA coating was reported to allow higher hydrogen production in comparison to thiolate ligands, this coating was found to be less favorable for production of hydrogen peroxide. This should be further investigated and may be attributed to different factors, such as the surface coating effect on HNPs surface defects passivation, band alignment, the ability of molecules to penetrate and migrate from the HNPs surface and the interaction of the surface coating with the formed radicals and peroxides. We next tried to challenge the use of this concept in biological systems where many biomolecules can interact and consume the produced radicals and hydrogen peroxide. To mimic the mammalian circulation conditions, we ran the same set of experiments in solutions containing 10% fetal calf serum (Figure S8a). In the absence of hole acceptor, this set of experiments showed similar results for stimulating HNPs with or without serum. Interestingly, in the presence of ethanol as hole acceptor the kinetics of HRP activation following HNPs stimulation was similar with or without serum in the first minutes. However, at longer irradiation times the HNPs in serum showed an improved stabilized behavior over the HNPs in buffer, which showed slight kinetic decreases. This result may indicate the formation of a protein corona on the HNPs in serum that protects them.43,44 Nevertheless, as ethanol is not biologically relevant, various biological molecules were examined and confirmed for their ability to increase the photocatalytic capacity of HNPs under biological conditions (Figure S7). This set of experiments supports the consideration of the quantification of ROS production, as was described above, to be valid for in vitro experiments and further predicts prospective use of HNPs stimulation for H2O2 production in biological systems.

reaction of hydrogen peroxide with sulfite (SO32−) that may have been formed by photo-oxidation of the CdS rod surface.39 Figure 3d presents time-resolved EPR measurements, following the superoxide signal at static magnetic field of 3448G (star mark in Figure 3c). A fast buildup of the signal was revealed during illumination (blue arrows) followed by a plateau when reaching equilibrium with the radicals consumed via further reactions. Decay of the signals was observed with the excitation light turned off (black arrows). The plateau in the NRs case (blue curve) emerged at higher signal values than those of the HNPs (gray curve), suggesting increased superoxide formation with the NRs. While this observation appears contrary to the faster consumption of molecular oxygen by HNPs (Figure 3a), this apparent discrepancy can be resolved considering that oxygen may be consumed also by the competing pathway forming directly H2O2 as seen in Figure 2b for HNPs without SOD. Furthermore, superoxide can also react with hydroxyl radicals, which and as seen in Figure 3b are indeed formed more efficiently by the HNPs compared to the NRs (Figure 3b).40 Taken together, the various ROS detection assays, the polarography study, and the EPR infer the following complex scheme of radicals formation, as presented schematically in Figure 4. Light excitation of the semiconductor rod results in charge separation with the holes remaining on the rod and the electrons transferring to the metal tip. The electrons can reduce molecular oxygen to directly produce H2O2 or to produce superoxide that could be transformed to H2O2 by SOD. The holes on the other hand can oxidize water and hydroxide to produce hydroxyl radicals. Addition of hole acceptors effectively competes with direct formation of hydroxyl radicals by scavenging the holes. The acceptors may also neutralize hydroxyl radicals by secondary reaction. Notably, they have a major effect in diminishing the competing hole−electron backrecombination and hence significantly increase the reduction pathways leading either to direct or SOD catalyzed H2O2 formation. After gaining these insights into the mechanisms of ROS formation by HNPs, we addressed an additional parameter affecting their photocatalytic activity, their surface coating. Surface coating effects on ROS formation by semiconductor nanocrystals were reported23 and for HNPs were studied in the context of the photocatalytic activity in hydrogen generation.26 To this end, nanoparticles were transferred to water by additional phase transfer techniques, including ligand exchange41 with glutathione (GSH) and 3-mercaptopropionic acid (MPA), or by poly(styrene-co-maleic anhydride) E

