Designing Quantum Rods for Optimized Energy Transfer with Firefly

Publication Date (Web): May 23, 2012. Copyright © 2012 American ... Citation data is made available by participants in Crossref's Cited-by Linking se...
0 downloads 0 Views 4MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Designing Quantum Rods for Optimized Energy Transfer with Firefly Luciferase Enzymes Rabeka Alam,† Danielle M. Fontaine,§ Bruce R. Branchini,§ and Mathew M. Maye*,†,‡ †

Department of Chemistry and ‡Syracuse Biomaterials Institute, Syracuse University, Syracuse, New York, 13244 United States Department of Chemistry, Connecticut College, New London, Connecticut, 06320 United States

§

S Supporting Information *

ABSTRACT: The bioluminescence resonance energy transfer (BRET) between firefly luciferase from Photinus pyralis (Ppy) with core/shell semiconductive quantum rods (QRs) has been studied as a function of QR aspect ratio and internal microstructure. The QRs were found to be ideal energy acceptors, and Ppy-to-core distances were optimized using rod-in-rod microstructures that were achieved by the synthetic control of rod morphology, surface chemistry, and Ppy:QR loading. The BRET ratios of >44 measured are the highest efficiencies to date. KEYWORDS: BRET, energy transfer, quantum rods, biotic/abiotic

R

purpose built semiconductive quantum rods (QRs). When bound by firefly luciferase Photinus pyralis (Ppy) using a new direct bioattachment capability, these Ppy-QR BRET nanosystems routinely achieve BR up to 44. We have studied the origins of this much-improved BR by using QR with tailored aspect ratios, donor−acceptor distances, and internal QR microstructures. Quantum dots and rods have been proven to be powerful energy donors in a number of fluorescence resonance energy transfer (FRET) designs.8−10 This is due in large part to a broad absorption profile, size tunable emission, photostability, and long excited-state lifetimes. However these same qualities limit the use of QDs as energy acceptors.11,19 BRET on the other hand utilizes a bioluminescent enzyme in the presence of substrate to generate an excited-state donor and a fluorescent protein,17 dye,18 or QD11−15 as the acceptor. Due to the microsecond lifetimes of bioluminescent processes and the chemical origin of excitation, the QD in particular is an ideal acceptor due to shorter lifetimes and broad absorption in the visible region.11,19 To date, QD-based BRET nanosystems use blue-emitting R. reniformis luciferase (rLuc) expressed with multiple copies of surface exposed lysine residues, which are conjugated to free carboxylates at polymer wrapped QDs.11 BRET was observed for a number of QD colors, including nearinfrared emitters, which have proven to be particularly useful for in vivo imaging.11−14 In the absence of recent advances, there remains an important need to improve BRET efficiency by tailoring of QD morphology and structure to better optimize and understand the BRET nanosystems.

esearch at the nanoscale biotic−abiotic interface centers on endowing an inorganic nanocrystal with the physical, chemical, or energetic properties of biosystems.1,2 This has led to an influx of biomimetic self-assembly,3−5 recognition,6,7 and energy-transfer designs.8−10 One frontier that combines these advances is the ability to harness the energy converting properties of enzymes, by effectively transferring chemical energy from a substrate to an inorganic nanomaterial. This would lead to revolutionary ways of converting chemical energy to light, for example, providing new lighting strategies, in vivo signaling routes, and improved understanding of biosystems. The seminal work in this area has been performed by Rao and co-workers,11−15 who first showed that bioluminescent enzymes mutated from R. reniformis luciferase (Luc8) can be conjugated to semiconductive quantum dots (QDs) via 1-ethyl3-(3-dimethylaminopropyl)-carbodiimide (EDC) coupling.11 In the presence of the substrate colenterazine, the QD accepts the excited-state energy of Luc8 via the nonradiative pathway known as bioluminescence resonance energy transfer (BRET). This allows for the blue-green emission of the enzyme to be converted to the red11−14 and even near-infrared (NIR)15 colors of the QD. Since these reports, BRET nanosystems have been used for in vivo NIR imaging, where it has proven to be particularly useful due to low background signals and the lack of a direct excitation requirement.11,12 The metric for BRET efficiency is the BRET ratio (BR), where the emission of the acceptor and donor are compared.16 To date, typical BR using QD acceptors is limited to only 0.5−2.0. This efficiency is much lower than comparable studies using fluorescent protein (BR = 1−4)17 or molecular fluorophore acceptors (BR = 10− 14).18 The origins of the comparatively low BR with QDs is unknown but is likely due to relatively large spatial distances (r) as well as limited stoichiometry (n). Herein, we have designed a BRET nanosystem that addresses these limitations using © XXXX American Chemical Society

