Near-infrared nanosecond-pulsed laser-activated highly efficient intracellular delivery mediated by nano-corrugated mushroom-shaped gold-coated polystyrene nanoparticles

Tuhin Subhra Santra *a, Srabani Kar b, Te-Chang Chen c, Chih-Wei Chen d, Jayant Borana e, Ming-Chang Lee cf and Fan-Gang Tseng *egh
aDepartment of Engineering Design, Indian Institute of Technology Madras, India. E-mail: tuhin@iitm.ac.in
bElectrical Engineering Division, University of Cambridge, CB3 0FA, Cambridge, UK
cInstitute of Photonics Technology, National Tsing Hua University, Taiwan
dInstitute of Molecular Medicine, National Tsing Hua University, Taiwan
eDepartment of Engineering and System Science, National Tsing Hua University, Taiwan. E-mail: fangang@ess.nthu.edu.tw
fDepartment of Electrical Engineering, National Tsing Hua University, Taiwan
gInstitute of Nanoengineering and Microsystems, National Tsing Hua University, Taiwan
hDivision of Mechanics, Research Center for Applied Sciences, Academia Sinica, Taiwan

Received 2nd March 2020 , Accepted 4th May 2020

First published on 9th May 2020


Abstract

Here, an efficient intracellular delivery of molecules with high cell viability is reported using nanosecond-pulsed laser-activated plasmonic photoporation, mediated by high-aspect-ratio nano-corrugated mushroom-shaped gold-coated polystyrene nanoparticles (nm-AuPNPs) at near-infrared wavelength. Upon pulsed laser illumination, nm-AuPNPs exhibit greater plasmonic extinction than spherical AuPNPs, which increase their energy efficiency and reduce the necessary illumination of light, effectively controlling cell damage and improving the delivery efficiency. Nm-AuPNPs exhibit surface plasmon absorption at near infrared region with a peak at 945 nm. Pulsed laser illumination at this plasmon peak triggers explosive nanobubbles, which create transient membrane pores, allowing the delivery of dyes, quantum dots and plasmids into the different cell types. The results can be tuned by laser fluence, exposure time, molecular size and concentration of nm-AuPNPs. The best results are found for CL1-0 cells, which yielded a 94% intracellular PI dye uptake and ∼100% cell viability at 35 mJ cm−2 laser fluence for 945 nm wavelength. Thus, the presented approach has proven to have an inevitable potential for biological cell research and therapeutic applications.


Introduction

Introducing foreign biomolecules into targeted living cells efficiently without compromising cell viability is a challenging, yet critical task for many biological and therapeutic applications.1,2 Researchers have developed many transfection techniques, including the use of viral vectors,3 chemical methods such as basic proteins, complexes with lipids and calcium phosphate,4 and physical methods such as electroporation,5,6 lipid-mediated entry into cells,7 microinjection,8 jet injection,9 and sonoporation.10 Among these methods, chemical formulation, such as polymers and cationic lipids, encounter targeting efficiency issues, endocytic degradation, and the need for cell-specific formulation.11,12 Moreover, limitations of using viral vectors for gene transfer include mutagenicity, triggering an immune response, toxicity, and its high cost,13,14 while physical methods, such as ultrasound, laser irradiation, and electroporation, are effective substitutes for gene transfer without these limitations,15 are often insufficient.16 For example, while microinjection directly injects compounds into cells, it is time consuming and requires a highly skilled person to operate the task and the delivery throughput is low. In recent years, sonoporation or cellular sonication has been employed to permeabilize cell plasma membranes before delivery. The acoustically induced cavitation bubbles form microjets and shockwaves that create transient pores on the cell membrane. However, strong shear forces cause cell damage or toxicity.17 In contrast, electroporation techniques prove advantageous, owing to their operational ease, short process time, low toxicity, and better-electroporated size control to reduce cytosolic leakage components.5,6 However, the process encounters issues such as pH variation, electric field distortion and thermal effect, due to the application of micro- to millimeter-sized large electrodes, which induces a strong current and electrolysis effect that deteriorates cell viability.18,19

In recent years, photoporation has emerged as a promising physical method for intracellular delivery. In this technique, a high-intensity laser applied to the cell membrane to create permeable membrane pores that can deliver biomolecules. Using a multiphoton process and submicron membrane pores, femtosecond laser pulses can create low-density plasma on the cell membrane surface,20,21 achieving a maximum of 80% to 90% cell-specific transfection efficiency.22,23 Continuous-wave lasers heat the cell membrane to form permeable pores with less than 30% transfection efficiency,24,25 whereas, nanosecond laser pulses induce cavitation bubbles with a high transfection efficiency up to 88%.26,27 In recent years, nanoparticle-mediated techniques have been introduced to further enhance the coupling efficiency of photonic energy into cell poration.28,29 For example, lasers can interact with gold nanoparticles (AuNPs) to produce nanoscale energy deposition to the surrounding medium through light amplification and heat generation. Among various nanoparticles (NPs), AuNPs possess many valuable properties, such as surface plasmon resonance, high chemical stability, biocompatibility, convenient surface bioconjugation with molecular probes and low toxicity.28,30,31 Different gold nanostructures such as nanocrescents, nanorods, nanoshells, nanocages, and half nanoshells have excellent photothermal properties.30,32,33 Due to the high surface plasmon resonance of AuNPs, upon pulsed laser illumination of the cell membrane-AuNPs interface, high photoporation efficiency can be achieved, when large amounts of cells are clumped together.28,29 When a high-intensity nanosecond (<10 ns) laser pulse interacts with AuNPs, the temperature increases rapidly and the surrounding medium is vaporized to form plasmonic nanobubbles (PNBs) or photothermal vapor nanobubbles (VNBs) across the AuNPs surface.33,34 After the bubbles collapse, the generated cavitation and stress waves cause heat transfer from the AuNPs to the surrounding area. The amount of generated heat depends on particle size and shape, aggregation of particles, laser intensity, wavelength and pulse duration. However, in case of a very short lifetime (<1 μs) of PNBs, the heat transferred from the AuNPs to the surrounding medium is negligible and all irradiated energy from the AuNPs is transferred as mechanical energy.35,36 When this mechanical energy exceeds the cell membrane threshold values, transient membrane pores are created, allowing biomolecules to be gently delivered into the cells.35,36 This approach was followed by Xiong et al., who used spherical AuNPs to achieve best delivery efficiency of 85% (FITC-dextran) and cell viability ∼90% at very high laser fluence of 2 J cm−2.37 However this energy is high enough by means of cell damage and energy savings, and thus still there is free room to reduce the energy and improve the delivery efficiency and cell viability. In addition, most of the studies used visible wavelength for cell poration.38,39 Whereas, it is a well-established fact that maximum penetration of electromagnetic radiation in biological tissue occurs approximately from 650 nm to 1350 nm-the near infrared (NIR) range; therefore possesses great promise for in vivo applications.40,41 Furthermore, NIR range is highly biocompatible and photothermal damage is much less than UV and visible range. Especially NIR range is an emerging field in photothermal therapy. Therefore it is highly demanding to achieve optoporation and biomolecular delivery in NIR. In the present study, our goal is to establish a pathway for utilizing NIR range with lower fluence for cell optoporation. With this background, we aim to shift operating wavelength towards NIR by using high-aspect-ratio nano-corrugated mushroom-shaped gold-coated polystyrene nanoparticles (nm-AuPNPs).

