Received: 9 November 2017
Accepted: 26 April 2018
DOI: 10.1002/jbio.201700329
FULL ARTICLE
siRNA release from gold nanoparticles by nanosecond pulsed laser irradiation and analysis of the involved temperature increase Florian Rudnitzki1*
| Susanne Feineis2 | Ramtin Rahmanzadeh1 | Elmar Endl3 | Johanna Lutz2 |
Jürgen Groll2 | Gereon Hüttmann1,4 1 Institute of Biomedical Optics, University of Lübeck, Lübeck, Germany 2
Department and Chair of Functional Materials in Medicine and Dentistry and Bavarian Polymer Institute (BPI), University of Würzburg, Würzburg, Germany 3 Institutes of Molecular Medicine and Experimental Immunology, University of Bonn, Bonn, Germany 4 Medizinisches Laserzentrum Lübeck GmbH, Lübeck, Germany
*Correspondence Florian Rudnitzki, Institute of Biomedical Optics, University of Lübeck, Peter-Monnik-Weg 4, 23562 Lübeck, Germany. Email:
[email protected] Funding information Bundesministerium für Bildung und Forschung, Grant/Award Number: 13N11832
Nanosecond pulsed laser irradiation can trigger a release of nucleic acids from gold nanoparticles, but the involved nanoeffects are not fully understood yet. Here we investigate the release of coumarin labeled siRNA from 15 to 30 nm gold particles after nanosecond pulsed laser irradiation. Temperatures in the particle and near the surface were calculated for the different radiant exposures. Upon irradiation with laser pulses of 4 nanosecond duration release started for both particle sizes at a calculated temperature increase of approximately 500 K. Maximum coumarin release was observed for 15 nm particles after irradiation with radiant exposure of 80 mJ cm−2 and with 32 mJ cm−2 for 30 nm particles. This corresponds to a temperature increase of 815 and 900 K, respectively. Our results show that the molecular release by nanosecond pulsed irradiation is based on a different mechanism compared to continuous or femtosecond irradiation. Local temperatures are considerably higher and it is expected that bubble formation plays a crucial role in release and damage to cellular structures. KEYWORDS
cavitation, cell manipulation, controlled release, gold nanoparticle bioconjugates, laser nanoeffects
1 | INTRODUCTION Gold nanoparticles (AuNPs) can be utilized in biomedical applications due to their good biocompatibility [1], special optical properties [2, 3], and diverse possibilities for surface functionalization [4, 5]. One successfully demonstrated application, which is based on these properties, is selective Florian Rudnitzki and Susanne Feineis contributed equally to this study. J. Biophotonics. 2018;e201700329. https://doi.org/10.1002/jbio.201700329
optical transfection with antibody conjugates [6–8]. In addition to targeting molecules, various gene-regulating biologically active molecules, such as drugs [9], DNA [10] and siRNA [11–13], were conjugated to the particle surface to modify specific functions of cells. Laser irradiation makes functionalized AuNPs optically addressable and enables a spatially-temporally controlled release of DNA [14] and siRNA [15–18] into the cytoplasm. In particular, the manipulation of cells with siRNA offers a broad range of possible
www.biophotonics-journal.org
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 of 12
2 of 12
applications based on the knockdown of certain target genes [19]. In principle, any gene expression can be silenced by appropriately designed siRNA interfering with the target mRNA [20]. AuNPs combine high optical absorption cross-sections at around 520 nm with high photostability [2] and allow for an effective transfer of optical energy to heat. Hence, high temperatures are created locally around AuNPs by pulsed laser irradiation [21, 22] and laser heating of AuNPs may cause thermal denaturation, disruption [23] or photochemical modifications of nearby biological structures. In principle, all these physical mechanisms can contribute to the dissociation of the SH-bond between the AuNP and DNA or siRNA. Release of single stranded DNA (ssDNA), which was either linked directly or by a complementary DNA strand, was demonstrated for femtosecond (fs) [24, 25], nanosecond (ns) [16] and continuous (cw) [25] irradiations. Different mechanisms for release were found for fs pulses and cw irradiation. Chemical bond breakage was associated with electrons, which under irradiation with fs pulses become 3 to 4 times hotter than the lattice [26], since energy is supplied within the charge carrier relaxation time by fs irradiation [24, 25]. Under cw irradiation, the melting of the double strand occurs at temperatures below 60! C, when the particle is in thermal equilibrium with the surrounding medium [25]. Long time exposure to temperatures of 80! C can also destroy covalent SH-bonds [27]. In the regime of nanosecond laser pulses, the electrons and lattice atoms of the AuNPs exhibit the same temperatures. However, a strong temperature gradient develops around an irradiated particle and the elevated temperature decreases to the bulk temperature within a distance of several nanometers from the particle surface [28]. Thus, for nanosecond pulsed irradiation, heating of the water is confined to the close vicinity of the AuNPs. Under cw irradiation, heat diffusion causes nonlocal increase of the bulk temperature, which may have unwanted detrimental effects on cell vitality. Release by fspulse requires a complex laser technology and may produce unwanted effects such as the near-field ablation from the AuNP [29] and fragmentation due to coulomb explosion [30] in the sample by the high peak powers. With ns-pulsed irradiation focused in the sample inside a cuvette, SH- bond breakage und dehybridization were observed [16]. Due to the partial irradiation of the sample with a Gaussian beam, this experimental setup was not able to exactly quantify the light dose and to discriminate local from non-local effects of the irradiation. Hence no attempts were made to correlate release with the local temperature and to discuss the impact of the elevated temperature on the siRNA and the surrounding of the particle. Especially high temperature effects, such as particle melting or cavitation, that can significantly impact the cell integrity, were not characterized. However, some possible effects of laser-induced particle melting and fragmentation by a heating-melting-
RUDNITZKI ET AL.
