Pulsed laser assisted high-throughput intracellular delivery in hanging drop based three dimensional cancer spheroids

Abstract

Targeted intracellular delivery of biomolecules and therapeutic cargo enables the controlled manipulation of cellular processes. Laser-based optoporation has emerged as a versatile, non-invasive technique that employs light-based transient physical disruption of the cell membrane and achieves high transfection efficiency with low cell damage. Testing of the delivery efficiency of optoporation-based techniques has been conducted on single cells in monolayers, but its applicability in three-dimensional (3D) cell clusters/spheroids has not been explored. Cancer cells grown as 3D tumor spheroids are widely used in anti-cancer drug screening and can be potentially employed for testing delivery efficiency. Towards this goal, we demonstrated the optoporation-based high-throughput intracellular delivery of a model fluorescent cargo (propidium iodide, PI) within 3D SiHa human cervical cancer spheroids. To enable this technique, nano-spiked core–shell gold-coated polystyrene nanoparticles (ns-AuNPs) with a high surface-to-volume ratio were fabricated. ns-AuNPs exhibited high electric field enhancement and highly localized heating at an excitation wavelength of 680 nm. ns-AuNPs were co-incubated with cancer cells within hanging droplets to enable the rapid aggregation and assembly of spheroids. Nanosecond pulsed-laser excitation at the optimized values of laser fluence (45 mJ cm⁻²), pulse frequency (10 Hz), laser exposure time (30 s), and ns-AuNP concentration (5 × 10¹⁰ particles per ml) resulted in the successful delivery of PI dye into cancer cells. This technique ensured high delivery efficiency (89.6 ± 2.8%) while maintaining high cellular viability (97.4 ± 0.4%), thereby validating the applicability of this technique for intracellular delivery. The optoporation-based strategy can enable high-throughput single cell manipulation, is scalable towards larger 3D tissue constructs, and may provide translational benefits for the delivery of anti-cancer therapeutics to tumors.

Introduction

Delivery of biomolecules, foreign genetic material, and other theranostic cargo into cells has been used for various applications including disease treatment, labeling and tracking of cells, and manipulation of fundamental cellular processes, amongst others.^1–5 Intracellular delivery can be broadly classified into (1) viral transfection, (2) chemical treatment and (3) physical disruption techniques. Viral transfection involves certain limitations including poor efficiency, off-target effects, mutagenicity, and cell-specificity. Chemical treatment involves the use of polymers and cationic lipids to disrupt the lipid bilayer of the cell membrane, but has certain limitations including chemical toxicity and cell-specificity.^6–8 Among the physical disruption techniques, the most prominent ones include electroporation, magnetoporation, sonoporation, and microinjection, which involve the creation of ultra-short duration pores in the cell membrane for rapid entry of the desired cargo.^9–14 Each of these techniques is associated with specific advantages and limitations and may be desirable in a case-specific manner depending on the end application. For example, microinjection is highly targeted and efficient for single cell handling, but due to time, cost, and labor limitations, it cannot be scaled for handling larger cell numbers.^15,16 Similarly, electroporation, magnetoporation, and sonoporation provide ready alternatives for high-throughput delivery but several operational problems are encountered including electric and thermal field distortion, mechanical shear force generation, pH variations, cellular shock and damage, and others.⁹

Laser-based photoporation (or optoporation) has emerged as a promising alternative that seeks to address several of the limitations of the existing techniques. Femto-second or nanosecond laser pulses can be used to create low-density plasma adjoining the cell membrane, which leads to the rapid evaporation of water molecules and the photothermal expansion of vapor nano-bubbles. The cavitation of these nano-bubbles creates ultra-short duration, transient pores on the cell membrane which allows the rapid entry of biomolecules and/or other genetic, nano-particulate, and molecular cargo into the cells.^17–19 This process can be further enhanced by employing gold nanoparticles (AuNPs) owing to their high surface plasmon resonance by which the incident optical energy is converted to electrical field energy leading to the efficient generation of heat in a nano-confined space. Changing the nanoparticle surface morphology from a spherical surface to one with surface irregularities can lead to highly localized, area-selective enhancement of the electric field potential and heat generation.²⁰ Thus, by employing surface-modified AuNPs, high transient temperatures can be obtained at lower fluence, incident optical energy can be efficiently converted to mechanical energy of the nano-bubble cavitation, and transient nano-pores can be created on the cell membrane which can be repaired, thereby ensuring low cellular damage.