Received: April 4, 2012 Revised: May 16, 2012

A

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 1. (a) The CdSe/CdS or CdSe/CdS/ZnS QRs are first synthesized via organometallic routes in step 1, resulting in trioctylphosphine oxide (TOPO) and alkylphosphonic acid (PA) capped QRs in chloroform. In step 2 the QRs undergo phase transfer and ligand exchange with L-histidine (His), which renders the dots hydrophilic and colloidially stable with a facile His-capping, the excess of which is removed via purification. (b) In step 3 the His-QRs are incubated with PpyGRTS at increasing loading ratios, L = [PPy]:[QR]. Due to the N-terminus 6×His modification of this PpyGRTS, it binds to the QR interface, displacing the His monolayer. (c) In step 4, an excess of the luciferin (LH2) substrate is added to the PpyGRTS-QRs, which immediately induces bioluminescence from PpyGRTS and BRET to the QR. (d) In step 5, the BRET energy transfer is measured, and the BRET efficiency is quantified as the BR. (e) A representative set of TEM micrographs (I−II) for PpyGRTS-QR(675) conjugates, in-which the PpyGRTS layer is visualized using negative staining.

emission spectrum of PpyGRTS is shifted to the blue and is resistant to spectral shifts at high temperature and low pH. The PpyGRTS was directly attached to the QR interface by the Nterminus hexahistidine tag (6×His), which has been shown to coordinate to QDs.23,24 This direct attachment was achievable because the as-synthesized hydrophobic-capped QRs were made hydrophilic and colloidially stable via the use of the histidine-mediated phase-transfer method recently developed in our laboratory (Figure 1a,b, see Supporting Information).24,25 This attachment approach allows for the shortest possible donor−acceptor distance.24 The green emitting PpyGRTS serves as the bioluminescence donor (λD = 547 nm) in the presence of LH2 substrates.18,20,21 The QR acceptors are constructed from core/shell designs, where the inner CdSe core serves as the source for emission (λA = 610−675 nm) that is

With this aim in mind we designed a QD BRET nanosystem based on firefly luciferase from PPy and semiconductive QRs. The native and recombinant Ppy luciferases are robust and emit yellow-green light (λ = 560 nm) at pH = 7.8 in the presence of the substrates firefly (beetle) luciferin (LH2), Mg-ATP, and oxygen. The Ppy-LH2 combination is one of the brightest systems in nature and has one of the highest known quantum yields for luciferases (41 ± 7.4%).21 We chose CdSe/CdS and CdSe/CdS/ZnS QRs as BRET acceptors due to long lifetimes (>20 ns), broad absorption, synthetic control of aspect ratio and microstructure, and increased surface area for tailored donor/acceptor stoichiometry.22 The assembly design and main structural concerns in this biotic−abiotic BRET nanosystem is shown in Figure 1. To provide a robust signal approximately half as intense as the wild-type protein, we have chosen the highly thermostable Ppy variant PpyGRTS.20 The B

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 2. (Top panel) TEM micrographs of (a) CdSe/CdS QR(675) l/w = 3.1 ± 0.5, l = 24.8 ± 3.6, w = 8.2 ± 1.3 nm and (b) CdSe/CdS/ZnS QR(675*) l/w = 3.0 ± 0.5, l = 26.6 ± 3.8, w = 9.1 ± 1.3 nm. (Bottom panel) The BRET emission profile for the respective QR(675) and QR(675*) at L = [Ppy]:[QR] = 5. (Insets) Measured BR dependence on L.

Figure 3. (Top panel) TEM micrographs of (a) CdSe/CdS QR(628) l/w = 8.7 ± 1.6, l = 50.1 ± 5.5, w = 5.9 ± 0.8 nm) and (b) CdSe/CdS/ZnS QR(628*) l/w = 7.4 ± 1.5, l = 52.1 ± 5.8, w = 7.2 ± 1.1 nm. (Bottom panel) The BRET emission profile for the respective QR(628) and QR(628*) at L = [Ppy]:[QR] = 5. (Insets) Measured BR dependence on L.