Compared to spherical metal nanoparticles, high-intensity surface plasmon can be achieved on non-spherical metal NPs, such as nano triangles and nanorods with sharp edges.42,43 In our previous articles, we showed the fabrication of high-aspect-ratio nm-AuPNPs, which enhanced the optical hot-spot intensity, resulting in surface-enhanced Raman spectra (SERS).44,45 Here, we utilized these nm-AuPNPs as a mediator and applied the nanosecond-pulsed laser to achieve highly efficient intracellular delivery of cargo molecules with high cell viability. After illumination of the laser pulse on nm-AuPNPs (generating ∼120–130 hot spots per particle), light absorption increased on the hotspots, resulting in higher production of surface plasmon across the horizon of nm-AuPNPs compared to spherical AuPNPs. Consequently, nm-AuPNPs transferred higher energy to generate PNBs, following a highly efficient intracellular delivery without any cell damage. The best delivery results were optimized at 945 nm with delivery efficiency ∼94% (PI dye uptake) and cell viability ∼100%, at laser fluence of 35 mJ cm−2 (for spherical AuPNPs, we obtained a delivery efficiency of 42% and cell viability of 91% at 35 mJ cm−2) for lung cancer (CL1-0) cells. Moreover, we found molecular uptake efficiency and cell viability depend on laser fluence, exposure time, molecular size and concentration of nm-AuPNPs. Using pulsed laser-assisted photoporation technique, we successfully delivered dyes, quantum dots (QDs) and different plasmids into lung cancer (CL1-0), gastric cancer (AGS) and P-19 embryonic stem cells, with minimum laser fluence (25 mJ cm−2–35 mJ cm−2) and 30 seconds exposure time at 516 nm for nm-AuPNPs. Therefore our methods pave a way to achieve high efficient intracellular delivery with high cell viability at much lower energy NIR radiation.

Results and discussion

nm-AuPNPs mediated intracellular delivery

Fig. 1 shows the schematic experimental procedure for nm-AuPNPs mediated intracellular delivery using a nanosecond-pulsed laser. After we cultured cells overnight in an incubator (37 °C with 5% CO2), polyethylene glycol (PEG) mediated nanoparticles were placed into the cell culture dish for an hour before the experiment and washed the unbound nm-AnPNPs from the cell surface using phosphate buffer saline (PBS). Then cell-impermeable cargo molecules (Fig. 1(b)) were introduced into the dish right before the experiment. Due to the nanosecond-pulsed laser illumination on the nm-AuPNPs (Fig. 1(c)) attached to the cell membrane, electromagnetic field enhancement occurred near the nano-corrugated edges, resulting in a lightning-rod effect, also known as optical hotspots. The resonantly excited optical hotspots generated local heat, causing a rapid temperature enhancement and subsequent formation of PNBs surrounding the nm-AuPNPs-cell membrane interface. After PNBs formed, they rapidly grew, coalesced, and collapsed to induce an explosion, which caused a strong fluid flow at the nm-AuPNPs-cell membrane interface. After PNBs collapsed, the generated cavitation and stress wave transferred heat from the nm-AuPNPs to the nano-restricted region of the cell membrane.37
image file: d0nr01792b-f1.tif
Fig. 1 Schematic overview of photoporation procedure for intracellular delivery (a) cells were incubated with PEG mediated nm-AuPNPs for attachment onto the cell membrane (b) cell surface was washed to remove unbound nm-AuPNPs and then cargo molecules were added just prior to laser exposure (c) nanosecond pulse laser-induced plasmonic nanobubbles formation at nm-AuPNPs and cell membrane interface resulting in transient pore formation into the cell membrane (d) successful intracellular delivery of cargo molecules with membrane reseal.

Furthermore, nanosecond-pulsed laser exposure (pulse duration ∼5 ns, 10 Hz repetition rate) results in PNBs having very short lifetime (<1 μs), which is in turn beneficial for negligible heat transfer from gold nanoparticle to the cell membrane.35,36 Therefore, almost all irradiated energy from the nm-AuPNPs was transferred as mechanical energy, in the form of strong fluid flows, which disrupted the plasma membrane forming transient membrane pores, allowing the delivery of cargos from outside to inside the cell (Fig. 1(c)). After the cargo was delivered, the cell membrane resealed again, maintaining cell viability (Fig. 1(d)). The transient pore formation on the cell membrane depends on laser fluence and the concentration of nm-AuPNPs. A detailed fabrication process of nm-AuPNPs is provided in ESI Fig. S1.Fig. 2(a) and Fig. S2 shows the scanning electron microscopy (SEM) image of fabricated high-aspect-ratio nm-AuPNPs is approximately 300 nm in diameter and each contains approximately 120–130 hotspots.


image file: d0nr01792b-f2.tif
Fig. 2 nm-AuPNPs fabrication and attachment on cell surface (a) SEM image of fabricated nm-AuPNPs (300 nm particle) (b) nm-AuPNPs synthesis and surface modification by using polyethylene glycol (PEG) for strong attachment between nm-AuPNPs and cell membrane (c) the extinction spectra of the particles have peaks at 516 nm and 945 nm wavelength (d) SEM image of nm-AuPNPs attachment on lung cancer (CL1-0) cells.

Details of the fabrication process are reported in our previous articles.44,45 After fabrication, we modified the particle surfaces with HS-PEG-COOH to form a covalent bond between nm-AuPNPs and polyethylene glycol (PEG) (Fig. 2(b)), for better adhesion onto the cell membrane.46 We detected two peaks at 516 nm and 945 nm as shown in Fig. 2(c). The absorption band peaked around 516 nm arises due to interband electronic transitions.47–49 The peak at 945 nm corresponds to surface plasmon resonance at the gold surface. The plasmon peak position can be tuned by changing the aspect ratio of the particle through either dry etching or thin-film coating methods,44,45 or wet-chemistry synthesis.50 ESI Fig. S3 shows the corresponding UV extinction spectra of spherical AuPNPs with peaks centered at 522 nm, 630 nm and 730 nm, which completely differs from the nm-AuPNPs spectra (corresponding field enhancement results are shown in ESI Fig. S4). Fig. 2(d) shows a scanning electron microscopy (SEM) image of nm-AuPNPs attached to the cell membrane surface. The average number of nm-AuPNPs attachment was approximately 50–60 per cell.

We exposed cells mixed with propidium iodide (PI) dye with spherical AuPNPs (at 522 nm) and asymmetry type nm-AuPNPs (at 516 nm) separately, using the same laser fluence (25 mJ cm−2) for 30 seconds at 10 Hz pulse frequency. We found that for CL1-0 cells (see ESI Fig. S5), spherical AuPNPs achieved a maximum of 24% delivery efficiency with 95% cell viability. Also, we examined different cell types by using spherical AuPNPs, as shown in ESI Fig. S6, resulting a maximum cellular delivery efficiency of around 17–24% at 522 nm. Interestingly, nm-AuPNPs achieved a maximum of 88% delivery efficiency along with 92% cell viability at 516 nm (Fig. 4a). This is the direct consequence of higher energy absorption of nm-AuPNPs, which clearly improves delivery efficiency and cell viability compared with spherical AuPNPs.