evaporation mechanism were reviewed by Hashimoto et al. [31]. Some of the side-effects need to be controlled to provide intracellular manipulation by molecular release only. In this study, we utilized AuNPs of 2 different sizes, to which fluorescence labeled double-stranded siRNA was bound, to estimate at which temperature increase the SHbond breaks. Sequences of 600 pulses at increasing radiant exposure were applied to the whole sample and release was quantified by changes of the particle absorption and the increase of unbound fluorescing molecules in the solution. The temperature increase was calculated by an analytical thermal model, which accounts for plasmonic absorption, temporal shape of the pulses and thermal conduction from the particles via the surface into the surrounding water. Temperatures at which release occurs, shall pin down the physical processes which are responsible for the release and potential side effects that can limit the intracellular application. 2 | M AT E R I A L A N D M E TH O D S 2.1 | Laser nanoparticle interaction and temperature increase The optical properties of nanoparticles were extensively reviewed by several publications [32–37]. Furthermore, physical effects based on the optical absorption by the nanoparticle and its application, such as cell elimination and manipulation, have been described extensively [6, 7, 28, 38–42]. Scattering of radiation is not relevant for cell targeting or molecular release from the nanoparticle surface, but may be used for visualizing of the particles or for a molecular detection [43–46]. Here, we will focus on the absorption properties of AuNPs and on thermal diffusion, which are both relevant for generating locally high temperatures. In noble metal particles, when smaller than the wavelength of the incident light, the electric field of incident light efficiently couples to quasi-free electrons inducing a charge separation, which collectively oscillates with the external electric field. This plasmon creates an additional electromagnetic field localized on the AuNP, its resonance frequency depends on size, complex dielectric functions of the particle and the surrounding medium. The plasmon causes an extraordinary high absorption and scattering off the AuNPs. For a spherical AuNP, the absorption was calculated by the Mie theory [34, 47] (cf. Eqs. (S1-S4) in File S1, Supporting Information), utilizing the dielectric function of gold measured by Johnson and Christy [48]. Absorbed energy of the pulsed laser irradiation causes a temperature increase in AuNPs, which is determined also by heat conduction. The exact calculation of the spatial and temporal temperature distribution within and around a spherical symmetric AuNP with radius Rp is based on 2 coupled differential equations, which describe the heat flow from the
3 of 12
RUDNITZKI ET AL.
particle into the surrounding water [49–51]. From the solution for a Dirac-shaped irradiation (Green’s function), spatial and temporal temperature profiles were calculated by integrating over the temporal shape of the laser pulse. Thermal properties were assumed to stay constant (cf. Eqs. (S5-S14) in File S1). Figure 1A shows the temporal temperature increase inside a 15 nm sized AuNP and Figure 1B for both, a 15 and 30 nm sized AuNP at specified locations with different distances from the nanoparticle surface for a 4 nanosecond at full width half maximum (FWHM) Gaussian laser pulse. The resulting temperature increase within the AuNP is constant due to the high thermal diffusivity of gold, while in the surrounding water temperature decreases nearly exponentially with distance to the AuNP surface. The finite thermal conductivity of the interface between the AuNP and the aqueous medium causes a discontinuity of temperature across the nanoparticle surface, which is located at r = Rp. An interface conductance of 105 MW m−2 K−1 was used in the calculations [51]. In general, the nanoparticle size, the pulse duration and the radiant exposure determine the maximum temperature increase within an irradiated AuNP and in the surrounding water. In the thermal model used here, the temperature of the AuNP and in the surrounding water scales linearly with the radiant exposure. This holds approximately until a cavitation bubble formed around the particle, which changes optical absorption and heat conduction [28] or the particle starts to melt. Outside this linear regime of constant optical and thermal properties mathematical modeling of the temperature evolution inside the nanoparticle and the surrounding medium becomes very difficult and has not been done so far. When a cavitation bubble is formed the absorption of the nanoparticle decreases. Calculations show at a wavelength
of 532 nm, the absorption cross section of a 30 nm AuNP decreases from 1.4 × 10−15 m2 to 0.30 × 10−15 m2, while scattering cross-section increases, due to the formed vapor (cf. Eqs. (S1-S4) in File S1, refractive index of 1.0 was assumed for the surrounding of the AuNP, when a bubble formed). Also, the heat conduction from the particle to the surrounding water breaks during the formation of the vapor bubble. Assuming adiabatic conditions, the energy deposition within the AuNP ∆Ep can be calculated from the integral over the irradiance H of the laser pulse as determined by Rudnitzki et al. [28]. ðt ∆Ep ðt Þ = σ abs ċH ðt 0 Þdt0 : ð1Þ tv
Under nanosecond pulsed irradiation, a nanoparticle passes through different heating phases. The initial phase is governed by thermal diffusion and high absorption in the presence of water around the particle, which causes a rapid temperature increase of the whole nanoparticle. During second phase, bubble formation thermally isolates the particle and changes local index of refraction at the surface of the particle, which decreases absorption. Absorption loss caused by surrounding vapor remains transient, while melting and subsequent reshaping of the nanoparticle causes permanent absorption loss [28]. During third phase, low energy deposition and still a temperature increase can be expected despite the strongly decreased absorption. In case the vaporization enthalpy is provided a loss of nanoparticle material will occur. The initial phase can be clearly distinguished from the subsequent phases by sudden bubble formation. In the course of energy density of the nanoparticle over the laser pulse time, the event of bubble formation is clearly marked by a discontinuity. Calculation of the temperature becomes extremely complex since cooling of the particle depends on
(A) Temperature increase calculated for a 15 nm AuNP (blue curve) and at different distances (blue to red curves) in the surrounding medium after irradiation with a Gaussian shaped laser pulse of 4 nanosecond FWHM duration. A wavelength of 532 nm was assumed. A radiant exposure of 69.4 mJ cm−2 leads to a maximum temperature increase of around 700 K within the particle. (B) Calculated temperature for a 15 nm (blue line) and a 30 nm AuNP (yellow line) as a function of the radial distance from the nanoparticle center at the time point of maximum nanoparticle temperature increase (vertical line in A). Due to the finite thermal conductivity across the AuNP-water interface a discontinuity with a temperature drop of 35% (15 nm particle) and 25% (30 nm particle) occurs FIGURE 1
4 of 12
RUDNITZKI ET AL.