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

those used for HRP stimulation (highest concentrations). These results were also supported by Live/Dead bioassay, which is based on staining mitochondrial ATPase that decorates live cells in fluorescent green and dead cells in red.45 Under dark conditions, HNPs did not show significant toxicity with the highest tested concentration (Figure S8b). The effect of ROS formation following light excitation on cell viability was also examined. Figure 5c shows results of Live/ Dead assay after short illumination with 405 nm 20 mW/cm2 LED on the K-562 cells following their incubation for 1 h with 2.5 nM (left) and 0.5 nM (right) of HNPs. For the higher HNPs concentration, the enhanced oxygen consumption and ROS production, as measured in the presence of fetal calf serum, lead to significant cells’ death (Figure 5c left). This observation suggests the possible utilization of these HNPs in photodynamic therapy applications. On the other hand, lower HNPs concentration mainly manifested live cells, consistent with limited ROS formation (Figure 5c right). This finding demonstrated that light-controlled production of ROS by HNPs can be performed without significant cell damage, and hence leaves open the doorway to utilize them for lightmediated modulation of biological activity. Notably, in these experiments with GSH coating some cellular uptake is expected46 that can be minimized by different surface coatings as shown for semiconductor nanocrystals overcoated with PEG.47−49 In conclusion, we reported on ROS formation via light irradiation of HNPs, showing significantly improved yield over NRs, and studied the various oxidation/reduction processes taking place in physiologically relevant conditions. Both systems produce superoxide radicals, while HNPs also showed an enhanced capacity to produce hydrogen peroxide and hydroxyl radicals in comparison to bare nanorods. The demonstrated ability to activate HRP by light using HNPs is a proof of concept for a general approach of applying such systems for modulating enzyme properties and affecting biological pathways through radicals and/or peroxides formation. Moreover, the ability to control nanoparticle properties by tuning the energy band alignment can open further new avenues for efficient modulation of specific biological pathways involving redox reactions with visible light. The addition of hole acceptors and SOD increases the efficiency of H2O2 production and can serve for increasing the specificity of the system by minimizing the diffusion and undesired interaction of radicals from the HNPs surface to biomolecules other than the peroxidases. Biotoxicity studies demonstrate the versatile utilization of HNPs in biomedical application by tuning the extent of ROS production. Aggressive production may be used for photodynamic therapy; in comparison, cell viability was reasonably maintained under the light-controlled limited formation of ROS. HRP is widely used in detection and quantification methods and other peroxidases play roles in metabolic and immune pathways, highlighting the relevance of this approach to biological processes. In fact, some peroxidases are expressed and function mainly outside the cells, which further limits toxicity concerns due to internalization of HNPs and intracellular ROS production. For example, thyroid peroxidase plays its biological role by incorporating hydrogen peroxide in thyroid hormones synthesis on the apical membrane of the thyroid follicular cell.50 Additionally, HNPs can be conjugated to delivery agents and to substrates, including implantable substrates, to provide local specific

Figure 5. HNPs surface effects and biocompatibility properties. (a) HRP activity upon light stimulation of HNPs with different surface coatings in the presence of SOD and ethanol. HNPs coated with PEI (black) show 5-fold higher efficiency in comparison to thiolate ligands, such as GSH (red) and MPA (blue). PSMA (orange) showed the lowest efficiency. (b) MTT viability assay with cultured K-562 cells shows that their incubation for 24 h with different concentrations of NRs and HNPs did not significantly affect their viability. (c) Live/ Dead assay 24 h after illumination on cells incubated with 2.5 and 0.5 nM of HNPs (left and right, respectively). Higher HNPs concentration shows significant cells’ death (red stained cells), an outcome that indicates a potential use for photodynamic therapy. Lower HNPs concentration showed mostly live cells (green stained cells), suggesting cells’ viability under light-controlled ROS production, scale bar is 100 μm.

To further address this potential, we investigated the biotoxicity of HNPs for the first time. For this purpose, we examined the biocompatibility of the HNPs by incubating cultured K-562 human bone marrow cells for 24 h under dark conditions with different concentrations of HNPs or NRs coated with GSH. In these sets of experiments, we used GSH surface coating to minimize toxicity resulting from the use of PEI. Figure 5b shows the results of an MTT viability assay, presenting no significant effects for either NP types following incubation with low NP concentrations, and limited toxicity when exposing the cells to similar concentrations of HNPs as F

