Gold Nanoantenna-Mediated Photothermal Drug Delivery from Thermosensitive Liposomes in Breast Cancer.

In this work, we demonstrate controlled drug delivery from low-temperature-sensitive liposomes (LTSLs) mediated by photothermal heating from multibranched gold nanoantennas (MGNs) in triple-negative breast cancer (TNBC) cells in vitro. The unique geometry of MGNs enables the generation of mild hyperthermia (∼42 °C) by converting near-infrared light to heat and effectively delivering doxorubicin (DOX) from the LTSLs in breast cancer cells. We confirmed the cellular uptake of MGNs by using both fluorescence confocal Z-stack imaging and transmission electron microscopy (TEM) imaging. We performed a cellular viability assay and live/dead cell fluorescence imaging of the combined therapeutic effects of MGNs with DOX-loaded LTSLs (DOX-LTSLs) and compared them with free DOX and DOX-loaded non-temperature-sensitive liposomes (DOX-NTSLs). Imaging of fluorescent live/dead cell indicators and MTT assay outcomes both demonstrated significant decreases in cellular viability when cells were treated with the combination therapy. Because of the high phase-transition temperature of NTSLs, no drug delivery was observed from the DOX-NTSLs. Notably, even at a low DOX concentration of 0.5 μg/mL, the combination treatment resulted in a higher (33%) cell death relative to free DOX (17% cell death). The results of our work demonstrate that the synergistic therapeutic effect of photothermal hyperthermia of MGNs with drug delivery from the LTSLs can successfully eradicate aggressive breast cancer cells with higher efficacy than free DOX by providing a controlled light-activated approach and minimizing off-target toxicity.

Introduction

Nanostructures have been used as carriers for transporting cargo where drug delivery is controlled endogenously by biological cues, such as pH or ions,[1−4] or exogenously by plasmonic photothermal materials with near-infrared-light-triggered release. Plasmonic gold nanostructure-mediated drug delivery has been extensively studied in the past decade to combat cancer and other inflammatory diseases.[5−11] There are several advantages to exploiting the light-to-heat conversion abilities of gold nanostructures for exogenous control of drug delivery. First, by tuning the laser flux density and nanostructure concentration, the light-triggered approach generates mild hyperthermia (40–43 °C) that is sufficient for drug release with minimal damage to healthy tissue.[12−14] Second, light activation enables controlled drug delivery at the tumor site while minimizing off-target toxicity. Third, photothermal hyperthermia is highly localized and noninvasive, thus eliminating the need for whole-body heating or invasive heating probes used in current clinical hyperthermia. Finally, hyperthermia improves vascular permeability and increases blood perfusion in the hypoxic tumor areas, thereby enhancing drug uptake and therapeutic efficacy.[15−17] For example, plasmonic nanostructures have been combined with thermoresponsive polymers for controlled delivery.[18,19] In this work, we demonstrate the use of multibranched gold nanoantennas (MGNs) as photothermal actuators to induce delivery of the anticancer drug doxorubicin (DOX) from low-temperature-sensitive liposomes (LTSLs). The intense photothermal properties of MGNs are attributed to their unique geometry where each spherical core behaves as an antenna absorbing near-infrared light and the protrusions act as emitters localizing the absorbed light at the tips, thus efficiently converting light to heat.[20,21] The 50–60 nm size of MGNs is ideal for these studies, enabling rapid endocytosis and accumulation in cells.[22−25] Further, their straightforward synthesis in aqueous media with a nontoxic ligand, 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), improves their biocompatibility for cancer treatments.

LTSLs, currently under phase III clinical trial (Thermodox, Celsion Inc.), are ideal drug delivery vehicles because of their low phase-transition temperature, Tm, of ∼41.7 °C.[26,27] This enables drug release at mild hyperthermia, minimizing unnecessary heating of healthy tissues. Their size can be controlled to ∼100 nm, which favors particle accumulation in the tumor microenvironment through the enhanced permeability and retention (EPR) effect.[23,28] Further, they have high drug-loading efficiency and minimal drug leakage at physiological temperature, which can reduce the toxic side effects of chemotherapy.[29,30] LTSLs have been shown to deliver 25 times more DOX into tumors than an IV infusion of unencapsulated DOX and 5 times more DOX than Doxil (commercial non-temperature-sensitive liposomal formulation of DOX).[26] For enhanced utility of LTSLs, improved hyperthermia strategies that are noninvasive and administer heat uniformly throughout the cancerous tissue are necessary. However, current clinical approaches to induce local hyperthermia can be invasive, which requires a heating probe directly in contact with solid tumors (e.g., radiofrequency ablation) or can give rise to heterogeneous tumor temperatures (e.g., ultrasonic energy), resulting in unpredictable drug release and toxicity to healthy tissues.[31−34]