In the past, optoporation-based delivery of biomolecules has been employed in single cells cultured as two-dimensional (2D) monolayers.^21–24 A study by Patheja et al. has also explored the optoporation method for transferring cytoplasmic vesicles and injected dye molecules in the tumour spheroid in conjugation with an optical tweezer to optically trap the target cell.²⁵ In our previous study, the delivery efficiency was improved through the use of nano-corrugated mushroom-shaped gold-coated polystyrene nanoparticles (nm-AuPNPs) that allowed facile entry of dyes, quantum dots and plasmids into multiple cancer cell types in 2D culture while maintaining high cell viability.¹⁹ However, the intracellular delivery of molecular cargo into complex three-dimensional (3D) multi-cellular aggregates or micro-tissues is yet to be demonstrated. Extending the ability of optoporation-based delivery into 3D tissues is important in the context of drug delivery, labeling and tracking of target cell types within a heterogeneous tissue volume, selective spatio-temporal regulation of cell function and behavior of cell populations within a larger tissue mass.

To demonstrate the applicability of optoporation-based intracellular delivery of molecular cargo into 3D tissue aggregates, we demonstrate the delivery of a model fluorescent dye, propidium iodide (PI), into 3D spheroids of SiHa cervical cancer cells. The biological characteristics of the SiHa cell line have been found to be consistent with those of human squamous cell carcinomas of the cervix, along with an epithelial morphology and phase-microscopy appearance similar to those of other squamous cell carcinoma lines from different sites, e.g., epidermis and tongue and head and neck. These cells were spontaneously aggregated into tumor spheroids of diameter ~1000 μm by culturing cells within hanging droplets. AuNPs were specially modified to display a nano-spiked surface morphology, which we term as ns-AuNPs. The ns-AuNPs were co-incubated with SiHa tumor spheroids within hanging droplets. Upon nano-second laser irradiation at 680 nm wavelength, electric field enhancement and high transient temperature caused by the surface plasmon resonance of ns-AuNPs enabled the formation of transient nano-scale pores on the cell membrane. This was validated using the cell-impermeable dye PI, which provides red fluorescence after entry into the cell cytosol due to optoporation. The schematic for the concept is shown in Fig. 1, where single cells form aggregates within hanging droplets (Fig. 1(a)) which are then exposed to laser irradiation for optoporation-based delivery (Fig. 1(b)). The delivery efficiency of PI and cell viability were quantified as 89.6 ± 2.8% and 97.4 ± 0.4% respectively, which demonstrates the potential of the optoporation technique for intracellular delivery of biomolecules and therapeutic cargo for future applications in nanomedicine.

Fig. 1. Schematic of the optoporation-mediated delivery in 3D SiHa tumor spheroids.

(a) SiHa cancer cells (green) and ns-AuNPs were pipetted on the lid of a Petri dish as droplets which was then inverted and allowed to co-incubate, thereby forming 3D spheroids at the apex of the hanging droplets. (b) The 3D spheroids were exposed to a pulsed infrared laser, resulting in the generation of photothermal cavitation bubbles, transient disruption of the plasma membrane and rapid entry of biomolecules into the cells.

Results and discussion

Fabrication and characterization of ns-AuNPs

The schematic for the fabrication process of ns-AuNPs is shown in Fig. 2 and described in the Methods section. Briefly, 200 nm sized gold-coated polystyrene nanoparticles (ns-AuNPs) with nano-spiked morphology were fabricated via reactive ion etching and electron beam evaporation (Fig. 2).

Fig. 2. Fabrication of nano-spiked gold nanoparticles (ns-AuNPs).