It is important to note that the best results were found at 945 nm wavelength, which yielded a 94% intracellular PI dye uptake and ∼100% cell viability at 35 mJ cm−2 laser fluence for CL1-0 cells. Following paragraphs will discuss the impact of different experimental parameters only for the excitation wavelength of 945 nm which is the main focus of this study.

To justify the PNBs as the physical origin of the cell transfection at 945 nm, we have carried out electromagnetic simulations in 3D finite element method by using Comsol Multiphysics software. The model particles are in core–shell configuration where core is the polystyrene and shell is the gold material. The spherical particle has gold coating with thickness of 30 nm on top of spherical polystyrene of size around 280 nm. For mushroom-shaped particle the gold has an average thickness of ∼30 nm coated on the corrugated surface of polystyrene. The wavelength dependent complex refractive index (n + ik) of are taken from ref. 51 for Au and ref. 52 for water medium in the simulation. The complex refractive index of polystyrene was kept fix for all wavelengths to n + ik = 1.59 + i0.02.53Fig. 3 shows the cross-sectional view of COMSOL Multiphysics simulated electric field enhancement results of the spherical particle and nm-AuPNPs placed in water for the excitation wavelengths 945 nm. From the simulations, it is clear that plasmonic electromagnetic field enhancement is approximately 4.5 times higher at the edge of the nm-AuPNPs than that of spherical AuPNPs. Usually, a few nm size gold nanoparticle shows plasmonic peak at around 532 nm.38,39,54 In contrast, the strong field enhancement at 945 nm arises due to the larger size of the particle, which causes electromagnetic retardation and the shifts of the plasmonic peak towards the infrared region.54,55 The higher enhancement in mushroom-shaped particle occurs because of corrugated surface, which causes nanolocalized plasmonic enhancement in the vicinity of the tips of corrugated surface and the inter-spike nano-space. The field enhanced regions act as plasmonic hot spots and also result in enhancement of Joule heating. As a consequence, the temperature of the gold nanoparticle increases above its initial temperature and diffuses into its surrounding medium. The transient rise of local temperature has been simulated and discussed in the following paragraph.


image file: d0nr01792b-f3.tif
Fig. 3 (a and b) Cross-sectional view of Comsol Multiphysics simulated electric field distribution at 945 nm for spherical (a) and (b) nm-AuPNPs in core–shell configuration with polystyrene as core and Au taken as shell. (c and d) Corresponding temperature distribution when Tl in Au shell reaches the maxima. Temperature outside (inside) the shell corresponds to the distribution of water (polystyrene) temperature Tm (Tc). Excitation pulse energy of ∼35 mJ cm−2 with pulse width of 5 ns was used in this simulation. (e and f) Transient temperature rise at an arbitrary point on gold/water interface for (e) spherical and (f) nm-AuPNPs.

To obtain temperature rise, we have adapted two temperature model to account the temperature rise of gold. The pulse laser exposure results in the rise of electron temperature, Te in gold shell. The increase of Te leads to the rise of gold lattice temperature Tl as well due to strong electron–phonon coupling. The temperature of the surrounding medium Tm also starts to increase within a few picoseconds due to the transfer of heat from nanoparticle to the adjacent medium. The variation of these temperatures in time and space are described by the following rate equation model:56

 
image file: d0nr01792b-t1.tif(1)
 
image file: d0nr01792b-t2.tif(2)
 
image file: d0nr01792b-t3.tif(3)
 
image file: d0nr01792b-t4.tif(4)
where Ce(Te) = γTe, with γ = 70 J m−3 K−2 is the electron heat capacity for gold,56Cl = 3 × 106 J m−3 K−1 is lattice heat capacity of gold,56ke = 300 W m−1 K−1 for the electron thermal conductivity;56kl ∼ 2.7 W m−1 K−1 for the lattice thermal conductivity;57g = 2.0 × 1016 W m−3 K−1 for gold electron–phonon coupling coefficient;56Cm = 4182 J m−3 K−1 the water heat capacity;56km = 0.6 W m−1 K−1 for the water thermal conductivity.56 The terms gl(TlTm) in eqn (1)–(4) represent the boundary interface heat exchange between the gold lattice and the surrounding water, where gl = 105 × 106 W m−2 K−1 are the corresponding thermal boundary conductance.56 The terms gc(TlTc) in eqn (1)–(4) represent the boundary interface heat exchange between the gold lattice and the polystyrene core, where gc = 508 × 106 W m−2 K−1 is interface (gold/polystyrene) thermal conductance calculated following eqn (5) in ref. 58 and 59. In eqn (1), the source term Q(t) is defined by the transient resistive heating due to a 5 ns Gaussian laser pulse excitation at the fluence of 35 mJ cm−2. The above equations do not include any phase change phenomena due to the rise of temperature. The spatial distribution of the temperature at peak position (at the point where Tl reaches its maximum value) is shown in Fig. 3c and d. The transient dynamics of Te, Tl, and Tm at an arbitrary interface point (where the field enhancement is maximum) are shown in Fig. 3e for spherical particle and Fig. 3f for mushroom-shaped particle. Again, all the transient simulations were done by using 3D Comsol Multiphysics.

It is worth noting that Te and Tl increase synchronously for both particles, and Tl reaches the melting temperature (∼1337 K) and the boiling temperature (∼3129 K) of bulk gold at the early stage of the arrival of the laser pulse.60,61 Thus, though the simulation results show that Te and Tl increase far above the melting temperature of gold, in practice, the real values cannot exceed the boiling temperature of bulk gold (∼3129 K).62 Therefore, it is very likely to obtain particle deformation, surface melting, and evaporation followed by the size-reduction of the gold nanoparticles for both shapes.62 Lombard et al. also estimated that cavitation bubbles start to generate when the lattice temperature of gold at gold/water interface rises above a certain critical temperature – named as spinodal temperature, Tspin ∼550 K.61 Therefore we expect the generation of plasmonic cavitation bubbles surrounding the nanoparticle/water interface. According to the above simulations, those phenomena are more likely to occur in mushroom-shaped particles compare to spherical shaped particles because the temperature rise is around one order of magnitude higher in the former one (Fig. 3(c–f)). The post pulses (2nd, 3rd … so on) will act on the deformed shaped particles and reduce the probability of temperature rise and bubble generation. However, during these processes, whenever Tl becomes more than 550 K, bubble generation will occur at the gold/water interface.61 The experimental visualization of bubble formation and its temporal dynamics at each step after pulse excitation is beyond the scope of the present work. All these results reinforce the claim of dynamic plasmonic bubbles to be the origin cell poration at 945 nm.

Parameters to control delivery efficiency and cell viability

nm-AuPNPs concentration. To determine the highest delivery efficiency and cell viability results for PI dye uptake, we tested different concentrations of nm-AuPNPs. We reported optimum delivery efficiency, and cell viability by using: 1.75 × 1011 particles per ml at 30–35 mJ cm−2 laser fluence with 10 Hz pulse frequency for 30 seconds laser exposure at 945 nm for CL1-0, AGS, and P-19 stem cells. We found that increasing particle concentrations above 1.75 × 1011 particles per mL, reduced delivery efficiency and cell viability (Fig. S7). As shown in Fig. S8, higher concentrations of nm-AuPNPs (3.8 × 1011 particles per mL) produced more debris (bright-field image after laser exposure) and dramatically reduced cell viability (calcein-AM for live cell staining) and delivery efficiency of PI dye (Fig. S8 (c and d)). Moreover, when we introduced pulsed lasers (30 mJ cm−2 for 30 seconds with 10 Hz pulse frequency) on AGS cells without nm-AuPNPs, we reported no uptake of PI dye into the cells (Fig. S9). Therefore, appropriate particle concentrations play an important role in intracellular delivery.