(A) Calculated temperature increase inside a 30 nm AuNP (black line) and of the water on the particle surface (red line), when irradiated with 25.2 mJ cm−2 at a wavelength of 532 nm. Assumed pulse width was 4 nanosecond. If the change of optical and thermal conditions after bubble formation is neglected temperature would follow the dashed lines. (B) Energy density within the 30 nm AuNP calculated with Eq. (1). Three heating phases (I-III) can be distinguished. (I) initial phase of heating the AuNP until spinodal temperature is reached in the water for r = Rp. The spinodal temperature is assumed to be the minimum temperature for bubble formation. We assumed that water evaporates in 100 nm layer. Energy providing the latent heat of evaporation is comparably small and not visible on this scale. (II) Energy density grows during particle temperature increase to 1336 K, while bubble formation causes thermal isolation and decreased absorption. The particle melts at constant particle temperature. (III) Energy deposition in a liquid gold particle. Further calculations for AuNP with different diameters shown in Figure S2 in File S1 FIGURE 2
gas density and size of the vapor bubble. In contrast to fs and picosecond (ps) pulses, cooling has an immense influence on the temperature for ns heating time. The temperature of AuNPs with diameter up to 30 nm closely follows the temporal shape of the laser pulse. Maximum temperature is reached after maximum laser pulse irradiance at a time delay of 285 ps for 15 nm AuNP and 585 ps for 30 nm AuNP. For larger particles, the thermal inertia causes an increasing temporal delay between irradiance and particle temperature. After reaching maximum temperature, the nanoparticles cool down to nearly room temperature in less than 100 nanosecond. Thus, irradiation with a pulse sequence at a repetition rate of a few tens of a Hertz can be considered as single pulsed irradiation. Mutual heating of the particles is avoided at AuNP concentration used in the experiments of around 1012 AuNP/mL, which corresponds to an inter-particle distance of around 600 nm. 2.2 | siRNA functionalization of AuNPs In order to ensure RNase-free conditions, all following solutions were treated with 0.1% diethyl pyrocarbonate (DEPC). The 15 nm AuNPs with a concentration of 1.4 × 1012 particles mL−1 and 30 nm AuNPs with a concentration of 2.0 × 1011 particles mL−1, both citrate-modified (BBI, Cardiff, UK), were incubated with a freshly reduced fluorescent GAPDH-siRNA (GAPDH: Glyceraldehyde 3-phosphate dehydrogenase; BioSpring GmbH, Frankfurt am Main, Germany) solution in water (14 μL of an 18 μM or 14 μL of a 27 μM solution, respectively). Within 1 hour, a phosphate buffer (50 mM, pH 7.5) was slowly added to the colloidal solution, followed by ultrasound treatment for 10 seconds. After incubation at room temperature for 12 hours under the exclusion of light, AuNPs were purified by 2 successive
centrifugations (22 000g, 30 minutes, 4! C) and resuspension cycles in 800 μL phosphate buffer (5 mM, pH 7.5). To probe the breakage of the thiol bond, the coumarin molecule was conjugated to the sense strand as illustrated in Figure 3. Dynamic light scattering (DLS) measurements showed no aggregation of the conjugates after synthesis (cf. Figure S1 in File S1). 2.3 | Irradiation of the samples Samples were irradiated with a frequency-doubled, Qswitched Nd:YAG Laser (Surelite I-20, Continuum, Santa Clara, California), which generates pulses of a duration of 4 nanosecond at FWHM with a wavelength of 532 nm, at which the irradiated AuNP-siRNA conjugates exhibit around 90% of their maximum absorption. Figure 3 schematically illustrates the optical arrangement for pulsed irradiation of the samples. The pulse energy was adjusted by combination of a broadband λ/2 plate and a polarizing beam splitter. The diameter of the irradiation spot in the sample plane was adjusted by a lens to allow for an irradiation of the whole sample, which was placed in 384-well plates (Quartz well plate, well diameter 3.4 mm). In order to determine the single pulse radiant exposure, the beam radius, where intensity falls to 1/e of the peak intensity in the beam center, was determined with the help of the knife edge method and the pulse energy was measured with a pyroelectric detector
Schematic illustration of the siRNA-AuNP conjugates used to demonstrate the targeted release by pulsed laser irradiation
FIGURE 3
5 of 12
RUDNITZKI ET AL.