(a) A silicon wafer is oxygen plasma treated to render the top surface hydrophilic. A PDMS layer with a punched hole is bonded on top of the hydrophilic substrate to form a well. Polystyrene beads (510 nm in size) in a solution of ethanol and water are added to the well and allowed to pack as a monolayer in a hexagonal pattern. After removal and drying of the solution, the beads are etched under oxygen plasma to create a nano-spiked morphology and further coated with gold under electron beam evaporation to form ns-AuNPs (200 nm in diameter). (b–d) Scanning electron micrographs of representative polystyrene beads and ns-AuNPs with hexagonal conformation (yellow dotted line). Scale bars: (b and c) 200 nm, (d) 500 nm.

Fig. 3(a) shows the ultrastructural morphology of the fabricated ns-AuNPs with an average size of ~200 nm. Surface morphology plays a vital role in governing the intensity of the plasmon field and the optical extinction spectra of the nanoparticles. Extremely high electromagnetic fields can be generated at the surface of metals with corrugated structures.^19,26 Plasmonic coupling between two interacting nanoparticles can even intensify the local electric fields. The local spots exhibiting highly intense electric fields are known as “hot spots”. The electric field in the “hot spot” region depends on the direction of polarization of the electromagnetic radiation relative to the interparticle axis, reaching the maximum when the polarization vector is parallel to this axis.²⁷ Fig. 3(b) shows the UV-visible absorption spectra of ns-AuNPs showing two absorption bands peaked around 532 nm and 680 nm. The peaks at 532 nm is attributed to the interband electronic transitions of Au.^28–30 The peak at 680 nm arises due to surface plasmon resonance at the gold surface.

Fig. 3. Characterization of nano-spiked gold nanoparticles (ns-AuNPs).

(a) Scanning electron micrographs of ns-AuNPs with high surface-area-to-volume ratio (~200 nm in diameter) post-fabrication on the silicon substrate. Scale bar: 200 nm. (b) UV-visible absorption spectra of ns-AuNPs reveal two major peaks in the wavelength range of 500–1000 nm.

Simulation of surface plasmon resonance effect of ns-AuNPs

Plasmon-induced cavitation nanobubbles (PNBs) are generated when a nanosecond laser pulse irradiates metal nanoparticles in a water-based medium at specific wavelengths. The process of generation and collapse of these PNBs causes a jet flow, which is sufficiently strong to deform cell membranes in the vicinity and create transient membrane pores for intracellular delivery. However, in order for PNBs to be generated, a specific critical temperature (spinodal temperature), which is around 550 K for the Au/water interface, needs to be attained. To confirm the generation of PNBs in our experiment, we have simulated the electric field enhancements and associated temperature rise of a model ns-AuNP and the surrounding water by a finite element method using the Comsol Multiphysics software. The model ns-AuNP with an average diameter of ~200 nm contains a polystyrene core with a corrugated surface and gold as a shell with an average thickness of ~30 nm coated on the core. The values of the wavelength-dependent complex refractive index (n + ik) are taken from ref. 31 for Au and ref. 32 for water medium. The complex refractive index of polystyrene was considered to be n + ik = 1.59 + i0.02.³³

The spatial distribution of field enhancements is shown in Fig. 4(a) and (b). The double-sided arrows show the polarization of the electric field. Fig. 4(a) shows the field enhancement across the cross-section in the middle of the nanoparticle with light propagating perpendicular to the paper surface, as shown by the propagation vector, $\overrightarrow{k}$. Fig. 4(b) shows the field enhancement when light is propagating along the surface of the paper. These results clearly show that the enhancements occur at the tips as well as at the inter-tip regions depending on the polarization and the propagation directions. These enhanced regions act as hotspots and increase the local temperature of the surrounding medium.

Fig. 4. Comsol simulation of surface plasmon resonance of nS-AuNPs.

(a) Enhancement of the electric field across the middle cross-section of a model ns-AuNP (200 nm in diameter) when laser light propagates perpendicular to the plane of view. The white double sided arrow denotes the polarization of the electric field (E) and the white cross denotes the direction of the propagation vector ($\overrightarrow{k}$). (b) Electric field enhancement across the middle cross-section of the model ns-AuNP when laser light propagates parallel to the plane of view. The enhancements are localized both at the apex (a) of the nano-spikes and ridges (b) between the nano-spikes on the ns-AuNP surface. (c) Simulation of the temporal rise and decay in electron (T _e) and lattice temperatures (T _l) of the gold surface and water temperature (T _w) upon laser irradiation (indicated by delay time). (d) Simulated spatial distribution of temperature around a model ns-AuNP at peak transience demonstrating the T _l profile at the gold interface, the T _w profile at the water interface, and the T _c profile (core temperature) at the ns-AuNP core.

