Near Infrared-Triggered Liposome Cages for Rapid, Localized Small Molecule Delivery

Photolabile chelating cages or protecting groups need complex chemical syntheses and require UV, visible, or two-photon NIR light to trigger release. Different cages have different solubilities, reaction rates, and energies required for triggering. Here we show that liposomes containing calcium, adenosine triphosphate, or carboxyfluorescein are tethered to plasmon-resonant hollow gold nanoshells (HGN) tuned to absorb light from 650–950 nm. Picosecond pulses of near infrared (NIR) light provided by a two-photon microscope, or by a stand-alone laser during flow through microfluidic channels, trigger contents release with spatial and temporal control. NIR light adsorption heats the HGN, inducing vapor nanobubbles that rupture the liposome, releasing cargo within milliseconds. Any water-soluble molecule can be released at essentially the same rate from the liposome-HGN. By using liposomes of different composition, or HGN of different sizes or shapes with different nanobubble threshold fluences, or irradiating on or off resonance, two different cargoes can be released simultaneously, one before the other, or in a desired ratio. Calcium release from liposome-HGN can be spatially patterned to crosslink alginate gels and trap living cells. Liposome-HGN provide stable, biocompatible isolation of the bioactive compound from its surroundings with minimal interactions with the local environment.

Monodisperse silver nanocrystals are favored by a rapid burst of nucleation of silver crystal nuclei. This burst nucleation is followed by the slow, controlled growth of the nuclei 4,5 .
Poly(vinylpyrrolidone) (PVP) selectively binds to the (100) silver surface, directing the growth of silver nanocrystals with cubic shapes. During growth, Ag atoms are preferentially added to the {111} facets of a single-crystal seed, leading to the formation of sharp cornered cubic nanocrystals.
PVP also stabilizes the silver templates against aggregation. The reaction time determines the size of the silver templates; 30 minutes for 18 nm particles to 180 min for 32 nm particles 4,5 . (Fig.   S1). The solution goes through three color changes during growth: dark reddish brown, reddish green, and bright yellowish green as the edge length of the cubic Ag seeds increased. The reaction was quenched by placing the beaker in a cold-water bath. The reaction product was washed with acetone followed by centrifugation at 4000 rpm for 30 min to remove the remaining Ag precursor and DEG, and then washed with water followed by centrifugation at 13,000 rpm for 10 min (repeated three times) to remove excess PVP. The cubic Ag nanoparticles were redispersed in DI water. The surface plasmon resonance for the silver templates was 420 -430 nm (Fig. S1C).
For silver templates smaller than 20 nm, the oil bath was replaced by electrical resistance tape wrapped around the reaction beaker (Briskheat, Sigma Aldrich). The beaker heater keeps the

Wavelength (nm)
Ag templates HGN Fig. S1. Silver templates of mean size (A) 11 ± 3 nm and (B) 31 ± 3 nm. The HGN size is dictated by the template dimensions as shown in Fig. S2. The size distributions were confirmed by single particle tracking with a Nanosight instrument. (C) The silver template plasmon resonance is centered at 420 nm for all sizes of template. The SPR red-shifts on conversion to HGN depending on the relative amount of gold to silver in the galvanic replacement reaction 4 .

A B C
reaction temperature constant by supplying heat to the solution at a rate such that the reaction temperature of 150 °C is recovered in less than a minute following the nucleation step. The improved temperature control and uniformity promotes single burst nucleation and a monodisperse population with a mean edge length of ~9 nm for 10 minutes of reaction and ~15 nm sized Ag templates for a reaction time of 15 min (Fig. S1). from yellow to blue. The amount of HAuCl4 was adjusted to fine tune the LSPR peak by following the reaction with UV-Vis spectroscopy. (Fig. S2C). On completion, the solution was cooled, silver chloride precipitated out of solution and was separated, and the supernatant containing the gold nanoshells stored at 4°C.

Hollow Gold Nanoshell Synthesis
For a given silver template, increasing the ratio of Au to Ag red-shifts the LSPR peak ( Fig.   S2C), by decreasing the thickness of the shell walls 3-5 because silver is replaced by gold at a 3:1 stoichiometry. For 5 ml of the as-prepared 27 nm cubic Ag templates, adding 300 -600 µl of 1 mM HAuCl4 solution increased the LSPR absorption maximum from 600 -900 nm (Fig. S2C).
The silver templates in Fig. S1 and the HGN in Fig. S2 were prepared for transmission electron microscopy by spreading ~ 2 µl of sample suspension as onto lacey carbon TEM grids (Electron Microscopy Sciences), allowing the solvent to evaporate and imaging with an FEI Technai Sphera G2. Silver template and HGN concentration and size distributions were measured using single particle tracking with a Nanosight NTA 2.3 particle-tracking device. The mean size given by TEM and particle tracking showed good agreement.