Optical arrangement for pulsed irradiation of the samples
FIGURE 4
(PE25-C, Ophir). Independent of stirring and Brownian motion, all particles were irradiated with exactly 600 pulses at a repetition rate of 10 Hz for 60 seconds. Different radiant exposure up to 250 mJ cm−2 were applied. 2.4 | Detection of released siRNA In this study, we concentrated on absorption measurements of the AuNPs and fluorescence measurements of the released dye to bring the theoretical temperature calculations in line with the experimental observations. Particle extinction was measured before and after irradiation with a UV-Vis spectrometer (U-2900, Hitachi, Tokyo, Japan) to detect changes of particle extinction due to changes of the local index of refraction, when siRNA is released, or due to size changes and aggregation of the particles. The measurement of refractive index change by shift of the absorption peak is expected to small and therefore unsuitable to quantify the molecular release. Thus, the amount of siRNA released by pulsed laser irradiation was also determined by the increase of fluorescence (Fluoromax, Spex, Horiba, Kyoto, Japan) in the supernatant after centrifugation of the sample with 21 100 g at 4 ! C for 20 min (Heraeus Fresco 21, Thermo Fisher Scientific, Waltham, Massachusetts). To verify the optically induced release, the remaining pellet was redispersed in a dithiothreitol (DTT, SigmaAldrich, St. Louis, Missouri) for 20 hours, which releases the remaining siRNA, and the fluorescence in the supernatant was measured again after a further centrifugation. Nonirradiated conjugates served as negative control. For quantification of the fluorescence after full release and as a positive control, non-irradiated samples were incubated in DTT for 20 hours. Figure 5 illustrates schematically the controls and proof of release. The relative increase of the fluorescence of released siRNA was calculated from the integral of the measured fluorescence over the wavelength range between 420 and 600 nm divided by the integral of the fluorescence to the
positive control. Assuming that the dye molecule is not separated from the siRNA or destroyed by the irradiation, the relative fluorescence intensity should be proportional to the amount of released siRNA. 3 | R E S U L T S AN D D IS C US S IO N 3.1 | Change of the nanoparticle absorption To study the molecular release of siRNA, extinction spectra were measured after irradiation at varying radiant exposure (Figure 6). At both particle sizes, a blue shift of up to 4 nm was observed. For the conjugates with 30 nm AuNPs, the peak position of the extinction spectra started to change for irradiations above 10 mJ cm−2, whereas the 15 nm conjugate required irradiation above 50 mJ cm−2 (Figure 7A). Relating the extinction changes to the calculated increase of particle temperature, the course of the blue shift is the same for both particle sizes (Figure 7B). In first approximation, the position of the absorption peak is proportional to the index of refraction of the medium around the particles. The observed shift indicates a decrease of the local index of refraction, which is caused by a release of siRNA [24]. Most of the absorption shift occurs at particle temperatures below 1300 K. At higher particle temperature, the particle extinction at a wavelength above 600 nm increases significantly (Figure 8). This effect is caused by an increase of the effective particle size. It was observed previously under ns laser irradiation of AuNP nucleic acid conjugates [16] and was related to aggregation, where the electron cloud is delocalized over all adjacent AuNPs with a multitude of plasmon modes resulting in a broadening and damping of the SPR. This can be ascribed to a loss of the surface stabilization, since at the calculated particle temperatures of 1300 K most of the siRNA will be released from the AuNP surface, which subsequently leads to aggregation. Relating the absorption peak shift and the absorption increase in the range above 600 nm to calculated particle
6 of 12
RUDNITZKI ET AL.
Measurement of siRNA release by fluorescence measurements of the supernatant after centrifugation. (A) Non-irradiated conjugates exhibit low concentration of fluorescence dye labeled siRNA in the supernatant after centrifugation and served as negative control. (B) Addition of DTT leads to a complete release of the siRNA. Full release of the siRNA by DTT was used for positive control and for reference fluorescence measurements. Fluorescence was measured in the supernatant of samples after irradiation with different radiant exposure
FIGURE 5
temperatures corrects of the different absorption cross sections, 0.16 × 10−15 m2, for 15 nm diameter and 1.40 × 10−15 m2 for 30 nm. Also the differences in thermal conduction due to the particle size are taken into account. The good correspondence for the 2 different particle sizes suggests that particle temperature has large impact on the release of siRNA. However, changing absorption and heat conduction make the calculated temperatures after phase changes of water and gold completely unrealistic. The temperatures given here are only extrapolations from a regime, were all optical and thermal parameters remain constant. Further, TEM images were investigated for shape changes of the particles before and after irradiation (Figure 9).
Significant reshaping of the particles shows for irradiation with radiant exposure of 23 mJ cm−2. For irradiation above 50 mJ cm−2 fragmentation appears. Coverage of the particle with siRNA was not visible in our hand under TE microscopy. The siRNA has the size of only 19 to 20 base pairs. Visualization of the corona around the particle by an increased inter-particle distances or negative staining needs a better resolution of the microscope and more sophisticated preparation the samples than visualizing antibodies or larger proteins on the particle surface. Absorption measurements of non-functionalized 30 nm AuNP also showed increasing red shift of the peak wavelength (Figure 10, red dots), due to surface melting of the
RUDNITZKI ET AL.