The two-temperature model is employed to obtain the temporal and spatial distribution of the temperature using the rate equations³⁴ given below:

$$C_{\text{e}}\left(T_{\text{e}}\right)\frac{\partial T_{\text{e}}}{\partial t}=\nabla \left(k_{\text{e}}\nabla T_{\text{e}}\right)-g\left(T_{\text{e}}-T_1\right)+Q\left(t\right)$$ (1)

$$\begin{array}{ll}{C}_{\text{l}}\frac{\partial T_{\text{l}}}{\partial t}=\hfill & \nabla \left(k_{\text{l}}\nabla T_{\text{l}}\right)+g\left(T_{\text{e}}-T_{\text{l}}\right)-g_{\text{w}}\left(T_{\text{l}}-T_{\text{w}}\right)\hfill \\ \hfill & -g_{\text{c}}\left(T_{\text{l}}-T_{\text{c}}\right)\hfill \end{array}$$ (2)

$${\rho }_{\text{w}}{C}_{\text{w}}\frac{\partial T_{\text{w}}}{\partial t}=\nabla \left(k_{\text{w}}\nabla T_{\text{w}}\right)+g_{\text{w}}\left(T_{\text{l}}-T_{\text{w}}\right)$$ (3)

$${\rho }_{\text{c}}{C}_{\text{c}}\frac{\partial T_{\text{s}}}{\partial t}=\nabla \left(k_{\text{c}}\nabla T_{\text{c}}\right)+g_{\text{c}}\left(T_{\text{l}}-T_{\text{c}}\right)$$ (4)

where T _e and T _l are the electron and lattice temperatures of Au respectively, T _w is the water temperature, and T _c is the core temperature. The electron heat capacity for Au is given as C _e(T _e) = γT _e, with γ = 70 J m⁻³ K⁻²,³⁴ the gold lattice heat capacity C ₁ = 3 × 10⁶ J m⁻³ K⁻¹,³⁴ the electron thermal conductivity k _e = 300 W m⁻¹ K⁻¹,³⁴ the lattice thermal conductivity k _l ~ 2.7 W m⁻¹ K^−1,35 the Au electron–phonon coupling coefficient g = 2.0 × 10¹⁶ W m⁻³ K⁻¹,³⁴ the water heat capacity C _w = 4182 J m⁻³ K⁻¹,³⁴ and the water thermal conductivity k _w = 0.6 W m⁻¹ K⁻¹.³⁴ The term g _w(T _l – T _w) in eqn (1)–(3) represents the boundary interface heat exchange between the gold lattice and the surrounding water, where the corresponding thermal boundary conductance g _w = 105 × 10⁶ W m⁻² K⁻¹.³⁴ The term g _c(T _l – T _c) represents the boundary interface heat exchange between the core and Au, where the interface thermal conductance g_c = 508 × 10⁶ W m⁻² K⁻¹ is calculated following eqn (5) in ref. 36 and 37. In eqn (1), the source term Q(t) is defined by the transient resistive heating due to single 10 ns Gaussian laser pulse excitation with a fluence of 45 mJ cm⁻². The above equations do not include any phase change phenomena due to the rise of the temperature.