Fig. S2: A, B)
Hollow gold nanoshells made from the silver templates in Fig. S1 A, B, respectively by the galvanic replacement of silver by gold. The sizes of the HGN increased 2-4 nm due to the gold shell plating on the silver template. The walls are porous so water is both inside and outside the shell 4 . C) The gold plating and hollow shape red-shifts the spectra to the NIR 4 depending on the gold to silver ratio in the synthesis For a fixed silver template size of 27 nm, adding increasing amounts of 1 mM HAuCl4 red-shifts the LSPR peak from 600 -900 nm, without significantly altering the magnitude of the peak (From Ref. 3). concentrations of liposomes and HGN were determined by particle counting using a Nanosight system and HGN were added to the liposome suspensions at a 10:1 nanoshell to liposome ratio.
Our goal was to provide 1-3 HGN per liposome to ensure that all liposomes would be ruptured by a given light pulse. As shown in Fig. S3 Figure 8. The gel was tilted at 30° to induce flow following irradiation and the gold colored stripes were followed (highlighted by gold lines in Fig. 8) to show that the gel stiffness could be spatially altered by different NIR fluences, which in turn led to different calcium release from the liposomes.

Cell Trapping and Growth in Irradiated Gels
A 2 % (w/v) alginate gel was mixed with PC3 cells 6

Quasistatic Approximation
Mie theory in the quasistatic limit can relate the local surface plasmon resonance (LSPR) frequency for core-shell spheres (See Figure S2C) with particle diameters much smaller than the wavelength of the incident light 2 as a function of the ratio of shell thickness, t, to overall size, R 2 . Water, with a constant, real dielectric function, ; , makes up the nanoshell core as well as the external medium. The gold-silver alloy shell has a frequency dependent dielectric function < ( ).
Eqns. S1-S2 give the absorption cross section, sabs 2 : ?@A = D EF G HIJ K N I O I P /I J I Q I O I P 15I J I Q R (S1) ? = ; (3 − 2 ) + 2 < (S1a) @ = < (3 − ) + ; (S1b) and the shape factor, P: (S1c) ?@A ≈ ( 5 ) \ This shows that sabs ~ xR 2 t, in which x is a frequency-dependent function of the dielectric constants of water and metal, and R 2 t is proportional to the volume of the metal portion of the nanoshell. The results for the gold-silver alloy HGN are shown in Fig. S4.

Electromagnetic simulations
We used Full-field finite-difference time-domain (FDTD) electromagnetic simulations to extract the absorption cross sections 7 as described in Ref. 4. The metal alloy dielectric function was , in which e is the high-frequency dielectric constant, j is the plasma frequency, and Γ is the damping parameter following the data of Peña-Rodriguez et al. 8 for gold-silver alloys. The complex permittivity as a function of silver mole fraction, xAg is Figure S4 shows a comparison between the quasistatic analytical model predictions and those of the full FDTD simulations.
For a fixed outer diameter (or silver template diameter), increasing the ratio of gold to silver during galvanic replacement decreases the shell thickness as the stoichiometry dictates that 3 silver ions are removed per gold atom. This causes the resonance to red-shift to higher wavelengths ( Fig.   Figure S4. Comparisons of the quasistatic (QS) analytical approximation (Eqn. S1 -S3) for spherical core-shell spheres using the properties of pure Au or the 50-50 Au-Ag alloy to the full field electromagnetic simulations for spheres and cuboids. There is a small red-shift and decrease of the magnitude of the cross section for the full field calculation compared to the quasistatic approximation for spheres of the same composition. Changing from a sphere to a cube provides a much larger redshift and a significant increase in the cross section. The alloy blue-shifts both spheres and cubes and decreases the cross section ( S2C) 2,9 . Similarly, for a fixed shell thickness, increasing the particle diameter red-shifts the resonance to higher wavelengths (which also increases the ratio of the inner to outer radii). Cubic shapes with sharp edges are red-shifted relative to spherical shapes for the same diameter and shell thickness (Fig. S4) 10 . Therefore, the LSPR wavelength can be tuned by changing the size of the template particle or the shell thickness, or by controlling the shape of the resulting particle.

Nanobubble Generation and Detection in Flow
Nanobubbles in flow were detected by the scattering of a Helium-Neon probe laser (632.8 nm, 2 mW, Thorlabs, Inc.). The liposome suspensions were pumped through a 0.2 mm ID square, hollow glass capillary of 0.1 mm wall thickness (#8320 Vitro Tubes, VitroCom, Mountain Lakes, NJ) at controlled rates. The pump beam diameter was set to 300 µm and was tilted at 15° so the pump beam would miss the lens used to collect the light from the probe beam, which was aligned normal to the capillary. The pump beam fluence was measured by registering the image of the pump beam on the capillary and measuring the beam diameter at the sample plane with a photodetector/imaging device (Luka, Andor Technology, Northern Ireland). The pulse energy was measured using a pulse energy meter (Ophir Optronics, Ltd., Israel).
The continuous low power probe beam is scattered by the refractive index difference caused by the generation of nanobubbles around the HGN, and the light intensity measured by the photodetector is collected by an oscilloscope (Teledyne LeCroy, Wavesurfer MXs-8). A decrease in the transmitted light intensity is the characteristic signal of nanobubble generation and growth, followed by a rapid rise (100 nsec) in the transmitted intensity as the nanobubbles collapse.
Nanobubbles are transient events, lasting a few hundred nanoseconds. The decrease in transmitted light intensity is due to the collective light scattering from a large number of nanobubbles being generated within the irradiated volume, rather than the scattering signal from single bubbles.