7 of 12
Absorption spectra of (A) 15 nm and (B) 30 nm gold conjugates after irradiation at different radiant exposures. Spectra were normalized to their absorption at 400 nm. The absorption band peak of irradiated conjugates exhibits a shift to shorter wavelengths. At higher radiant exposures, a decrease of the absorption at 520 nm and an increase in the range above 600 nm was observed, which indicates particle aggregation
FIGURE 6
AuNP and reshaping. The shift is not as prominent as the one for siRNA release; however, it cannot be clearly differentiated from the absorption shift due to reshaping of the particles. The change in absorption spectra and TEM imaging showed that high temperatures are involved that can lead to melting and furthermore to cavitation bubble formation, at the light doses, at which siRNA release was observed. Increase of coumarin fluorescence. After irradiation of the coumarin-siRNA AuNPs conjugates fluorescence increased in the supernatant for both 15 and 30 nm particles (Figure 11). In correspondence to the blue shift of the nanoparticle extinction, the conjugates with 15 nm AuNP exhibit an increase of fluorescence in the supernatant to a maximum of 20% of the positive control, when irradiated with 80 mJ cm−2 (Figure 11A). A slightly higher increase to 30% compared to the positive control was achieved for 30 nm AuNP at a radiant exposure of 32 mJ cm−2 (Figure 11B). This corresponds to the irradiation parameter, at which the blue shift of the extinction reached
its maximum. With the increase of fluorescence in the supernatant, fluorescence of the bound siRNA decreased as expected (see Figure S5 in File S1 for exemplary fluorescence spectra measured in the supernatant of irradiated 30 nm AuNP). Above 80 mJ cm−2, respectively, 32 mJ cm−2 fluorescence in the supernatant decreased whereas the amount of fluorescence of bound siRNA did not increase. Obviously, coumarin is destroyed during the irradiation. Photobleaching by absorption of the ns pulses can be excluded as a reason for the loss of total coumarin fluorescence. When irradiating coumarin labeled siRNA in aqueous solution, the fluorescence remains stable under radiant exposure with up to 100 mJ cm−2 (Figure 12A). In contrast, the Cy5-siRNA, which served as a positive control for photobleaching, loses its ability to emit fluorescence in a linear dose-dependent manner. The photochemical destruction can be explained by the residual absorption of Cy5 at the irradiation wavelength, which is not present for coumarin (Figure 12B). The
Blue shift of the peak extinction wavelength occurs for both particle sizes at different radiant exposure (A), but at similar calculated particle temperature (B)
FIGURE 7
8 of 12
RUDNITZKI ET AL.
(A) Increase of particle extinction at 650 nm after irradiation with different radiant exposure. (B) In both particles, increase of absorption is observed at calculated temperatures above 1300 K
FIGURE 8
plasmon resonance absorption of an AuNP is spectrally located in a wavelength range above from the emitted peak fluorescence of coumarin.
Reduction of total coumarin can be related to the calculated maximum temperature increase above 1300 K within the AuNP and 1000 K in a surrounding water layer on the AuNP surface.
TEM images of 30 nm AuNP siRNA conjugates before irradiation (A) and after irradiation with radiant exposure of 5 mJ cm−2 (B), 10 mJ cm−2 (C), 23 mJ cm−2 (D), 50 mJ cm−2 (E) and 100 mJ cm−2 (F) FIGURE 9
9 of 12
RUDNITZKI ET AL.
3.2 | Mechanism of siRNA release
The absolute shift of the absorption peak to shorter wavelength increased with the radiant exposure of the irradiation. Data represented by blue dots correspond to the TEM images shown in Figure 9. Irradiation of non-functionalized 30 nm AuNP (red dots) leads also to a shift of the peak wavelength. Corresponding TEM images and absorption spectra of 30 nm AuNP without siRNA functionalization are shown in Figures S3 and S4 in File S1
FIGURE 10
Three mechanisms may be responsible for the molecular release of covalently linked siRNA from the nanoparticle surface: (1) cleavage of the thiol-bonding by hot electrons; (2) thermal dissociation; and (3) the effect of the bubble formation. Goodman [25] and Jain [24] proposed that fs pulsed laser excitation causes a hot-electron transfer, which leads to breaking of the Au−S bond with low local heating. Our experimental results show that irradiation with ns laser pulses releases siRNA at particle temperatures of at least 500! C, which is just above the expected threshold for bubble formation. Cold bond breakage is obviously not possible under ns irradiation. Since thermalization of the electronic system and subsequently of the atomic lattice take place on the ps scale. Under our irradiation conditions, electrons have the same temperature as the gold lattice. This is in contrast with fs irradiation, where directly after irradiation the electronic system, which has a lower heat capacity than the
Fluorescence in the supernatant after irradiation of conjugates with 15 and 30 nm AuNP shows a release of siRNA due to thiol-bond breaking. (A) For 15-nm AuNP, the measured fluorescence of released siRNA increased to a maximum of 20% after irradiation with 80 mJ cm−2 (blue). Up to this radiant exposure, the fluorescence of bound siRNA behaves complementary (yellow). For higher radiant exposures, fluorescence of both fractions decreases. (B) Relative fluorescence of released (blue) and bound (yellow) siRNA for irradiated conjugates with 30 nm AuNP. A maximum of around 30% is reached after irradiation with 32 mJ cm−2. Above this value, fluorescence of both siRNA fractions decreases as for 15 nm particles. (C) Plotted against the calculated temperature, the same behavior is observed for both particle sizes
FIGURE 11
10 of 12
RUDNITZKI ET AL.