The toxicity test results were provided in Fig. S10 for CL1-0 cells, where pulse laser was exposed on first day with nm-AuPNPs and then cell viability was tested on different days. The results show that, most of the cells are viable after three days.

Exposure time. ESI Fig. S11 shows the delivery efficiency and cell viability into CL1-0 cells as a function of laser exposure time, using 35 mJ cm−2 laser fluence at 945 nm. We reveal that increasing exposure times to 30 seconds, increases the delivery efficiency to a maximum of ∼94%, and cell viability to ∼100%; above this exposure time, both delivery efficiency and cell viability began to decrease.
Laser fluence and wavelength. Fig. 4 shows the effect of laser fluence on intracellular delivery efficiency and cell viability of CL1-0 cells at 516 nm and 945 nm with 30-second laser exposures (Fig. S12 shows the data for AGS and P-19 stem cells at 945 nm). Upon increasing laser fluence at 516 nm, we achieved the highest delivery efficiency of 88% and 92% cell viability at 25 mJ cm−2, which then decreased with further laser fluence increases. At 945 nm, the maximum delivery efficiency was 94%, with ∼100% cell viability at 35 mJ cm−2 laser fluence.
image file: d0nr01792b-f4.tif
Fig. 4 Effect of nanosecond pulse laser fluence on intracellular delivery efficiency and cell viability of CL1-0 cells at 516 nm and near-infrared at 945 nm wavelength under nm-AuPNPs mediation (a) the delivery efficiency and cell viability at 516 nm with different laser fluence. (b) The delivery efficiency and cell viability at 945 nm with different laser fluence. Data show the average ± standard deviation (n = 3 replicate).

Our results demonstrate that delivery efficiency and cell viability at 945 nm is higher than at 516 nm. To confirm these results, we also tested a non-resonant wavelength at 660 nm (Fig. S13), where PI dye delivery was not achieved, indicating no cell poration. However, a very few cells showed red fluorescence imaging (Fig. S13b, f and j) indicating the nucleus of dead cells.

The optimum fluence of illumination was 25 mJ cm−2 for 516 nm. Fig. S14 shows PI dye delivery and viability of CL1-0 cells at 50 mJ cm−2 laser fluence with a 10 Hz pulse frequency for 30-second laser exposure at 516 nm. The results clearly show that most of the cells are dead at higher laser fluence; only about 3.7% of cells were viable.

Here, it is worth noting that a previous study on cell opto poration induced by plasmonic nanobubbles generated through pulsed laser heated Au nanoparticles required much higher fluence (∼0.38 J cm−2–2.05 J cm−2) to achieve successful cell poration.37 Their study used 70 nm spherical Au nanoparticles in the experiments.37 There are several studies showing field enhancement (|E/Ein|) in the range of 6–8 for similar size (60–80 nm) Au nanoparticles,63,64 which is around 10 times lower than our spherical AuPNP (Fig. 3a) and around 14 times lower than the nm-AuPNPs (Fig. 3b) used in our study at 945 nm. Therefore absorption efficiency and the heating effect would be much higher in the later particles resulting in the requirement of lower threshold fluence (EPNB) for the generation of PNBs. In addition, the increase of particle size in core–shell configuration reduces the EPNB as per previous studies by Lukianova-Hleb et al.65 It showed EPNB is lower in core–shell structure compare to same size all-Au spherical particle.65 For example, EPNB for 250 nm solid Au sphere was ∼100 mJ cm−2 whereas ∼40 mJ cm−2 for silica@gold shell at 532 nm.65 It is also important to note that bigger size particle shifts central wavelength for plasmonic resonance towards infrared region. Thus, at resonant excitation EPNB is expected to be more lower value compare to nonresonant excitation. All these previous understandings65 along with our temperature simulation results readily support our observation of lower threshold fluence for the onset of plasmonic nanobubbles compare to the previous study.37

Intracellular delivery

Molecular dye delivery. We have used PI dye to test the delivery efficiency at the molecular level. PI dye is cell impermeable, but it delivers into a live cell cytosol, visualized by red fluorescence, only after physical rupture of the cell membrane.5,6 We added PI dye (∼45 μl mL−1, Fig. S15 showing delivery efficiency and viability as different dye concentrations) to the dish immediately before we activated the nanosecond-pulsed laser. To test cell viability, we added cell-permeable calcein-AM dye after three hours of laser exposure. The calcein-AM can hydrolysis inside live cell and produce green fluorescence color.66,67Fig. 5 shows successful PI dye delivery into CL1-0, AGS and P-19 embryonic stem cells, which remained highly viable, using 35 mJ cm−2 laser fluence at 945 nm at 10 Hz pulse frequency for 30 seconds. Fig. 5(a, d and g) show successful PI dye delivery into the CL1-0, AGS and P-19 stem cells, respectively, whereas Fig. 5(b, e and h) show the corresponding live cells (calcein-AM staining) after PI dye delivery and Fig. 5(c, f and i) show the merged images of PI dye delivery and live-cell imaging.
image file: d0nr01792b-f5.tif
Fig. 5 nm-AuPNPs mediated PI dye delivery into different cell type at 945 nm laser excitation (a) red fluorescence indicating successful PI dye delivery into lung cancer (CL1-0) cells (b) green fluorescence indicating maximum cells are viable by using calcein-AM (c) merged image of PI dye and calcein-AM (d–f) PI dye delivery, cell viability and merge image of gastric cancer (AGS) cells (g–i) PI dye delivery, cell viability and merge image of P-19 embryonic stem cells. Data show the average ± standard deviation (n = 3 replicate).

To confirm that nm-AuPNPs effectively delivers PI dye into cells, we conducted area restricted pulse laser exposure on cells with nm-AuPNPs. The results depicted that PI dye is only successfully delivered through the exposed area (Fig. S16(b)) into CL1-0 cells, and all cells remained viable (calcein-AM staining) after delivery in both exposure and non-exposure areas (Fig. S16(c)). We observed similar results for AGS and P-19 stem cells, which are shown in ESI Fig. S17.