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

(7) Aronovitch, E.; Kalisman, P.; Mangel, S.; Houben, L.; Amirav, L.; Bar-Sadan, M. J. Phys. Chem. Lett. 2015, 6, 3760−3764. (8) Ben-Shahar, Y.; Scotognella, F.; Kriegel, I.; Moretti, L.; Cerullo, G.; Rabani, E.; Banin, U. Nat. Commun. 2016, 7, 10413−10419. (9) Viste, P.; Plain, J.; Jaffiol, R.; Vial, A.; Adam, P. M.; Royer, P. ACS Nano 2010, 4, 759−764. (10) Manzi, A.; Simon, T.; Sonnleitner, C.; Döblinger, M.; Wyrwich, R.; Stern, O.; Stolarczyk, J. K.; Feldmann, J. J. Am. Chem. Soc. 2015, 137, 14007−14010. (11) Freeman, R.; Willner, I. Chem. Soc. Rev. 2012, 41, 4067−4085. (12) Mattoussi, H.; Palui, G.; Na, H. B. Adv. Drug Delivery Rev. 2012, 64, 138−66. (13) Kovtun, O.; Arzeta-Ferrer, X.; Rosenthal, S. J. Nanoscale 2013, 5, 12072−81. (14) Medintz, I. L.; Stewart, M. H.; Trammell, S. A.; Susumu, K.; Delehanty, J. B.; Mei, B. C.; Melinger, J. S.; Blanco-Canosa, J. B.; Dawson, P. E.; Mattoussi, H. Nat. Mater. 2010, 9, 676−84. (15) Chan, W. C. W.; Nie, S. M. Science 1998, 281, 2016−2018. (16) Bareket, L.; Waiskopf, N.; Rand, D.; Lubin, G.; David-Pur, M.; Ben-Dov, J.; Roy, S.; Eleftheriou, C.; Sernagor, E.; Cheshnovsky, O.; Banin, U.; Hanein, Y. Nano Lett. 2014, 14, 6685−6692. (17) Timor, R.; Weitman, H.; Waiskopf, N.; Banin, U.; Ehrenberg, B. ACS Appl. Mater. Interfaces 2015, 7, 21107−21114. (18) Brown, K. A.; Wilker, M. B.; Boehm, M.; Hamby, H.; Dukovic, G.; King, P. W. ACS Catal. 2016, 6, 2201−2204. (19) Malvindi, M. A.; Carbone, L.; Quarta, A.; Tino, A.; Manna, L.; Pellegrino, T.; Tortiglione, C. Small 2008, 4, 1747−1755. (20) Wang, S. H.; Riedinger, A.; Li, H. B.; Fu, C. H.; Liu, H. Y.; Li, L. L.; Liu, T. L.; Tan, L. F.; Barthel, M. J.; Pugliese, G.; De Donato, F.; D’Abbusco, M. S.; Meng, X. W.; Manna, L.; Meng, H.; Pellegrino, T. ACS Nano 2015, 9, 1788−1800. (21) He, W.; Kim, H. K.; Wamer, W. G.; Melka, D.; Callahan, J. H.; Yin, J. J. J. Am. Chem. Soc. 2014, 136, 750−7. (22) Ipe, B. I.; Lehnig, M.; Niemeyer, C. M. Small 2005, 1, 706−709. (23) Rajendran, V.; Lehnig, M.; Niemeyer, C. M. J. Mater. Chem. 2009, 19, 6348−6353. (24) Rajendran, V.; Konig, A.; Rabe, K. S.; Niemeyer, C. M. Small 2010, 6, 2035−2040. (25) Fruk, L.; Rajendran, V.; Spengler, M.; Niemeyer, C. M. ChemBioChem 2007, 8, 2195−2198. (26) Ben-Shahar, Y.; Scotognella, F.; Waiskopf, N.; Kriegel, I.; Dal Conte, S.; Cerullo, G.; Banin, U. Small 2015, 11, 462−471. (27) Gough, D. R.; Cotter, T. G. Cell Death Dis. 2011, 2. (28) Dickinson, B. C.; Peltier, J.; Stone, D.; Schaffer, D. V.; Chang, C. J. Nat. Chem. Biol. 2011, 7, 106−12. (29) Krainer, F. W.; Glieder, A. Appl. Microbiol. Biotechnol. 2015, 99, 1611−25. (30) Grasberger, H. Mol. Cell. Endocrinol. 2010, 322, 99−106. (31) Reis, A. C.; Alessandri, A. L.; Athayde, R. M.; Perez, D. A.; Vago, J. P.; Avila, T. V.; Ferreira, T. P. T.; de Arantes, A. C. S.; Coutinho, D. D.; Rachid, M. A.; Sousa, L. P.; Martins, M. A.; Menezes, G. B.; Rossi, A. G.; Teixeira, M. M.; Pinho, V. Cell Death Dis. 2015, 6, e1632. (32) Klebanoff, S. J.; Kettle, A. J.; Rosen, H.; Winterbourn, C. C.; Nauseef, W. M. J. Leukocyte Biol. 2013, 93, 185−198. (33) Zhu, H. M.; Song, N. H.; Lv, H. J.; Hill, C. L.; Lian, T. Q. J. Am. Chem. Soc. 2012, 134, 11701−11708. (34) Nann, T. Chem. Commun. 2005, 1735−1736. (35) Varnavski, O.; Ramakrishna, G.; Kim, J.; Lee, D.; Goodson, T. J. Am. Chem. Soc. 2010, 132, 16−17. (36) Berr, M. J.; Wagner, P.; Fischbach, S.; Vaneski, A.; Schneider, J.; Susha, A. S.; Rogach, A. L.; Jackel, F.; Feldmann, J. Appl. Phys. Lett. 2012, 100, 223903. (37) Wu, K. F.; Chen, Z. Y.; Lv, H. J.; Zhu, H. M.; Hill, C. L.; Lian, T. Q. J. Am. Chem. Soc. 2014, 136, 7708−7716. (38) Buettner, G. R. Free Radical Biol. Med. 1987, 3, 259−303. (39) Hsieh, Y. H.; Huang, C. P. Colloids Surf. 1991, 53, 275−295. (40) Cohen, G.; Heikkila, R. E. J. Biol. Chem. 1974, 249, 2447−2452. (41) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013−2016.