tailored by novel tuning of the QRs aspect ratios (l/w), which in this study were synthetically modified from 2 to ∼7. Figure 2a summarizes the BRET observed between the PpyGRTS donor and the CdSe/CdS QR acceptor with λA = 675 nm (denoted as QR(675)). The TEM micrograph of the QR(675) reveals an aspect ratio (l/w) of 3.1 ± 0.5 (l = 24.8 ± 3.6, w = 8.2 ± 1.3 nm). The BRET energy-transfer efficiency, BR, was found to be highly susceptible to the molar loading ratio, L = [PpyGRTS]:[QR]. Figure 2a shows typical BRET emission spectra at L = 5, the characteristic feature of which is the low PpyGRTS donor emission and the presence of an intesnse QR emission. Unlike traditional QR photoluminescence spectra, no direct excitation is present in the system, which serves as an attractive internal control that demonstrates energy transfer from PpyGRTS. After spectral deconvolution, a BR of ≈44.2 at L = 5 was calculated based on integration of emission areas. Multiple repeat experiments were performed that resulted in similar efficiencies (±30%). In addition, control experiments in which the PpyGRTS is not assembled at the QR

showed no BRET efficiency. Interestingly, a decrease in BR to 32.0, 24.4, and 17.8 was observed despite an increase in L = 10, 20, 40, respectively (Figure 2a inset). These results suggest that an optimum L exists, possibly because some of the bound PpyGRTS is inactive or linked at distances too great for efficient Förster transfer (see below). These BR values are the highest achieved to date for a BRET nanosystem and comparable to systems using molecular fluorophores with the highest quantum efficiency closest donor−acceptor distance.11−15,18 The resonance energy-transfer efficiency of a donor/acceptor pair is highly sensitive to center-to-center (dipole-to-dipole) distances, r, and scales as 1/r6.19,26 As a morphological control, we increased r by reprocessing the CdSe/CdS QR(675) to deposit a 0.5 nm thick layer of ZnS, creating CdSe/CdS/ZnS QRs (denoted as QR(675*)). As a result, the QR(675*) has a slightly decreased aspect ratio, l/w = 3.0 ± 0.5 (l = 26.6 ± 3.8, w = 9.1 ± 1.3 nm). Interestingly, this additional layer and larger distance have a significant effect on the BRET response (Figure C

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

Figure 4. (Top panel) TEM micrographs of (a) CdSe/CdS QR(613) l/w = 1.8 ± 0.2, l = 9.1 ± 0.9, w = 5.0 ± 0.4 nm) and (b) CdSe/CdS/ZnS QR(613*) l/w = 1.9 ± 0.3, l = 10.3 ± 1.3, w = 5.3 ± 0.5 nm. (Bottom panel) The BRET emission profile for the respective QR(613) and QR(613*) at L = [Ppy]:[QR] = 5. (Insets) Measured BRET ratio (BR) dependence on L.

Figure 5. (a) BRET efficiency plots for PpyGRTS donors and QR acceptors. The summary of the BR measured with respect to aspect ratio at L = 5 and 10 for (b) CdSe/CdS and (c) CdSe/CdS/ZnS QRs. (d) Illustration of the microstructure of the particular QRs studied, including dot-in-dot, rod-in-rod, and dot-in-rod types.

further investigate the BR dependence on aspect ratio, CdSe/ CdS QRs with smaller aspect ratio than both QR(675) and QR(628) were fabricated. Figure 4a shows the TEM micrograph for CdSe/CdS with l/w = 2.1 ± 0.3 (l = 9.9 ± 0.8, w = 4.7 ± 0.4 nm) and λA = 613 nm (denoted as QR(613)). After identical phase transfer and incubation with PpyGRTS as described above, these QR(613) have stable BR of 0.4−0.3 (Figure 4a), with similar values observed for QR(613*) with ZnS shells (Figure 4d). The BRET efficiencies measured in Figures 2−4 are the result of both spectral and morphological characteristics of the PpyGRTS-QR nanosystem. For example, because each QR has a different absorption profile, extinction coefficient, and photoluminescence emission, the resonance energy-transfer parameters will be different. In BRET, the same resonance energy-transfer model is used for that of FRET, were energytransfer efficiency (E) is related to Förster distance (R0) and r by E = R06/R06 + r6.19,26 The BRET efficiency plots for the PpyGRTS donor and the CdSe/CdS or CdSe/CdS/ZnS QR acceptors discussed above are presented in Figure 5a. A key feature of this calculation is the extended BRET efficiency region for QR(675), that due to a high spectral overlap integral