(A) Fluorescence signal of siRNA-coumarin and siRNA-Cy5 conjugates measured after 4 nanosecond pulsed irradiation for 60 seconds. With increasing irradiation Cy5 fluorescence is bleached in linear dose-dependent manner, while fluorescence of the coumarin fluorescence remains stable. (B) Absorption and fluorescence spectra of coumarin, cy5 and 30 nm AuNP
FIGURE 12
lattice, is 5 times hotter than the particle and its surroundings [26]. Li et al. showed, in a study of the thermal stability of DNA functionalized 20 nm AuNPs, that the Au-thiol bonding becomes instable upon temperature increase lower than 100! C. By measuring the fluorescence of labeled DNA cleavage of the AuNP-thiol bonding was observed for temperatures up to 85! C within hours [27]. In contrast to these experiments with homogeneous heating of the whole solution, ns pulsed laser irradiation induce considerably higher temperatures for a significantly shorter time. The possibility of a thermally activated bond breaking within the few ns of increased temperature can be estimated by Arrhenius relation of temperature and reaction rate. When extrapolating the Arrhenius data given by Li et al., a reaction rate of 11.5 s−1 is found for a nanoparticle temperature of 500! C. Even at a temperature of 2000! C, the rate will increase only to 2.5 × 104 s−1. Assuming 5 nanosecond increased temperature in 600 pulses gives a total thermal exposure of 3 μs, which is too short to release at this rate significant amounts of covalently bound siRNA
from the particles. Given the validity of the Arrhenius extrapolation, siRNA release must be caused by a nonthermal effect. Onset of siRNA release occurs at irradiation with radiant exposure of 55 mJ cm−2 for 15 nm AuNP conjugates and at 16 mJ cm−2 for 30 nm AuNP conjugates. Here, temperatures inside the 15 and 30 nm AuNPs increase to a maximum of 580! C and 469! C, respectively. The particles heat a water shell near their surface to approximately 305! C, which is near of the spinodal point of water. At the irradiance of maximal fluorescence in the supernatant (80 and 32 mJ cm−2), a temperature in the liquid near the particle of 437! C (15 nm) and 584! C (30 nm) is expected, which exceeds the critical temperature. Due to the overheating of the liquid within a thin layer on the AuNP surface, phase transition and a rapid expansion of the vapor will follow [52, 53]. Experimentally, an explosive vaporizing of liquid around laser irradiated AuNPs was shown for fs pulsed laser irradiation at temperatures of 302! C and 452! C for 36 and 100 nm AuNP, respectively [54].
Schematic illustration of temperature increase and bubble formation involved in the release of molecules from the nanoparticle surface. Irradiation of the AuNPs functionalized by dye labeled siRNA leads to a strong temperature increase inside the particles and in the surrounding medium. Due to thermal dissociation of the thiol-bondings, the molecules detach from the AuNP. Further, upon sufficient temperature increase a vapor bubble can form rapidly and expand, which carries the molecules further into the surrounding
FIGURE 13
11 of 12
RUDNITZKI ET AL.
Therefore, both increased temperature and bubble formation may be responsible for the releases of the siRNA as schematically illustrated in Figure 13. So far, the contribution of shear and tensile forces at the vapor-liquid interface from an expanding cavitation bubble to dissociation of the Au-thiol bond or the dehybridization of the double-stranded siRNA is unclear and difficult to predict. The effort to release higher amounts of siRNA caused coumarin and possibly the siRNA as well to lose its integrity. Besides this, for targeted gene silencing, the cavitation may cause undesired effects, such as increased cell elimination. Denaturation of proteins coupled to nanoparticles at even larger distances from the particle than the siRNA was demonstrated [55–57] and was related to cavitation, not thermal denaturation. It is conceivable that cavitation limits an application of the laser induced molecular release from AuNPs, at least when it is performed after cell uptake or very close to the cell membrane. The data for experiments with fs laser pulses published by Halas et al. [25] exhibit a decrease in the detected DNA release when certain pulse energy was exceeded. Also a color change of the irradiated sample of gold nanoparticle and a loss of nanoshell integrity were observed, which could also be attributed to cavitation. 4 | CON CLU SION In this study, we showed siRNA release from gold particles after ns pulsed laser irradiation due to Au-S bond breakage. Our study demonstrates that different radiant exposures are needed, when 15 or 30 nm particles were used. Maximum fluorescence of released siRNA was observed at 80 mJ cm−2 for 15 nm particles and at 32 mJ cm−2 for 30 nm particles. With these distinct radiant exposures for release, it should be possible to address different molecules from different sized particles within the same biological sample. The detection of the release by measuring the fluorescence of the dye used to label the siRNA, revealed a destruction of the dye molecule itself. Our calculations with an analytical model show that, for 15 and 30 nm AuNPs, the temperature increase is sufficient to cause cavitation. It is conceivable that high temperatures and the predicted bubble formation are responsible for this effect, as photobleaching was ruled out. Furthermore, our results show that ns pulsed laser irradiation may be used for drug release from AuNPs. However, irradiation parameters need to be controlled to avoid nontargeted effects, such as elimination through intracellular induced cavitation. Also, modifications of the conjugates to provide stability in the cell environment can affect the irradiation parameters, which provide molecular release but not cell elimination. When targeting extracellular matrix in the vicinity of cells, vapor bubble formation may be utilized for
transient permeabilization of the targeted cell. Further studies have to show, if side effects, such as cell elimination, can be avoided more effectively, while providing the desired cellular manipulation by simultaneously released siRNA inside living cells.
ACKNOWLEDGMENTS
This work was supported by the German ministry of Education and Research (grant no. 13N11832). Conflict of interest The author reports no conflicts of interest in this work.