Here, we demonstrate that MGN-mediated photothermal heating triggered by near-infrared light enables DOX delivery from LTSLs with high efficiency. Co-delivering liposomes and nanoantennas has several advantages over encapsulating nanoantennas within the liposomes. First, co-delivery allows loading a higher amount of drug into the liposomes. Second, by co-delivering, the size of the liposomes can be controlled to ∼100 nm, increasing their accumulation in cancer cells. Third, co-delivery also enables more nanostructures to localize in cells for enhanced photothermal transduction and drug delivery from liposomes. Finally, nanostructure encapsulation in the liposomes is typically achieved by random mixing of the two components, which often results in unsuccessful encapsulation and compromises the overall functionality of the hybrid structure. Therefore, by co-delivering MGNs and LTSLs, we observe that the MGN-mediated release of DOX from LTSLs far surpasses that from non-thermal-sensitive liposomes (NTSLs) under similar conditions because the phase-transition temperature of NTSLs is above the clinically relevant hyperthermia range (40–43 °C). Furthermore, our work shows that the combination treatment of MGN-mediated photothermal hyperthermia with LTSL-delivered chemotherapy has a higher therapeutic efficacy in triple-negative breast cancer (TNBC) than with free DOX or photothermal therapy with MGNs alone. TNBC is highly aggressive and is known to develop multidrug resistance (MDR), limiting the successful usage of many chemotherapy drugs, including DOX.[35−37] Photothermal hyperthermia has been shown to reverse resistance to several anticancer drugs,[38,39] which is particularly useful because MDR cells do not exhibit cross-resistance to hyperthermia.[40−42] We envision that a combination treatment with LTSLs and plasmonic nanostructures will ultimately provide a clinically relevant platform for noninvasive drug delivery to treat TNBC.

Results and Discussion

We synthesized MGNs by following our previously described procedure using a one-step seedless growth mechanism.[21] In this synthesis, the gold precursor Au3+ is reduced in the presence of 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), which plays the role of both a reducing and a capping agent. HEPES is a widely used biological buffer that promotes the overall biocompatibility of MGNs. The weak adsorption of HEPES along the ⟨111⟩ planes enables kinetically driven protrusion growth along that direction, where the core-to-protrusion ratio is controlled by [HEPES]. By using 270 mM HEPES, the size of MGNs (tip-to-tip distance) was controlled to 56 ± 4 nm (Figure a), which is ideal for therapeutic applications. This size regime tunes the plasmon resonance of MGNs to the near-infrared biological window (700–980 nm) where deep tissue penetration can be achieved with the incident light.[43] The MGNs have three distinct plasmon resonances (Figure b): a bonding plasmon mode at 800 nm, an antibonding plasmon mode at 550 nm, and a shoulder peak at 1100 nm. The antibonding mode is primarily contributed by the core plasmons, whereas the bonding plasmon resonance is predominantly composed of MGN protrusions but with a finite contribution of the core plasmons.[21,44] The shoulder peak arises from the hybridization of the plasmons of multiple protrusions of an MGN.[45] The MGNs were further coated with polyethylene glycol (PEG) to enhance biocompatibility, reduce cytotoxicity, and provide steric stabilization.[46−50] Successful PEG coating was confirmed by detecting a clear 10 nm shift in the plasmon resonance (Figure b) because of the higher refractive index of PEG relative to water. A decrease in the surface charge of MGNs is also indicative of successful PEG coating. The surface charge of the MGNs before coating was −35.3 mV, which was reduced to −2.9 mV after PEG coating. A near-neutral surface charge on nanoparticles is desirable to reduce nonspecific binding of MGNs to the cell surface.[22,51−54]

Figure 1

Characterization of PEG-coated MGNs. (a) TEM micrograph of MGNs. (b) Extinction spectra of MGNs before and after coating with PEG showing near-infrared resonance. Photothermal characteristics of PEG-coated MGNs showing (c) temperature at the laser illumination spot at 5.5 W/cm2 achieved by varying the concentrations of PEG-coated MGNs. (d) The photothermal steady-state temperature of PEG-coated MGNs after 10 min of laser irradiation at 4, 5.5, and 7 W/cm2 at varying concentrations, showing photothermal tunability with particle density and laser flux.