ROS production on-demand, paving the way for their use in unique biomedical applications.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b01298. Experimental details, Figure S1 and S2 showing absorption spectra, TEM images and size distribution statistics for the used nanoparticles. Figure S3 shows control measurements for HRP activation by light stimulation of HNPs. Figure S4 shows comparison of HRP activity upon light irradiation between CdS-Au HNPs and CdS NRs in the presence of SOD and hole scavenger. Figure S5 shows control measurements and Table S1 shows simulation parameters for EPR measurements. Figure S6 shows oxygen consumption, ROS production, and HRP activation for nanoparticles coated with GSH coating. Figure S7 shows kinetic measurements of hydrogen production with different biological molecules as hole acceptor agents. Figure S8 shows HRP activation in the presence of serum and control measurements for live/dead assay. (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Fax: (+ 972) 26584148. Author Contributions

N.W. and Y.B.S. contributed equally to this work. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Professor Dorit Ben-Shachar’s lab from the department of psychiatry, B. Rappaport Faculty of Medicine, Technion, Rambam Health Care Campus, Haifa for assistance in the oxygen consumption measurements. The research leading to these results has received funding from The Israel Science Foundation (Grant 1560/13). U.B. thanks the Alfred and Erica Larisch memorial chair. H.S. acknowledges support by The Israel Science Foundation (Grant 817/13) and the Edmond and Lily Safra Center of Brain Science. N.W acknowledges support by a Clara Robert Einstein Scholarship. Y.B.S. acknowledges support by the Ministry of Science, Technology, and Space, Israel and The Camber Scholarship.

■

REFERENCES

(1) Banin, U.; Ben-Shahar, Y.; Vinokurov, K. Chem. Mater. 2014, 26, 97−110. (2) Costi, R.; Saunders, A. E.; Elmalem, E.; Salant, A.; Banin, U. Nano Lett. 2008, 8, 637−641. (3) Mokari, T.; Rothenberg, E.; Popov, I.; Costi, R.; Banin, U. Science 2004, 304, 1787−1790. (4) Adams, D. M.; Brus, L.; Chidsey, C. E. D.; Creager, S.; Creutz, C.; Kagan, C. R.; Kamat, P. V.; Lieberman, M.; Lindsay, S.; Marcus, R. A.; Metzger, R. M.; Michel-Beyerle, M. E.; Miller, J. R.; Newton, M. D.; Rolison, D. R.; Sankey, O.; Schanze, K. S.; Yardley, J.; Zhu, X. Y. J. Phys. Chem. B 2003, 107, 6668−6697. (5) Wu, K. F.; Zhu, H. M.; Lian, T. Q. Acc. Chem. Res. 2015, 48, 851−859. (6) Wilker, M. B.; Shinopoulos, K. E.; Brown, K. A.; Mulder, D. W.; King, P. W.; Dukovic, G. J. Am. Chem. Soc. 2014, 136, 4316−4324. G

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (42) Lees, E. E.; Nguyen, T. L.; Clayton, A. H. A.; Mulvaney, P.; Muir, B. W. ACS Nano 2009, 3, 1121−1128. (43) Rocker, C.; Potzl, M.; Zhang, F.; Parak, W. J.; Nienhaus, G. U. Nat. Nanotechnol. 2009, 4, 577−80. (44) Ellis-Davies, G. C. R. Nat. Methods 2007, 4, 619−628. (45) Mitra, K.; Lippincott-Schwartz, J. Current Protocols in Cell Biology 2001, 46 (4.25), 1−21. (46) Zhu, Z. J.; Yeh, Y. C.; Tang, R.; Yan, B.; Tamayo, J.; Vachet, R. W.; Rotello, V. M. Nat. Chem. 2011, 3, 963−968. (47) Waiskopf, N.; Rotem, R.; Shweky, I.; Yedidya, L.; Soreq, H.; Banin, U. BioNanoScience 2013, 3, 1−11. (48) Pelaz, B.; del Pino, P.; Maffre, P.; Hartmann, R.; Gallego, M.; Rivera-Fernandez, S.; de la Fuente, J. M.; Nienhaus, G. U.; Parak, W. J. ACS Nano 2015, 9, 6996−7008. (49) Oh, E.; Liu, R.; Nel, A.; Gemill, K. B.; Bilal, M.; Cohen, Y.; Medintz, I. L. Nat. Nanotechnol. 2016, 11, 479. (50) McLachlan, S. M.; Rapoport, B. Endocr. Rev. 1992, 13, 192−206.

H

DOI: 10.1021/acs.nanolett.6b01298 Nano Lett. XXXX, XXX, XXX−XXX