Fig. 6(a) shows the delivery efficiency and cell viability at 945 nm using nm-AuPNPs, where we recorded maximum delivery efficiency of 94% with ∼100% cell viability of CL1-0 cells. For AGS and P-19 cells, the delivery efficiency was 85% and 91%, respectively, while cell viability was 94% and 98%, respectively. In general, we found cell viability was higher at 945 nm, compared with those at 516 nm (see Fig. S18 and S19). The best results at 516 nm were achieved with 88% efficiency and 94% viability for CL1-0 cells (Fig. S19). For AGS and P-19 stem cells, 81% and 84% efficiency were achieved with 85% and 88% cell viability, respectively (Fig. S19). On the other hand, we tested spherical AuPNPs which required ∼120 mJ cm−2 laser fluence at 630 nm (Fig. S20) and ∼110 mJ cm−2 laser fluence at 730 nm (Fig. S21) for 30 seconds. From these results, it is clear that the energy requirement is almost 3–4 times higher for spherical AuPNPs compare to nm-AuPNPs.


image file: d0nr01792b-f6.tif
Fig. 6 PI dye and QDs delivery efficiency and cell viability of different cancer cells and one stem cell at 945 nm wavelength (a) The PI dye delivery efficiency (94%) and viability (∼100%) were higher for CL1-0 cells, whereas the delivery efficiency (85%) and viability (94%) were lower for AGS cells (b) QDs delivery efficiency and viability for deferent cells, where maximum efficiency (91%) and viability (95%) were found for CL1-0 cells and lower delivery efficiency (70%) with higher viability (94%) for P-19 stem cells. For both cases, the laser exposure was 30 seconds at 10 Hz pulsing frequency with 35 mJ cm−2 laser fluence. Data show the average ± standard deviation (n = 3 replicate).
QDs delivery. To examine the delivery efficiency of QDs, using nm-AuPNPs, we initially washed unbound nm-AuPNPs from the cell dish using PBS solution, before introducing CdSxSe1−x/ZnS core nanocrystals coated with COOH functional ligands into the cell dish. After laser exposure, we immediately washed the QDs from the dish and introduced cell culture medium (DMEM for CL1-0 and AGS; MEM for P-19 cells) into the dish for cell imaging. Fig. 6b shows the QDs delivery efficiency and cell viability of different cancer cells and one type of stem cells.

Fig. 7 shows fluorescence images of QDs delivery and describes cell viability of CL1-0, AGS, and P-19 embryonic stem cell at 35 mJ cm−2 laser fluence with 945 nm laser excitation at 10 Hz pulse frequency for 30 seconds. Fig. 7a shows QDs delivery for CL1-0 cells exhibited 91% delivery efficiency and 95% cell viability (see Fig. 6b). The calcein-AM staining of live cells in Fig. 7b indicates that the majority of the cells remained viable after QDs delivery. The merged image of QDs delivery and live cell staining is shown in Fig. 7c, which indicated that QDs delivered into cells. For AGS cells and P-19 stem cells, the delivery efficiency was 83% and 70%, respectively, and cell viability was 90% and 94%, respectively. The QDs were about 5.5–6.5 nm in size, whereas the PI dye was approximate 0.69 nm in diameter and 1.55 nm in length (cylindrical approximation). Therefore, the lower delivery efficiency of the QDs compared with the PI dye is due to its larger molecular size and weight, resulting in a lower diffusion rate. Fig. S22 shows QDs delivery on the non-exposure area of CL1-0, AGS cells, and P-19 stem cells at 945 nm with 35 mJ cm−2 laser fluence for 30 seconds. The results clearly show that in non-exposure areas, QDs were not delivered into different cancer cells and stem cells (Fig. S22c, g and k), and the cells remained viable (Fig. S22d, h and i).


image file: d0nr01792b-f7.tif
Fig. 7 QDs delivery and cell viability for CL1-0, AGS and P-19 embryonic stem cell at 945 nm wavelength (a) red fluorescence indicating successful QDs delivery into lung cancer (CL1-0) cells (b) green fluorescence indicating all cells remain viable after using calcein-AM (c) merge image of QDs and calcein-AM (d–f) QDs delivery, cell viability and merge image of gastric cancer (AGS) cells (g–i) QDs delivery, cell viability and merge image of P-19 embryonic stem cells. For each data shows the average ± standard deviation (n = 3 replicate).
Plasmid delivery with dose control. For efficient plasmid delivery, we used two different plasmids, pMax-E2F1, and pCAG-GFP. For these experiments, after washing unbound nm-AuPNPs from the cell surface, we added different concentrations of the plasmids into the cell culture dish, before exposing laser pulses at 40 mJ cm−2 laser fluence with 945 nm at 10 Hz pulse frequency for 30 seconds. After laser exposure and before cell incubation, we washed all residual plasmids from the cell culture dish and added the required fresh medium onto the dish. Then approximately after one day, we captured the final protein expression image. Fig. 8(a) shows the dose control transfection efficiency and cell viability for CL1-0, AGS and P-19 stem cells. For pMax-E2F1, at 71.81 ng μL−1 plasmid concentration, the maximum transfection efficiency occurred using CL1-0 cells, which reached 86% with 96% cell viability. Fig. 8(b and c) shows confocal scanning images for pMax-E2F1 and pCAG-GFP delivery and GFP expression results using CL1-0 cells, whereas Fig. 8(d–g) show GFP expression of AGS and P-19 stem cells. Fig. 8(h and i) shows the transfection efficiency and cell viability of CL1-0, AGS and P-19 stem cells using pMax-E2F1 and pCAG-GFP at 945 nm pulse laser excitation. For pMax-E2F1, we achieved 82% and 73% transfection efficiency, and 98% and 96% cell viability, for AGS and P-19 stem cell, respectively. While for pCAG-GFP, we achieved 79%, 81%, and 63% transfection efficiency, and 92%, 96% and 94% cell viability, for CL1-0, AGS and P-19 stem cells, respectively.
image file: d0nr01792b-f8.tif
Fig. 8 Doses control, transfection efficiency, cell viability and confocal scanning images of plasmid delivery for GFP expression of CL1-0, AGS and P-19 stem cell at 945 nm wavelength (a) transfection efficiency and viability of CL1-0 cells for pMax-E2F1 delivery with different plasmid concentrations (b & c) confocal scanning images of pMax-E2F1 and pCAG-GFP plasmid delivery and protein expression of CL1-0 cells after 24 hours (d & e) pMax-E2F1 and pCAG-GFP plasmid delivery and protein expression of AGS cells after 24 hours (f & g) pMax-E2F1 and pCAG-GFP plasmid delivery and protein expression of P-19 embryonic stem cells after 24 hours (h) transfection efficiency and cell viability using pMax-E2F1 of CL1-0, AGS and P-19 stem cells (i) transfection efficiency and cell viability using pCAG-GFP of CL1-0, AGS and P-19 stem cells. Data show the average ± standard deviation (n = 3 replicate).

We examined both plasmids for delivery in laser exposure and non-exposure areas, as shown in Fig. S23 and S24. The results demonstrate successful plasmid delivery occurred only into the exposure areas. Noteworthy, using our nm-AuPNPs mediated photoporation process, traditionally hard to transfect cells, such as stem cells (currently P-19 cells) can be successfully transfected with plasmids at an efficiency of 63% and 73%, for pCAG-GFP and pMax-E2F1, respectively, which has tremendous potential for basic cellular research and biomedical application purposes.68–70

Cell viability test. The viability of cells was assessed via (MTT), 3-[4,5-dimethylthiozol-2-yl]-2,5 diphenyl tetrazolium bromide (SRL, Mumbai, India) assay for HCT-8 and HeLa cell lines. Briefly, 5000 cells were seeded in 96 well plates (Tarsons, India) along with 20 μl nanoparticles and 180 μl of DMEM high glucose medium (Gibco™) in sterile conditions (Fig. 9).
image file: d0nr01792b-f9.tif
Fig. 9 Cell viability of HCT-8 and HeLa cells using MTT assay on different days after laser exposure at 945 nm wavelength corresponding to 35 mJ cm−2 energy for the 30 s each.