2b). For instance, at L = 5, a 3.5× decrease in BRET was observed (BR ≈ 12.5) compared to the QR(675). Similarly to the QR(675) parent, the system also showed a decreased in BR as L increases. This is likely due to the underlying morphology staying very similar despite the increase in r. The dependence of QR aspect ratio and L on BRET response was next studied using both longer and shorter QRs. For example, Figure 3a shows a TEM micrograph of the longer and narrower CdSe/CdS QR with l/w = 8.7 ± 1.6 (l = 50.1 ± 5.5, w = 5.9 ± 0.8 nm). Due to the thinner diameter, the QR emitted at λA = 628 nm (denoted as QR(628)). This morphology change was achieved by using smaller initial CdSe cores and tuned growth conditions (see Supporting Information). Compared to the QR(675) shown above, a significant decrease in BR was observed for QR(628). For instance, a BR of 1.2, 1.6, and 0.9 was measured at L = 2, 5, and 10 (Figure 3a). Similar to the study above, a ZnS layer was deposited, resulting in an average thickness of ∼0.7 nm (l/w = 7.4 ± 1.5, l = 52.1 ± 5.8, w = 7.2 ± 1.1 nm). This resulted in the decrease in BR to 1.4−0.9 for QR(628*) (Figure 3b). Surprisingly, these longer QRs with the capability to bind more PpyGRTS did not achieve BR as high as the QR(675). To D

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

(JQR(675) = 1.19 × 10−11 M−1 cm3), accounts for an exceptionally large Förster distance of R0 = 9.4 nm (see Table S2, Supporting Information). By comparison, the R0 of 8.2 and 7.3 for QR(628) and QR(613), results in decreased efficiency ranges. The addition of the ZnS layers slightly changes the absorption profile of the QR, increasing R0 due to higher J values (dashed lines, Figure 5a, Table S2, Supporting Information). This is offset however due to the increased r from the ZnS shell, resulting in decreased BR for each. As an example, consider the CdSe/CdS QR(675) and CdSe/CdS/ ZnS QR(675*) shown in Figure 2a,b, respectively. The LH2 binding site at Ppy is located ∼3.5 nm from the (His)6 Nterminus of PPy,27 and since the radius of the QR(675) is ∼4.1 nm, the shortest donor-to-acceptor distance can be approximated as r ∼ 7.6 nm (Figure 5a, inset). At this r, a BRET efficiency of ∼77% is expected for the PpyGRTS and thus the high measured BR. The added 0.5−1.1 nm from the ZnS shell thus increases r, resulting in the loss of BRET efficiency. In addition to spectral considerations, the BRET characteristics observed also indicate that QR morphology plays a critical role in influencing energy transfer. The first way is related to the number of PpyGRTS donors at each QR. For example, at aspect ratio of l/w = 3.1, each QR(675) has a surface area ∼750 nm2, allowing it to accommodate ∼42 PpyGRTS/QR (Ppy = ∼18 nm2 footprint). The longer QR(628) at l/w = 8.7 has a surface area of ∼983 nm2 and can accommodate up to ∼55 PpyGRTS/QR. On the other hand, the QR at low aspect ratio of l/w = 1.8 has surface area of ∼216 nm2 and can only load ∼12 PpyGRTS/QR. Thus, the rods are able to accommodate more Ppy donors, compared to previously reported Ppyfluorophore16−18 and Ppy-QD11−15 examples, in which the acceptor/donor ratios are 3 and 10, respectively. However, as summarized in Figure 5b,c, the BRET efficiency does not necessarily increase with aspect ratio (or loading) but instead shows a considerable maxima with QR(675) at l/w = 3.1. Since each QR can accommodate five PpyGRTS, this suggests that the rod microstructure must also be considered when describing the BRET. It is known that the QR emission originates at the CdSe core and that the core of these QRs is asymmetrically positioned roughly one-third of the rod distance.22,28 With this in mind, the PpyGRTS distance to the core was considered, since the PpyGRTS located furthest from the core, particularly those at r ≫ R0, will not participate in BRET. This is particularly important in the present study, as while all the QRs in the present system have the same composition and general core/shell rod morphology, there are differences in the internal microstructure that are the result of the synthetic parameters. This is illustrated in Figure 5d, in which each QR is shown at scale. For instance, the CdSe/CdS QR(675) with the highest efficiency was synthesized using rodlike CdSe cores (l/w = 2.4 ± 0.3, see Supporting Information), upon which longer core/shell rods were grown. Thus, it is these rod-in-rod morphologies that provide the highest efficiencies. A calculation of the average r distance (rAve) of the PpyGRTS from the core yields ≈9.2 nm, which is close to R0 (9.4 nm). The longest QR(628) with l/w = 8.7, on the other hand, used smaller and more spherical cores (d = 3.9 nm), and thus the rods possess a dot-in-rod morphology. Considering the length of the rod, this results in a rAve ≈ 19.2 nm, well beyond the R0 (8.2 nm). Finally, the QR(613) with l/w = 1.8 has a more traditional dot-in-dot morphology, with the calculated rAve ≈ 8.8 nm being larger than R0 (7.3 nm). An added component not included in this analysis is the potentially unique dipole