AUTHOR BIOGR APHI ES
Please see Supporting Information online. OR CID Florian Rudnitzki
http://orcid.org/0000-0001-8363-425X
REFERENCES [1] D. A. Giljohann, D. S. Seferos, W. L. Daniel, M. D. Massich, P. C. Patel, C. A. Mirkin, Angew. Chem. (Int. Ed. Engl.) 2010, 4 9 , 3280. [2] V. Amendola, R. Pilot, M. Frasconi, O. M. Marago, M. A. Lati, J. Phys. 2017, 29 , 203002. [3] S. Eustis, M. A. El-Sayed, Chem. Soc. Rev. 2006 , 3 5, 209. [4] W. Gao, C.-M. J. Hu, R. H. Fang, B. T. Luk, J. Su, L. Zhang, Adv. Mater. 2013, 25, 3549. [5] C. Vericat, M. E. Vela, G. Corthey, E. Pensa, E. Cortes, M. H. Fonticelli, F. Ibanez, G. E. Benitez, P. Carro, R. C. Salvarezza, RSC Adv. 2014, 4 , 27730. [6] C. Yao, R. Rahmanzadeh, E. Endl, Z. Zhang, J. Gerdes, G. Hüttmann, J. Biomed. Opt. 2005, 10, 064012. [7] C. Yao, X. Qu, Z. Zhang, G. Hüttmann, R. Rahmanzadeh, J. Biomed. Opt. 2009, 14 , 054034. [8] C. Yao, F. Rudnitzki, G. Hüttmann, Z. Zhang, R. Rahmanzadeh, Int. J. Nanomedicine 2017, 12, 5659. [9] J. D. Gibson, B. P. Khanal, E. R. Zubarev, J. Am. Chem. Soc. 2007, 129 , 11653. [10] N. L. Rosi, D. A. Giljohann, C. S. Thaxton, A. K. R. Lytton-Jean, M. S. Han, C. A. Mirkin, Science 2006 , 3 12, 1027. [11] A. C. Bonoiu, S. D. Mahajan, H. Ding, I. Roy, K.-T. Yong, R. Kumar, R. Hu, E. J. Bergey, S. A. Schwartz, P. N. Prasad, Proc. Natl. Acad. Sci. U.S.A. 2009, 106 , 5546. [12] S. Guo, Y. Huang, Q. Jiang, Y. Sun, L. Deng, Z. Liang, Q. Du, J. Xing, Y. Zhao, P. C. Wang, A. Dong, X.-J. Liang, ACS Nano 2010, 4 , 5505. [13] D. A. Giljohann, D. S. Seferos, A. E. Prigodich, P. C. Patel, C. A. Mirkin, J. Am. Chem. Soc. 2009, 13 1, 2072. [14] A. Wijaya, S. B. Schaffer, I. G. Pallares, K. Hamad-Schifferli, ACS Nano 2009, 3 , 80. [15] R. Huschka, O. Neumann, A. Barhoumi, N. J. Halas, Nano Lett. 2010, 10, 4117. [16] L. Poon, W. Zandberg, D. Hsiao, Z. Erno, D. Sen, B. D. Gates, N. R. Branda, ACS Nano 2010, 4 , 6395. [17] S. E. Lee, D. Y. Sasaki, T. D. Perroud, D. Yoo, K. D. Patel, L. P. Lee, J. Am. Chem. Soc. 2009, 13 1, 14066. [18] S. E. Lee, G. L. Liu, F. Kim, L. P. Lee, Nano Lett. 2009, 9 , 562-570. [19] L. J. Scherer, J. J. Rossi, Nat. Biotech. 2003, 21, 1457. [20] M. A. Minks, D. K. West, S. Benvin, C. Baglioni, J. Biol. Chem. 1979, 254 , 10180.
12 of 12
[21] G. Hüttmann, B. Radt, J. Serbin, B. I. Lange, R. Birngruber, Med. Laser Appl. 2002, 17, 9. [22] G. Hüttmann,R. Birnugruber, IEEE J. Sel. Top. Quantum Electron. 1999, 5, 954-962. [23] C. M. Pitsillides, E. K. Joe, X. Wei, R. R. Anderson, C. P. Lin, Biophys. J. 2003, 84 , 4023. [24] P. K. Jain, W. Qian, M. A. El-Sayed, J. Am. Chem. Soc. 2006 , 128, 2426. [25] A. M. Goodman, N. J. Hogan, S. Gottheim, C. Li, S. E. Clare, N. J. Halas, ACS Nano. 2017, 11, 171. [26] M. Perner, Optische Untersuchung der Elektronen- und Gitterdynamik in Edelmetall-Nanopartikeln, Shaker, Herzogenrath, Germany 1999. [27] F. Li, H. Zhang, B. Dever, X.-F. Li, X. C. Le, Bioconjug. Chem. 2013, 24 , 1790. [28] F. Rudnitzki, M. Bever, R. Rahmanzadeh, K. Brieger, E. Endl, J. Groll, G. Hüttmann, J. Biomed. Opt. 2012, 17, 05800313. [29] A. Plech, V. Kotaidis, M. Lorenc, J. Boneberg, Nat. Phys. 2006 , 2, 44. [30] D. Werner, A. Furube, T. Okamoto, S. Hashimoto, J. Phys. Chem. C. 2011, 115, 8503. [31] S. Hashimoto, D. Werner, T. Uwada, J. Photochem. Photobiol. C. 2012, 13 , 28. [32] H. C. van de Hulst, Light Scattering by Small Particles, John Wiley & Sons Ltd, New York, NY 1958. [33] C. F. Bohren, D. R. Huffman, Absorption and Scattering of Light by Small Particles, Wiley, New York, NY 1983. [34] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer, Berlin/New York 1995. [35] S. Link, M. A. El-Sayedink, Ann. Rev. Phys. Chem. 2003, 54 , 331. [36] E. Hutter, J. H. Fendler, Adv. Mater. 2004, 16 , 1685. [37] P. Mulvaney, Langmuir 1996 , 12, 788. [38] J. Neumann, R. Brinkmann, BIOMEDO 2006 , 11, 041112. [39] E. Bergeron, C. Boutopoulos, R. Martel, A. Torres, C. Rodriguez, J. Niskanen, J.