After overnight incubation at 37 °C with a 5% CO2 incubator, the cultured cells were exposed using pulsed laser at 522 nm and 945 nm wavelength corresponding to 35 mJ cm−2 energy for the 30 s each (Fig. S25 shows viability results using MTT assay at 516 nm wavelength). Afterward, cells were washed with PBS and then cell culture media was changed at alternate days. The MTT assay was performed after 2 h, 1 day, 2 days and 3 days of laser exposure. For each MTT experiment, 30 μl of 10 mg ml−1 MTT was added to the individual well followed by 4 h incubation at 37 °C. After incubation in dark for 4 h, the medium containing MTT dye was removed and 100 μl of DMSO (SIGMA, India) was added for dissolving the formazone crystals formed after adding MTT dye. Then the optical density of formazone dissolved DMSO was measured at 570 nm using ELISA plate reader. The measured optical density corresponds to the viability of cells, higher the optical density, greater is the cell viability.

Experimental

nm-AuPNPs fabrication

We considered 510 nm diluted polystyrene (PS) beads as the base templates, which are arranged as dense-packed monolayer hexagonal structure on a substrate for receiving the nanofabrication process only on the top side. ESI Fig. S1 shows the process step of nm-AuPNPs fabrication. Initially, the substrate was cleaned by the piranha solution with a 7[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio of sulfuric acid (H2SO4) and hydrogen peroxide (H2O2) at 80 °C for 10 minutes. After that, the substrate was treated with oxygen (O2) plasma under 100W RF power, 75 mTorr pressure and 10 sccm oxygen flow rate to prepare the substrate as hydrophilic in nature. Then the substrate boundary was covered by PDMS films with 250 μm in height to form as a wall. The PDMS layer was made by spin coating with 5 g of uncured PDMS at 1000 rpm for 30 seconds onto a 4 inch glass wafer, pre-coated with a 1% Teflon solution and finally heated with 45 °C overnight to form as 250 μm thickness of the layer. Then 3 μg ml−1 of the PS beads solution was added into the PDMS mounted well. After water evaporation, the close-packed monolayer PS beds started to form, from the center part to the PDMS wall by the combined effect of capillary force and convective force.71 After peeling off the PDMS well, the sample was etched by reactive ion etching (RIE) to form as nanocorrugated structure on top of the PS beads (power = 55 W, O2 flow = 5 sccm, time = 260 s, pressure = 60 mTorr). Then we deposited Ti/Au (8/30 nm) by electron beam evaporation to form final nm-AuPNPs. A detail fabrication process of nm-AuPNPs is described in our previously published articles.44,45

Cell culture

For nanosecond pulse laser-induced intracellular delivery, we have used three different types of cells – lung cancer CL1-0, gastric adenocarcinoma cell line AGS and P-19 embryonic stem cells. All of these cell lines were originally from the American Type Culture Collection (ATCC, Manassas, VA, USA). Both CL1-0 and AGS cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 1% sodium pyruvate, 1% non-essential amino acids and 1% penicillin–streptomycin (Gibco, USA). While P-19 embryonal carcinoma cells were maintained and cultured according to the culture method of ATCC CRL-1825 from atcc.org. Cells were cultured with Minimum Essential Medium (MEM) (Invitrogen 61100) and supplemented with 10% fetal bovine serum (BI 04-001-1A), 100 units ml−1 penicillin/100 μg ml−1 streptomycin (Invitrogen 15140), 2 mM L-glutamine (Life Technologies 25030081), and 1 mM sodium pyruvate (Life Technologies 11360070). The exposure circumstances were 5% CO2 and 37 °C and a subcultivation ratio was 1[thin space (1/6-em)]:[thin space (1/6-em)]10 every 48 h. Before the experiment, cells were dissociated from the dish surface with trypsin (0.05% trypsin–EDTA, GIBCO), and suspended in the regular culture medium at a density of approximately 3 × 106 cells per ml.

Materials (dyes, QDs and plasmids preparation)

The PI dye (P1304MP, Invitrogen, USA), calcein AM (C3099, Invitrogen, USA), Hoechst 33342 (H3570, Invitrogen, USA) and ethidium homodimer (ETHD-1) (E1169, Invitrogen, USA) were purchased from Life Technologies, USA. The poly (ethylene glycol) 2-mercaptoethyl ether acetic acid (HSC2H4O (C2H4O)nCH2CO2H) (average Mn 3500) was purchased from Sigma Aldrich, USA, and cdSxSe1−x/Zns core/shell nanocrystals coated with COOH functional ligands (Nanocrystals NC665C, size 5.5 ∼6.5 nm) was purchased from Crystalplex, Pittsburgh, USA. Both the plasmids were purchased from Addgene (http://www.addgene.com), where pMax-E2F1 (addgene #16007, Cambridge, MA, USA) is 3500 bp and expresses GFP-E2F1 driven by a pCMV promoter. The pCAG-GFP (addgene #11150, Cambridge, MA, USA) is 4779 bp and expresses GFP via chicken beta-actin promoter with CMV enhancer. For both plasmid preparation, DH5α E. coli strain was grown overnight in 100 ml LB/ampicillin or LB/kanamycin medium. The plasmid purification is guided by QIAGEN® plasmid kit midi (#12145).

Particle synthesis with PEG

The poly(ethylene glycol) 2-mercaptoethyl ether acetic acid (HSC2H4O (C2H4O) nCH2CO2H) was purchased from Sigma Aldrich. For increasing the colloidal stability of nm-AuPNPs, it was encapsulated with HS-PEG-COOH (molecular weight 3500 g mol−1) to form nm-AuPNPs@PEG. In summary, 1 ml of gold-coated nanoparticle (1.75 × 1011 particles per ml) aqueous suspension was mixed with 1 mg of thiolate PEG. The concentrated solution of NaOH was used to maintain the pH of the solution around 12.0. The reaction mixture was left overnight to undergo a ligand-exchange reaction under mild stirring conditions. Then, the reaction mixture was centrifuged twice for 15 min at 9500 rpm and dispersed in 1 ml Milli-Q water. The high disparity and stability of the gold-coated nanoparticles proves the efficient coating with thiolate PEG.50

SEM images of cells with nanoparticles attachment

The PEG mediated nm-AuPNPs were introduced into the cell culture dish and then (after one hour) we washed out unbound nm-AuPNPs from the cell membrane by using phosphate buffer saline (PBS). After that, samples were treated with a different combination of distilled water (DI) and ethanol and then, it was treated again with different ratios of ethanol and hexamethyldisilazane (HMDS). Finally, samples were treated with 100% HMDS and left it for evaporation into the fume hood overnight. After evaporation, we dried the samples at 50 °C for two hours and coated 5 nm gold on top of the cells. Finally the SEM (Hitachi TM-1000, version = 03-02, magnification = 3000, working distance = 5880 μm) images were captured.