The temporal rise and decay of T _e, T _l, and T _w at the fluence of 45 mJ cm⁻² are shown in Fig. 4(c) at temporal positions of 32.5 ns (the moment when T _e reaches its maximum value). Note that T _e and T _l rise synchronously with time, as the exchange of thermal energy happens in sub-picosecond time scale from the electron to the lattice of Au. T _l reaches its maximum value of ~2.5 × 10⁴ K which is far higher than the melting temperature (~1337 K) and the boiling temperature (~3129 K) of bulk gold.^38,39 Therefore, the simulation suggests that the first laser pulse can induce melting and deformation of the gold coating with possible generation of smaller particles.⁴⁰ Thus the subsequent pulses will act on the deformed shape or smaller sized nanoparticles. It is clear from Fig. 4(c) that T _w rises above the spinodal temperature of 550 K, before T _l reached the boiling point. Therefore, the experimental conditions are suitable for generating cavitation bubbles and a jet flow that can cause membrane deformation followed by the formation of transient hydrophilic pores at the cell membrane to facilitate intracellular delivery. Fig. 4(d) shows the spatial distribution of temperatures T _l, T _w, and T _c which reveals that the temperature variation is a local event and is minimal after a certain distance from the core.

Formation of 3D tumor spheroids

One of the prominent methods of 3D cell culture is the use of cellular spheroids which can be formed spontaneously in hanging droplets or through forced centrifugal aggregation. The hanging droplet method has been used extensively as a facile, high-throughput strategy that involves pipetting a small volume (~20 μL) of cell suspension on the lid of a Petri dish, inverting the lid and allowing the cells to aggregate at the apex of the droplet over time due to gravity. The spheroids so formed can be harvested and used for further downstream experiments. 3D tumor spheroids have become the ‘gold standard’ in the pharmaceutical industry for high-throughput drug-testing applications owing to their facile generation and rapid assay read-outs. Additionally, tumor spheroids also recapitulate certain aspects of the complex tumor microenvironment (including hypoxia, nutrient gradient, differences in proliferation, metabolism, and chemosensitivity amongst others) that are not achievable in standard 2D cultured monolayers.

Although optoporation-based delivery has been demonstrated in 2D cultured cancer cells, the ability to deliver molecular cargo in 3D cancer cell aggregates has not been explored before. Hence, 3D tumor spheroids were fabricated using the hanging droplet method for further testing of the delivery efficacy and safety of the optoporation-based strategy. Fig. 5(a) shows the hanging droplets on the Petri dish lid with cancer spheroids aggregated within them. Fig. 5(b) and (c) show brightfield images of the tumor spheroids on day 1 (with lower cell density) and day 4 (with higher cell density) respectively.

Fig. 5. Formation of hanging droplet tumor spheroids.

(a) 20 μL droplets of cell culture media containing SiHa cervical cancer cells are pipetted on the lid of a Petri dish and allowed to incubate for 4 days in culture. Cells aggregate spontaneously at the apex of the droplet due to gravity to form spheroids (denoted by white arrowheads). Scale bar: 20 mm. (b) Bright field image of a tumor spheroid on day 1 and (c) day 4 of incubation. Due to proliferation, the spheroids become more dense, tightly packed, and larger in size within the droplet. Scale bar: 500 μm.

Optoporation-based delivery in 3D tumor spheroids

Although intracellular delivery using nano-spiked morphology of nanoparticles has been studied in the past,¹⁹ the ability to perform high throughput intracellular delivery on tissue-mimicking cellular aggregates is demonstrated here (Fig. 6). The 3D tumor spheroids were incubated with ns-AuNPs to integrate them into the cell mass and ensure close contact of the ns-AuNPs with the cancer cells. After the formation of tumor spheroids with ns-AuNPs within the hanging droplets, propidium iodide (PI) was added to the droplets and individual spheroids were exposed to an infrared pulsed laser.

Fig. 6. Demonstration of optoporation-mediated delivery in 3D SiHa tumor spheroids.

(a and e) Bright field images of tumor spheroids acquired post laser irradiation. (b and f ) Corresponding fluorescence images of tumor spheroids demonstrating successful delivery of propidium iodide (PI) dye (red) after optoporation. (c and g) Corresponding fluorescence images of tumor spheroids demonstrating relatively high cell viability as visualized using calcein AM dye (green). (d and h) Merged images of red and green fluorescence demonstrating high delivery efficiency and high viability within tumor spheroids. Scale bar: 200 μm.