-J. Lebrun, F. M. Winnik, P. Sapieha, M. Meunier, Nanoscale 2015, 7, 17836. [40] M. Schomaker, D. Heinemann, S. Kalies, S. Willenbrock, S. Wagner, I. Nolte, T. Ripken, H. M. Escobar, H. Meyer, A. Heisterkamp, J. Nanobiotechnol. 2015, 13 , 10. [41] X. Huang, S. Neretina, M. A. El-Sayed, Adv. Mater. 2009, 21, 4880. [42] E. Y. Hleb, J. H. Hafner, J. N. Myers, E. Y. Hanna, B. C. Rostro, S. A. Zhdanok, D. O. Lapotko, Nanomedicine (Lond.) 2008, 3 , 647. [43] C. Sönnichsen, S. Geier, N. E. Hecker, G. von Plessen, J. Feldmann, H. Ditlbacher, B. Lamprecht, J. R. Krenn, F. R. Aussenegg, V. Z. H. Chan, J. P. Spatz, M. Möller, Appl. Phys. Lett. 2000, 77, 2949. [44] G. Raschke, Dr. rer. nat. Dissertation, Ludwig-Maximilians-Universität München (Fakultät für Physik, München, Bayern, Germany) 2005. [45] J. S. Becker, M. Zoriy, A. Matusch, B. Wu, D. Salber, C. Palm, J. S. Becker, Mass Spectrom. Rev. 2010, 29 , 156. [46] I. Ament, J. Prasad, A. Henkel, S. Schmachtel, C. Sönnichsen, Nano Lett. 2012, 12, 1092. [47] G. Mie, Ann. Phys. 1908, 3 3 0, 377. [48] P. B. Johnson, R. W. Christy, Phys. Rev. B. 1972, 6 , 4370. [49] H. S. Carslaw, J. C. Jaeger, Conduction of Heat in Solids, Clarendon Press, Wotton-under-Edge 1959. [50] A. Plech, S. Grésillon, G. von Plessen, K. Scheidt, G. Naylor, Chem Phys. 2004, 29 9 , 183. [51] A. Plech, V. Kotaidis, S. Grésillon, C. Dahmen, G. von Plessen, Phys. Rev. B. 2004, 70, 195423. [52] E. Lukianova-Hleb, Y. Hu, L. Latterini, L. Tarpani, S. Lee, R. A. Drezek, J. H. Hafner, D. O. Lapotko, ACS Nano 2010, 4 , 2109. [53] A. Vogel, V. Venugopalan, Chem. Rev. 2003, 103 , 577. [54] V. Kotaidis, C. Dahmen, G. von Plessen, F. Springer, A. Plech, J. Chem. Phys. 2006 , 124 , 184702.
RUDNITZKI ET AL.
[55] B. Radt, Dr. rer. nat. Dissertation, University of Lübeck (Lübeck, Schleswig-Holstein, Germany) 2002. [56] B. Radt, J. Serbin, B. I. Lange, R. Birngruber, G. Huettmann, Proc. SPIE 2001, 4 4 3 3 , 16. [57] M. Bever, Dr. rer. nat. Dissertation, University of Lübeck (Lübeck, Schleswig-Holstein, Germany) 2011.
SUPPORTING I NFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of the article. File S1. Supporting Information Figure S1. DLS measurements of (A) 15 nm and (B) 30 nm AuNP (red lines) and corresponding AuNP-siRNA conjugates (blue lines). Figure S2. Energy density in particles with 27 nm (short dashed), 30 nm (solid) and 33 nm (long dashed) irradiated with 4 ns pulse duration and radiant exposure of 25.2 mJ cm−2. Three heating phases (I – III) can be distinguished. I) initial phase of heating the AuNP until spinodal temperature is reached in the water. II) Energy density grows during AuNP temperature increase to 1336 K, while bubble formation causes thermal isolation and decreases sabs. The particle melts at constant particle temperature. III) Energy deposition in a liquid gold particle. The discontinuity at the beginning of phase II indicates bubble formation. Figure S3. TEM images of 30 nm AuNP without siRNA coverage before irradiation (A) and after radiant exposure of 5.76 mJ cm−2 (B), 10.13 mJ cm−2 (C), 24.50 mJ cm−2 (D), 50.0 mJ cm−2 (E), 100 mJ cm−2 (F) Figure S4. Absorption peak of 30 nm AuNP without siRNA shift by 5.5 nm after irradiation with 50 mJ cm−2. Figure S5. a) Fluorescence spectra the supernatants after irradiation with different radiant exposure. Increase of fluorescence is caused by increasing release of the coumarin labeled siRNA b) Fluorescence spectra in the supernatants after irradiation and the DTT treatment particle after irradiation. Particle diameter was 30 nm. How to cite this article: Rudnitzki F, Feineis S, Rahmanzadeh R, et al. siRNA release from gold nanoparticles by nanosecond pulsed laser irradiation and analysis of the involved temperature increase. J. Biophotonics. 2018;e201700329. https://doi.org/10.1002/ jbio.201700329