Conclusions

This study proposes that nm-AuPNP mediated photoporation by employing nanosecond laser pulses enhances intracellular delivery efficiency and cell viability. The asymmetrical nature of nm-AuPNPs raises 120–130 fold higher hotspots compared to spherical AuPNPs, allowing to harvest more laser energy at 945 nm due to plasmon resonance. This leads to the generation of local heat and subsequent formation of PNBs, which induce a strong fluid flow at the nm-AuPNPs-cell membrane interface, creating transient membrane pores that gently deliver exogenous cargos from outside to inside of the cell. We found this method allowed successful delivery of dyes, QDs and plasmids with high delivery efficiency and high cell viability for different cancer cells and stem cells. For PI dye uptake into CL1-0 cells, the lower laser fluence (35 mJ cm−2) at 945 nm with low nanoparticle concentrations (50–60 nm-AuPNPs per cell) effectively enhanced delivery efficiency up to 94% with ∼100% cell viability. For pMAX-E2F1 plasmid delivery, we successfully achieved 86% and 73% delivery efficiency, and 96% and 96% cell viability, for CL1-0 and P-19 stem cells, respectively. Moreover, we obtained around 90–95% cell viability three-day after laser exposure with calcein-AM test. This technique is potentially beneficial for biological cell research and therapeutic applications.

Conflicts of interest

The authors declare that there are no conflicts of interest.

Acknowledgements

The authors greatly appreciate the financial support from the DBT/Wellcome Trust India Alliance Fellowship under grant number IA/E/16/1/503062, Frontier Research Center on Fundamental and Applied Sciences of Matters, from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE 108QR001I5), the Ministry of Science and Technology (MOST) Taiwan ROC through National Nanotechnology and Nanoscience Program under contract no. 105-2221-E-007-072-MY3, 105-2321-B-007-003, as well as 108-2638-E-007-001-MY2 and Excellent University Center for Biomedical Technology, NTHU, 2013–2016. We are also thankful to Dr Pallavi Gupta and Ms Kavitha Illath for their experimental help from the Department of Engineering Design, Bio-μ-nano lab at IIT Madras, India.