originating from the dimensionality of the CdSe core in the rod-in-rod morphology. Recent empirical and theoretical studies have shown that such rod morphology leads to a linear (1D) transition dipole, instead of a spherical point dipole considered in the FRET calculations above.29−34 Such a dipole has been shown to result in a resonance energy-transfer efficiency comparable to 1/r5, thus extending the effective range.30,31,34 This is likely a considerable factor in the improved BRET efficiency for the rod-in-rod systems. In addition, the acceptor quantum yield (QY) may also play a role in BRET efficiency, since in contrast to FRET, acceptor emission is a main factor in the BRET ratio calculation. In these systems, the rod-in-rod samples had QY slightly higher than dot-in-dot and dot-in-rod morphologies (Table S1, Supporting Information). Despite these analyses, it is still unclear as to why BRET efficiency is optimized at L ≈ 5. It may be possible that preferential histag binding occurs at the interface located close to the CdSe core, as those regions have been shown to be defect rich.28 It may also be possible that accessibility to substrates decreases at higher coverage, or cooperative effects result in quenching of bioluminescence. Very recently it was shown that QRs with rod-in-rod morphologies have improved donor properties in FRET studies with fluorophores,29 providing further evidence that these morphologies have importance in energy transfer. In their study the researchers employed careful analysis of PL decay characteristics to probe the number and location of the acceptors that were found to be in restricted geometries. Similar work by us is needed to fully understand the L dependence on BRET efficiency. Our future work will investigate these possibilities and focus on using QR will similar rod-in-rod morphologies of different aspect ratios and spectral characteristics. In summary, we have shown that quantum rods conjugated with the firefly P. pyralis luciferase variant PpyGRTS have high BRET ratios that are dependent upon enzyme loading, rod aspect ratio, and donor-to-acceptor distances. Moreover, optimum loading of ∼5 was demonstrated, with QRs of aspect ratio ∼3 that have rod-in-rod internal microstructure possessing the highest BRET efficiency of 44. Considering that such morphologies allow for a considerable red shift in QR emission, this approach may be useful when optimizing NIR emission in future applications ranging from night vision capabilities to in vivo imaging.

■

ASSOCIATED CONTENT

S Supporting Information *

QD synthesis and Ppy expression procedures, experimental details, BRET constants and calculations, Tables S1−S3, and Figures S1−S9. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M.M.M. acknowledges support from a Department of Defense PECASE award sponsored by the Air Force Office of Scientific Research (FA9550-10-1-0033) and thanks Syracuse University and the Syracuse Biomaterials Institute for generous start-up E

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

Letter

support. B.R.B. acknowledges financial support provided by the National Science Foundation (MCB0842831), the Air Force Office of Scientific Research (FA9550-10-1-0174), and the Hans & Ella McCollum ’21 Vahlteich Endowment.

■

(33) Artemyev, M.; Ustinovich, E.; Nabiev, I. J. Am. Chem. Soc. 2009, 131, 8061−8065. (34) Hernandez-Martinez, P. L.; Govorov, A. O. Phys. Rev. B 2008, 87, 035314.