The light-to-heat conversion ability of MGNs was studied to evaluate their photothermal efficiency and efficacy in drug delivery from LTSLs. The photothermal behavior of MGNs in cellular media was evaluated at four different concentrations (80, 110, 140, and 170 μg/mL) by illuminating with an 808 nm laser at 5.5 W/cm2 (Figure c) and measuring the media temperature over time with a thermal camera. With 170 μg/mL of PEG-coated MGNs, the temperature within the laser spot reached 57 °C within a minute of illumination. Varying the laser flux (4, 5.5, and 7 W/cm2) revealed a linear relationship between photothermal temperature and MGN concentrations (Figures d and S1). The light-to-heat conversion by MGNs is initiated by the electronic excitation of MGNs followed by relaxation, which gives rise to rapid nonequilibrium heating of the metal. This is followed by cooling to equilibrium, during which heat is dissipated from the MGNs into the surrounding medium. This dissipated heat is then harnessed by the LTSLs, enabling efficient drug delivery.[55]

The photothermal ability of MGNs was exploited to enable light-activated drug release from the LTSLs with a phase-transition temperature (Tm) of ∼41.7 °C. We synthesized LTSLs following a previously described procedure[56−58] by mixing the lipids DPPC–MSPC–DSPE-PEG2000 at a 90:10:4 ratio in chloroform.[59,60] The desired size of 110.1 ± 6.5 nm (Figure S2a) of LTSLs was achieved by reverse-phase evaporation and subsequent extrusion. At the Tm of the LTSLs, the fully extended hydrocarbon chains of the lipids transform into a disordered liquid crystalline phase, enabling drugs to diffuse through the bilayer membrane.[27] To evaluate the drug release ability of LTSLs, DOX was loaded into the LTSLs using the well-established pH-gradient method.[61−63] This approach resulted in ∼98% DOX encapsulation efficiency within the LTSLs. Rapid release of DOX from the LTSLs (within 5 min) with a 90% efficiency was achieved at a mild hyperthermia temperature of 42 °C (Figure a). At a higher temperature, 45 °C, >97% of DOX was released from the LTSLs in less than 5 min. The long-term stability of the DOX-loaded LTSLs (DOX-LTSLs) at 37 °C was also evaluated for 12 h (Figure S3). DOX leakage studies showed that <5% of DOX escaped from the LTSLs at 37 °C, demonstrating a very low cytotoxicity of LTSLs at physiological temperature (Figures a and S3a). Highly efficient and rapid drug release from LTSLs with mild hyperthermia enhances the bioavailability of the chemotherapy drug in cancer cells with minimal damage to healthy cells. We also compared the drug delivery efficacy of LTSLs with the traditional nontemperature sensitive liposomal DOX formulation, NTSLs. The NTSLs were synthesized by mixing the lipids HSPC–CHOL–DSPE-PEG2000 in a molar ratio of 75:50:3 and a hydrodynamic diameter of 152.7 ± 0.7 nm (Figure S2b) was achieved.[64,65] DOX was loaded in the NTSLs similar to the procedure followed for LTSLs. DOX release studies were performed, demonstrating <10% release at all three temperatures, 37, 42, and 45 °C (Figure b). In additional, long-term stability of the DOX-loaded NTSLs (DOX-NTSLs) also shows minimal DOX leakage from NTSLs (Figure S3b).

Figure 2

Cumulative release of DOX, at 37, 42, and 45 °C for (a) DOX-LTSLs and (b) DOX-NTSLs. (c) Confocal images of MDA-MB-231 breast cancer cells incubated with DOX-LTSLs (left panel), DOX-NTSLs (middle panel), and free DOX (right panel) at 2 μg DOX/mL. The cancer cells were incubated at 37 and 42 °C for 15 min and cellular viability was assessed 30 h after treatment. The cells were stained with calcein (live cells, green) and propidium iodide (PI) (dead cells, red).

The therapeutic ability of LTSLs in controlled drug release was evaluated in MDA-MB-231 TNBC cells (Figure c). TNBC is one of the most lethal types of breast cancer because of its lack of response to endocrine treatment and other targeted therapies.[66,67] Neoadjuvant and adjuvant chemotherapy currently remain the backbone of treatment in TNBC.[68] The survival of MDA-MB-231 cells was compared after treatment with DOX-LTSLs, DOX-NTSLs, and free DOX for 15 min at 37 and 42 °C to evaluate the effectiveness of LTSLs as chemotherapy drug carriers to treat TNBC. After 15 min, all cells were left undisturbed at 37 °C for 30 h to induce apoptosis. A live/dead cell assay was performed by staining live cells with calcein and dead cells with PI. Free DOX (2 μg/mL) resulted in cell death at both temperatures (37 and 42 °C), showing its severe toxicity at physiological temperatures, which leads to off-target cell death with chemotherapy (Figure c, right panel). DOX-NTSLs did not induce cell death at either temperature, showing unsuccessful drug delivery from NTSLs with mild hyperthermia (Figure c, middle panel). Notably, the DOX-LTSLs were not cytotoxic to cells at physiological temperature, and the drug release was triggered only at their phase-transition temperature (42 °C), resulting in cell death (Figure c, left panel). These results reinforce the temperature controllability of drug release for LTSLs, an advantage for co-delivery strategies with plasmonic nanostructures.