Notes and references

  1. D. W. Pack, A. S. Hoffman, S. Pun and P. S. Stayton, Nat. Rev. Drug Discovery, 2005, 4, 581 Search PubMed .
  2. D. Luo and M. Saltzman, Nat. Biotechnol., 1999, 18, 33 CrossRef PubMed .
  3. L. H. Yin, S. Q. Fu, T. Nanakorn, F. Garcia-Sanchez, I. Chung, R. Cote, G. Pizzorno, E. Hanania, S. Heimfeld and R. Crystal, Stem Cells, 1998, 16, 247 Search PubMed .
  4. R. M. Schmid, H. Weidenbach, G. F. Draenert, S. Liptay, H. Lührs and G. Adler, Gut, 1997, 41, 549 CrossRef CAS PubMed .
  5. T. S. Santra, P.-C. Wang, H.-Y. Chang and F.-G. Tseng, Appl. Phys. Lett., 2013, 103, 233701 CrossRef .
  6. T. S. Santra and F.-G. Tseng, Micromachines, 2013, 4(3), 333 CrossRef .
  7. K. K. Ewert, A. Ahmad, N. F. Bouxsein, H. M. Evans and C. R. Safinya, Methods Mol. Biol., 2008, 433, 159 CrossRef CAS PubMed .
  8. M. R. Capecchi, Cell, 1980, 22, 479 CrossRef CAS PubMed .
  9. D. H. Fuller, P. Loudon and C. Schmaljohm, Methods, 2006, 40, 86 CrossRef CAS .
  10. S. Ohta, K. Suzuki, Y. Ogino, S. Miyagawa, A. Murashima, D. Matsumaru and G. Yamada, Dev., Growth Differ., 2008, 50(6), 517 Search PubMed .
  11. M. D. Lavigne and D. C. Gorecki, Expert Opin. Emerging Drugs, 2006, 11, 541 CrossRef CAS PubMed .
  12. P. Chakravarty, W. Qian, M. A. El-Sayed and M. R. Prausnitz, Nat. Nanotechnol., 2010, 5, 607 CrossRef CAS PubMed .
  13. R. G. Crystal, Science, 1995, 270, 404 CrossRef CAS PubMed .
  14. Z. C. Hartman, D. M. Appledorn and A. Amalfitano, Virus Res., 2008, 132(1–2), 1 CrossRef CAS PubMed .
  15. D. Luo and W. M. Saltzman, Nat. Biotechnol., 2000, 18(1), 33 CrossRef CAS PubMed .
  16. S. Mehier-Humbert and R. H. Guy, Adv. Drug Delivery Rev., 2005, 57, 733 Search PubMed .
  17. I. Lentacker, I. De Cock, R. Deckers, S. C. De Smedt and C. T. W. Moonen, Adv. Drug Delivery Rev., 2014, 72, 49 CrossRef CAS PubMed .
  18. K. Kim, J. A. Kim, S.-G. Lee and W. G. Lee, Nanoscale, 2012, 4, 5051 RSC .
  19. J. A. Kim, K. Cho, M. S. Shin, W. G. Lee, N. Jung, C. Chung and J. K. Chang, Biosens. Bioelectron., 2008, 23, 1353 CrossRef CAS PubMed .
  20. U. K. Tirlapur and K. Konig, Nature, 2002, 418, 290 Search PubMed .
  21. J. Baumgart, L. Humbert, E. Boulais, R. Lachaine, J. J. Lebrun and M. Meunier, Biomaterials, 2012, 33, 2345 Search PubMed .
  22. A. Uchugonova, K. Konig, R. Bueckle, A. Isemann and G. Tempea, Opt. Express, 2008, 16, 9357 Search PubMed .
  23. D. J. Stevenson, F. J. Gunn-Moore, P. Campbell and K. Dholakia, J. R. Soc. Interface, 2010, 7, 863 CrossRef CAS PubMed .
  24. A. V. Nikolskaya, V. P. Nikolski and I. R. Efimov, Cell Commun. Adhes., 2006, 13, 217 CrossRef CAS PubMed .
  25. H. Schneckenburger, A. Hendinger, R. Sailer, W. S. Strauss and M. J. Schmitt, Biomed. Opt., 2002, 7, 410 CrossRef PubMed .
  26. A. Sengupta, S. C. Kelly, N. Dwivedi, N. Thadhani and M. R. Prausnitz, ACS Nano, 2014, 8(3), 2889 CrossRef CAS PubMed .
  27. Z. G. Li, A. Q. Liu, E. Klaseboer, J. B. Zhang and C. D. Ohl, Lab Chip, 2013, 13, 1144 RSC .
  28. X. Huang, P. K. Jain, I. H. El-Sayed and M. A. El-Sayed, Nanomedicine, 2007, 2(5), 681 CrossRef CAS PubMed .
  29. Z. P. Qin and J. C. Bischof, Chem. Soc. Rev., 2012, 41, 1191 RSC .
  30. J. Chen, B. Wiley, Z.-Y. Li, D. Campbell and F. Saeki, Adv. Mater., 2005, 17, 2255 Search PubMed .
  31. T. S. Santra, F.-G. Tseng and T. K. Barik, J. Nanopharmaceutics Drug Delivery, 2014, 2, 1 CrossRef .
  32. Y. Lu, G. L. Liu, J. Kim, Y. X. Mejia and L. P. Lee, Nano Lett., 2005, 5, 119 CrossRef CAS PubMed .
  33. C.-H. Chou, C.-D. Chen and C. R. C. Wang, J. Phys. Chem. B, 2005, 109, 11135 CrossRef CAS PubMed .
  34. E. Y. Lukianova-Hleb, M. B. G. Mutonga and D. O. Lapotko, ACS Nano, 2012, 6, 10973 CrossRef CAS PubMed .
  35. E. Y. Lukianova-Hleb, X. Y. Ren, J. A. Zasadzinski, X. W. Wu and D. O. Lapotko, Adv. Mater., 2012, 24, 3831 CrossRef CAS PubMed .
  36. E. Y. Lukianova-Hleb, D. S. Wagner, M. K. Brenner and D. O. Lapotko, Biomaterials, 2012, 33, 5441 CrossRef CAS PubMed .
  37. R. Xiong, K. Raemdonck, K. Peynshaert, I. Lentacker, I. D. Cock, J. Demeester, S. C. De Smedt, A. G. Skirtach and K. Braeckmans, ACS Nano, 2014, 8(6), 6288 CrossRef CAS PubMed .
  38. D. Lapotko, E. Lukianova, A. Shnip, G. Zheltov, M. Potapnev, A. Oraevsky, V. Savitskiy and O. Klimovich, Proc. SPIE, 2005, 5697, 82–89 CrossRef CAS .
  39. D. Lapotko, E. Lukianova, M. Potapnev, O. Aleinikova and A. Oraevsky, Proc. SPIE, 2006, 6086, 135–142 CrossRef .
  40. A. M. Smith, M. C. Mancini and S. Nie, Nat. Nanotechnol., 2009, 4, 710–711 CrossRef CAS PubMed .
  41. R. Weissleder, Nat. Biotechnol., 2001, 19, 316–317 CrossRef CAS PubMed .
  42. C. Ungureanu, R. Kroes, W. Petersen, T. A. M. Groothuis, F. Ungureanu, H. Janssen, F. W. B. van Leeuwen, R. P. H. Kooyman, S. Manohar and T. G. van Leeuwen, Nano Lett., 2011, 11, 1887 CrossRef CAS PubMed .
  43. J. Boneberg, J. Koenig-Birk, H.-J. Munzer, P. Leiderer, K. L. Shuford and G. C. Schatz, Appl. Phys. A, 2007, 89, 299 CrossRef CAS .
  44. H.-Y. Hsieh, T.-W. Huang, J.-L. Xiao, C.-S. Yang, C.-C. Chang, C.-C. Chu, L.-W. Lo, S.-H. Wang, P.-C. Wang, C.-C. Chieng, C.-H. Lee and F.-G. Tseng, J. Mater. Chem., 2012, 22, 20918 RSC .
  45. H.-Y. Hsieh, J.-L. Xiao, C.-H. Lee, T.-W. Huang, C.-S. Yang, P.-C. Wang and F.-G. Tseng, J. Phys. Chem. C, 2011, 115, 16258 CrossRef CAS .
  46. B. Pelaz, V. Grazu, A. Ibarra, C. Magen, P. del Pino and J. M. de la Fuente, Langmuir, 2012, 28, 8965 CrossRef CAS PubMed .
  47. G.-Y. Yao, Q.-Lu Liu and Z.-Y. Zhao, Catalysts, 2018, 8(6), 236 CrossRef .
  48. M. Guerrisi and R. Rosei, Phys. Rev. B: Solid State, 1975, 12, 557–563 CrossRef CAS .
  49. L. Liu, P. Li, B. Adisak, S. Ouyang, N. Umezawa, J. Ye, R. Kodiyath, T. Tanabe, G. V. Ramesh, S. Ueda and H. Abe, J. Mater. Chem. A, 2014, 2, 9875–9882 RSC .
  50. B. Nikoobakht and M. A. El-Sayed, Chem. Mater., 2003, 15(10), 1957 CrossRef CAS .
  51. P. B. Johnson and R. W. Christy, Phys. Rev. B: Solid State, 1972, 6, 4370 CrossRef CAS .
  52. G. M. Hale and M. R. Querry, Appl. Opt., 1973, 12, 555 CrossRef CAS PubMed .
  53. R. H. French, K. I. Winey, M. K. Yang and W. Qiu, Aust. J. Chem., 2007, 60, 251–263 CrossRef CAS .
  54. P. K. Jain, K. S. Lee, I. H. El-Sayed and M. A. El-Sayed, J. Phys. Chem. B, 2006, 110, 7238–7248 CrossRef CAS PubMed .
  55. D. Werner and S. Hashimoto, J. Phys. Chem. C, 2011, 115, 5063–5072 CrossRef CAS .
  56. A. Hatef, S. Fortin-Deschênes, E. Boulais, F. Lesage and M. Meunier, Int. J. Heat Mass Transfer, 2015, 89, 866–871 CrossRef CAS .
  57. Y. Wang, Z. Lu and X. Ruana, J. Appl. Phys., 2016, 119, 225109 CrossRef .
  58. G. Chang, F. Sun, J. Duan, Z. Che, X. Wang, J. Wang, M. J. Kim and H. Zhang, Acta Mater., 2018, 160, 235–246 CrossRef CAS .
  59. Y. Zhang, H. L. Zhang, J. H. Wu and X. T. Wang, Scr. Mater., 2011, 65, 1097–1100 CrossRef CAS .
  60. D. Werner, A. Furube, T. Okamoto and S. Hashimoto, J. Phys. Chem. C, 2011, 115, 8503–8512 CrossRef CAS .
  61. J. Lombard, T. Biben and S. Merabia, J. Phys. Chem. C, 2017, 121, 15402–15415 CrossRef CAS .
  62. D. Werner and S. Hashimoto, J. Phys. Chem. C, 2011, 115, 5063–5072 Search PubMed .
  63. J. Shin, M. Song, H. Hafeez, P. J. Jeusraj, D. H. Kim and J. C. Lee, et al. , Org. Electron., 2019, 66, 94–101 Search PubMed .
  64. Z. Starowicz, R. Wojnarowska-Nowak, P. Ozga and E. M. Sheregii, Colloid Polym. Sci., 2018, 296, 1029–1037 CrossRef CAS PubMed .
  65. E. Lukianova-Hleb, Y. Hu, L. Latterini, L. Tarpani, S. Lee, R. A. Drezek, J. H. Hafner and D. O. Lapotko, ACS Nano, 2010, 4(4), 2109–2123 Search PubMed .
  66. T. S. Santra, H.-Y. Chang, P.-C. Wang and F.-G. Tseng, Analyst, 2014, 139, 6249–6258 RSC .
  67. T. S. Santra, C.-W. Chen, H.-Y. Chang and F.-G. Tseng, RSC Adv., 2016, 6, 10979–10986 RSC .
  68. F.-G. Tseng and T. S. Santra, Essentials of Single-Cell Analysis: Concepts, Applications and Future Prospects, 2016 Search PubMed .
  69. T. S. Santra, Microfluidics and Bio-MEMS: Devices and Applications, Jenny Stanford Publisher, 1st edn, 2020 Search PubMed .
  70. T. S. Santra and F.-G. Tseng, Handbook of Single Cell Technologies, Springer Nature, 2020 Search PubMed .
  71. T. Sakai, Y. Takeoka, T. Seki and R. Yoshida, Langmuir, 2007, 23, 8651 CrossRef CAS PubMed .

Footnotes

Electronic supplementary information (ESI) available. See DOI: 10.1039/d0nr01792b
Equal contribution.

This journal is © The Royal Society of Chemistry 2020