REFERENCES

(1) Niemeyer, C. M. Angew. Chem., Int. Ed. 2001, 40, 4128−4158. (2) Katz, E.; Willner, I. Angew. Chem. 2004, 43, 6042−6108. (3) Maye, M. M.; Kumara, M. T.; Nykypanchuk, D.; Sherman, W. B.; Gang, V Nat. Nanotechnol. 2010, 5, 116−120. (4) Park, S. Y.; Lytton-Jean, A. K.; Lee, B.; Weigand, S.; Schatz, G. C.; Mirkin, C. A. Nature 2008, 451, 553−556. (5) Zhang, J. P.; Liu, Y.; Ke, Y. G.; Yan, H. Nano Letters 2006, 6, 248−251. (6) Sonnichsen, C.; Reinhard, B. M.; Liphardt, J.; Alivisatos, A. P. Nat. Biotechnol. 2005, 23, 741−745. (7) Chen, C. L.; Zhang, P.; Rosi, N. L. J. Am. Chem. Soc. 2008, 130, 13555−13557. (8) Jennings, T. L.; Singh, M. P.; Strouse, G. F. J. Am. Chem. Soc. 2006, 128, 5462−5467. (9) Lu, H.; Schops, O.; Woggon, U.; Niemeyer, C. M. J. Am. Chem. Soc. 2008, 130, 4815−4827. (10) Xia, Z.; Rao, J. Curr. Opin. Biotechnol. 2009, 20, 37−44. (11) So, M. K.; Xu, C.; Loening, A. M.; Gambhir, S. S.; Rao, J. Nat. Biotechnol. 2006, 24, 339−343. (12) So, M. K.; Loening, A. M.; Gambhir, S. S.; Rao, J. Nature protocols 2006, 1, 1160−1164. (13) Zhang, Y.; So, M. K.; Loening, A. M.; Yao, H.; Gambhir, S. S.; Rao, J. Angew. Chem. 2006, 45, 4936−4940. (14) Xing, Y.; So, M. K.; Koh, A. L.; Sinclair, R.; Rao, J. Biochem. Biophys. Res. Commun. 2008, 372, 388−394. (15) Kim, Y. P.; Daniel, W. L.; Xia, Z.; Xie, H.; Mirkin, C. A.; Rao, J. Chem. Commun. 2010, 46, 76−78. (16) Xu, Y.; Piston, D. W.; Johnson, C. H. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 151−156. (17) Dragulescu-Andrasi, A.; Chan, C. T.; De, A.; Massoud, T. F.; Gambhir, S. S. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 12060−12065. (18) Branchini, B. R.; Ablamsky, D. M.; Rosenberg, J. C. Bioconjugate Chem. 2010, 21, 2023−2030. (19) Clapp, A. R.; Medintz, I. L.; Fisher, B. R.; Anderson, G. P.; Mattoussi, H. J. Am. Chem. Soc. 2005, 127, 1242−1250. (20) Branchini, B. R.; Ablamsky, D. M.; Murtiashaw, M. H.; Uzasci, L.; Fraga, H.; Southworth, T. L Anal. Biochem. 2007, 3261, 253−262. (21) Ando, Y.; Niwa, K.; Yamada, N.; Enomot, T.; Irie, T.; Kubota, H.; Ohmiya, Y.; Akiyama, H Nat. Photonics 2008, 2, 44. (22) Talapin, D. V.; Nelson, J. H.; Shevchenko, E. V.; Aloni, S.; Sadtler, B.; Alivisatos, A. P. Nano Lett. 2007, 7, 2951−2959. (23) Mattoussi, H.; Mauro, J. M.; Goldman, E. R.; Anderson, G. P.; Sundar, V. C.; Mikulec, F. V.; Bawendi, M. G. J. Am. Chem. Soc. 2000, 122, 12142−12150. (24) Han, H. J.; Zylstra, J.; Maye, M. M. Chem. Mater. 2011, 23, 4975−4981. (25) Zylstra, J.; Amey, J.; Miska, N. J.; Pang, L.; Hine, C. R.; Langer, J.; Doyle, R. P.; Maye, M. M. Langmuir 2011, 27, 4371−4379. (26) Clapp, A. R.; Medintz, I. L.; Mattoussi, H. ChemPhysChem 2006, 7, 47−57. (27) Conti, E.; Franks, N. P.; Brick, P. Structure 1996, 4, 287−298. (28) Menagen, G.; Macdonald, J. E.; Shemesh, Y.; Popov, I.; Banin, U. J. Am. Chem. Soc. 2009, 131, 17406−17411. (29) Halivni, S.; Sitt, A.; Hadar, I.; Banin, U. ACS Nano 2012, 6, 2758−2765. (30) Lee, J.; Govorov, A. O.; Kotov, N. A. Nano Lett. 2005, 5, 2063− 2069. (31) Muller, J.; Lupton, J. M.; Logoudakis, P. G.; Schindler, F.; Koeppe, R.; Rogach, A. L.; Feldmann, J.; Talapin, D. V.; Weller, H. Nano Lett. 2005, 5, 2044−2049. (32) Hardzei, M.; Artemyev, M.; Molinari, M.; Troyon, M.; Sukhanova, A.; Nabiev, I. ChemPhysChem 2012, 13, 330−335. F

dx.doi.org/10.1021/nl301291g | Nano Lett. XXXX, XXX, XXX−XXX