In this study, the cells were exposed at 680 nm wavelength of pulse laser light with 9 mJ laser energy and 5 ns pulse width at 10 Hz pulsing frequency for 30 seconds. Cell impermeable PI dye was used as a model cargo for intracellular delivery within the spheroids in a hanging droplet culture. After optoporation experiments were conducted, the cells were also stained with cell permeable calcein AM dye to assess cellular viability. Due to the pulsed laser interaction with ns-AuNPs, high electromagnetic field enhancement occurred at ns-AuNP surfaces, resulting in local heat generation and high temperatures, and the formation of cavitation nanobubbles in the vicinity of cell membrane–nanoparticle interfaces in the tumor spheroids. The nanobubbles rapidly grew, coalesced, and collapsed to prompt explosion resulting in intense fluid flow at the ns-AuNP–cell membrane interfaces.⁴¹ The strong fluid flow disrupted the plasma membrane to form transient pores enabling the entry of PI dye into the cells. Fig. 6 shows images of different tumor spheroids post optoporation after the successful delivery of PI dye. The PI stained cells in different spheroids are indicated by red color (Fig. 6(b) and (f )). After PI delivery, cell permeable calcein AM was added to the spheroids. Calcein AM is hydrolysed inside cells by intracellular esterases^42,43 and produces green fluorescence indicating live cells (Fig. 6(c) and (g)). The merged images for different spheroids are shown in Fig. 6(d) and (h).

Earlier, multiple studies have explored the light-mediated surface plasmon resonance phenomenon in gold nanoparticles (GNPs) for drug delivery applications in various forms, i.e., immobilized GNPs and aggregates in water dispersion.^44,45 Upon 532 nm visible light irradiation, the GNPs converted the absorbed light energy into heat to generate explosive vapor nanobubbles (VNBs), promoting the delivery of the fluorescent dye (calcein AM) to attached HEK293T cells.⁴⁶ The plasmonic nanostructured substrate with gold-coated tipless pyramid arrays, under NIR irradiation, produced heat to create nanobubbles and a pressure wave, thereby increasing the permeability of the cell membrane transiently and thus facilitating the delivery of fluorescent molecules (calcein) in HeLa cells with a delivery efficiency of 80%.⁴⁷ In another study, gold-coated thermoplasmonic substrate achieved delivery of FITC-dextran molecules with different molecular weights (0.6–2000 kDa) into HeLa CCL-2 cells with high delivery efficiency (up to 95%) and high cell viability (up to 98%).⁴⁸ Furthermore, gold nanoparticle layers (GNPLs) exhibited a more efficient light-to-heat conversion owing to their unique micro-nanotopography. However, the transfection efficiencies varied from ~100% for GFP in easy-to-transfect HeLa cells to 53% for mEFs and 9% for HUVECs under NIR irradiation, possibly due to the degradation of naked pDNA by endonucleases.⁴⁹ It is interesting to notice that all these past systems are based on individual cells in normal adherent cell culture platforms. The present study focused on 3D cellular aggregates in the form of spheroids in a hanging droplet culture.

Quantification of optoporation-based delivery efficiency and safety

Fig. 7 shows the confocal images of spheroid post-delivery of PI (Fig. 7(a and d)), staining of calcein AM (Fig. 7(b and e)) and merged images (Fig. 7(c and f )). The step size was ~1 μm between each slice. The quantification of fluorescence images revealed a high delivery efficiency of PI (89.6 ± 2.8%), high cell viability (97.4 ± 0.4%) and high level of co-staining (87.0 ± 3.2%) (Fig. 7(g)).

Fig. 7.

Image-based quantification of optoporation-mediated delivery efficiency and cell viability. Representative average intensity z-projections from 3D confocal image stacks of SiHa tumor spheroids labelled with (a and d) propidium iodide (PI) dye (red) and (b and e) calcein AM (green) after laser irradiation. (c and f ) Merged images of average-z-intensity projections of PI and calcein AM stained cells. Scale bar: (a–c) 100 μm, (d–f) 50 μm. (g) Quantification of confocal images reveals high PI dye delivery (89.6 ± 2.8%), high cell viability (97.4 ± 0.4%) and high co-staining, i.e. successful live cell delivery efficiency (87.0 ± 3.2%) post optoporation. Values indicate average ± standard deviation. n = 3 representative image locations in 3 independent spheroids.

Since the experiments were performed on 3D spheroids in hanging droplets, it is necessary to assess dye intake in each cell of spheroid post laser exposure. Therefore, the split stack images are shown in Fig. 8(a and b) with red and green color respectively, which represent successful high-throughput intra-cellular delivery in live cells. Fig. 8(c) shows the resulting Z-axis intensity profile of the spheroid after PI dye delivery and calcein AM staining. Fluorescence intensity on a per cell basis for both PI and calcein AM proves uniform delivery and uniform cell viability throughout the depth of the spheroid. Measurements were made on the different slices (n = 10 random cells in each slice for each channel). The results show the efficacy of the proposed concept for the delivery of molecules into a dense cellular mass. It also substantiates the efficacy of the proposed concept for the delivery of drugs, genetic materials, proteins, and other biomolecules into a dense cellular mass i.e., healthy or diseased tissue models or in vivo via nanoparticle-mediated laser exposure.

Fig. 8.

Quantification of uniformity of delivery and viability as a function of the tumor spheroid depth. Montage of representative confocal z-stack slices of (a) propidium iodide (PI, red) and (b) calcein AM (green) in SiHa tumor spheroids. Scale bar: 100 μm, slice numbers are indicated. (c) Quantification of the fluorescence intensity of PI dye and calcein AM on a per cell basis indicates uniform delivery and uniform viability throughout the depth of the spheroid. Values indicate average ± standard deviation. No significant difference in the PI dye or calcein AM fluorescence intensity was found between the different slices. n = 10 random cells in each slice for each channel.

Further validation of the optoporation-based delivery efficiency and safety was analysed by flow cytometry. The average values from the studies conducted (n = 3, where 20 hanging drop cultures were taken each time) showed 98.7 ± 0.2% co-staining of PI dye and calcein AM for 3D tumor spheroids. The sample count was set for 10 000 counts each time. Fig. 9 shows the scatter plot and distribution plots of the calcein AM positive and PI dye positive events, validating the dual staining along with high optoporation efficiency on the tumor spheroids. Hence, the proposed nanoparticle mediated laser optoporation platform presents a facile method of high-throughput intracellular cargo delivery with high efficiency and high cell viability.

Fig. 9. Flow cytometry-based quantification of optoporation-mediated delivery efficiency and cell viability.

(a) Scatter graph of SiHa cells from tumor spheroids post optoporation with propidium iodide (PI) dye and calcein AM reveals 98.7 ± 0.2% dual staining indicating high delivery efficiency while maintaining cell viability. (b and c) Representative individual cell counts of PI positive and calcein AM positive cells under respective channels.

Experimental

Fabrication of ns-AuNPs

Initially, a 4-inch silicon (Si) substrate was treated with oxygen plasma to render the substrate hydrophilic in nature. A polydimethylsiloxane (PDMS) well was placed on top and bonded to the Si wafer. 510 nm polystyrene beads (PSBs), cleaned with DI water and gently resuspended in ethanol–water solution, were introduced on the wafer surface within the PDMS well. Diluted PSBs self-assembled on the wafer surface to form a monolayer. After completion of the self-assembly process, excess water was removed and the PSBs were left for air-drying at room temperature. A densely packed monolayer of PSBs with a hexagonal patterned arrangement (Fig. 2(d)) was obtained. The monolayer was treated with oxygen plasma to form nano-spiked morphology on the PSB surface via reactive ion etching (RIE) (Fig. 2(c)). This was followed by deposition of a gold (Au) layer on the PSB surface by electron beam evaporation to form the final ns-AuNPs (Fig. 2(b)) with an average diameter of 200 nm. The ns-AuNPs were released from the surface through sonication for downstream use with 3D spheroids.

Characterization of nanoparticles

The morphology of fabricated nanoparticles was assessed using scanning electron microscopy (SEM-Hitachi TM-1000, version = 03-02). The absorbance peak was traced using a UV-spectrophotometer (Shimadzu UV-1800 UV-Vis spectrophotometer) in the wavelength range of 500–1000 nm using water as the blank sample.

Nanoparticle simulation

The electric field enhancement was simulated using the Comsol Multiphysics software by the finite element method. A structured nanoparticle was designed using Comsol Multiphysics in 3D and then it was simulated for 680 nm wavelength using electromagnetic waves, with the frequency domain (emw) in the radio frequency module.

Cell culture and tumor spheroid formation

Hanging droplet spheroids were prepared using SiHa human cervical cancer cells. The SiHa cell line was procured from the National Centre for Cell Science (NCCS), Pune, India. The cultured cells were trypsinized and centrifuged, the supernatant was removed, and cell pellets were resuspended in the cell culture medium (DMEM-F12 (Invitrogen) supplemented with 1% MEM non-essential amino acids (Gibco), 5% fetal bovine serum and 1% penicillin–streptomycin. Small aliquots of the cell suspension (20 μl per drop) were pipetted on the lid of a Petri dish which was then inverted to produce the hanging droplets. The dishes were filled with PBS in order to maintain the humidity inside the plate and prevent drops from evaporating. The hanging drop cultures were maintained in a humidified incubator, with 5% CO₂, at 37 °C temperature. The media for the spheroids was replaced every three days.

Nanosecond pulsed laser mediated optoporation

The spheroids were incubated with ns-AuNPs to facilitate optoporation based high throughput intracellular delivery. The ns-AuNPs were mixed in cell culture medium at a concentration of 5 × 10¹⁰ particles per ml of media. The cells were exposed at 680 nm wavelength with 9 mJ energy for 30 seconds. Due to the pulse laser exposure on the hanging drop with ns-AuNPs, the nanoparticles generate strong surface plasmon resonance inducing temperature rise. As a result, plasmonic nanobubbles can be generated and they can rapidly grow, coalesce, and collapse, and induce a strong fluid flow on the cell membrane surface. Thus cell membranes can deform and create transient hydrophilic nanopores and delivery molecules from the outside to inside of the cell.¹⁹ In this experiment, cell impermeable propidium iodide (PI) dye was used as cargo to be delivered inside spheroids in the hanging drop culture. After the photoporation experiments are conducted, cell permeable calcein AM dye was introduced to conclude the viability of the cells.

Confocal microscopy and flow cytometry

The spheroids were stained with calcein AM after nanoparticle-mediated optoporation based PI dye delivery. The treated spheroids were washed with ice-cold PBS and fixed with 4% PFA to be visualised by confocal microscopy. Immunofluorescence was visualised with an LSM780 confocal microscope (Carl Zeiss). ImageJ software (Version 1.52n, NIH) was utilised to reconstruct the serial Z-stack images acquired via confocal microscopy.

For flow cytometry measurement, the PI–calcein AM dual stained cells, PI stained, calcein AM stained, and unstained spheroid cells were captured and isolated. The spheroids were dissociated by incubation with trypsin–EDTA (Thermo Fisher Scientific) for 10 min at 37 °C, with gentle pipetting every 2 min to aid dissociation. The cells were gently pipetted until a single cell population was achieved. Then all the segregated spheroid cell populations were sorted using fluorescence activated cell sorting (FACS) with an Aria Fusion flow cytometer (BD Biosciences, CA, USA), and the data were analyzed using the BD CellQuest Pro software (BD Biosciences, CA, USA).

Conclusions

The study represents an improved method for drug delivery and screening via optoporation on spheroids formed via the hanging droplet technique.⁵⁰ The optimization of laser exposure with nanostructured metallic nanoparticles helps to deliver sufficient intensity to the nano-localized area to drive higher optoporation efficiency. The average laser power that can be used is bound by the thermal damage threshold of the tissue, limiting the utility of optoporation approaches in vivo. The parameters can be optimized for in vitro studies in 3D tumor spheroids which mimic to a certain degree the in vivo microenvironment. Tumor spheroids can also be used to culture cells isolated from biopsy of the cancerous tissues. Therefore, personalized treatment can be optimized on this particular type of culture, and drug dosages and phototherapy procedures can be analysed prior to application in vivo. As a result, the proposed process is useful for a wide range of cellular therapy and diagnostic applications.

Supplementary Material

Supplementary

Electronic supplementary information (ESI) available. See DOI: 10.1039/d0an02432e