By co-delivering MGNs with the LTSLs in breast cancer cells, the photothermal ability of MGNs is harnessed to directly release DOX in the cells for a highly controlled, light-activated cancer therapy. We confirmed MGNs internalization in the cells by using three different approaches—phase contrast imaging (Figure S4), confocal fluorescence imaging (Figure a,b), and TEM imaging (Figure c–e). To visualize cellular uptake, the PEG-coated MGNs were covalently conjugated to fluorescent tag Alexa Fluor 647 via an amide bond (details in the Materials and Methods section) and incubated with MDA-MB-231 cells for 24 h and then imaged with a confocal microscope. At time point zero when the Alexa Fluor-conjugated MGNs were added to the culture media, Z-stack confocal fluorescence images show that no MGNs were co-localized at the same focal plane of the cells (Figure a). Here, the cells are stained green with calcein and the Alexa Fluor-conjugated MGNs appear as red fluorescence. Following 24 h of incubation, clear red fluorescence from the Alexa Fluor-conjugated MGNs was observed inside of the cells (Figure b). The orthogonal view of the Z-stack image (right and bottom panels in Figure b) confirmed that the MGNs were localized inside of the cells in both x–z and y–z directions. To further confirm the uptake and distribution of MGNs inside of the cells, TEM micrographs of MDA-MB-231 cells were obtained after 24 h of incubation with the PEG-coated MGNs (Figure c–e). The dark clusters in Figure c indicate that the MGNs were endocytosed in cells and entrapped in membrane-bound cytoplasmic vesicles.[69,70] We note that the MGNs were not observed in the nucleus (indicated by an arrow in Figure c) or the mitochondria (indicated by arrows in Figure d). The high-magnification TEM micrograph (Figure e) shows that the PEG-coated MGNs maintained their characteristic anisotropic morphologies even upon cellular internalization.

Figure 3

Cellular uptake of PEG-coated MGNs by MDA-MB-231 cells. Z-stack confocal fluorescence images of cells incubated with PEG-coated MGNs at time zero (a) and after 24 h of incubation (b). Orthogonal views (right panel in b) at both x–z and y–z direction show that PEG-coated MGNs were delivered into the cells. (c–e) TEM micrographs of PEG-coated MGNs show internalization by the cells and localization in the intracellular vesicles. In addition, PEG-coated MGNs were neither found in the nucleus indicated by an arrow in (c) nor in the mitochondria indicated by arrows in (d). High magnification micrograph in (e) shows that MGNs maintain their anisotropic morphology in cells.

Following the uptake of PEG-coated MGNs, the cells were incubated with either DOX-LTSLs or DOX-NTSLs at 2 μg DOX/mL and subsequently treated with an 808 nm laser at 5.5 W/cm2 for 15 min (Scheme S1). An infrared camera was used to monitor the temperature elevation during laser irradiation (Figure S5a). The temperature profile of the cellular media with the MGNs during these in vitro experiments is shown in Figure S5b. We note that at this laser flux, the MGNs generated 42 ± 1 °C in the culture dish within the laser beam (3–3.5 mm2 spot size), sufficient to release DOX from the LTSLs. Three hours post laser treatment, the culture media was removed and the cells were washed with phosphate-buffered saline (PBS) buffer. The cells were then stained with Hoechst and imaged with a confocal microscope. The cellular uptake and translocation of DOX into the nucleus were visualized by utilizing the intrinsic fluorescence of DOX (excitation at 488 nm, emission at ∼570 nm).[71] After drug delivery, a concentration gradient of DOX is generated across the cellular membrane which drives DOX influx into the cells; DOX then translocates into the nucleus.[72−74] DOX is known to induce cytotoxicity through both the inhibition of DNA synthesis and the production of free radicals.[75,76] Confocal fluorescent images demonstrate the ability of LTSLs to successfully deliver DOX into the cells by utilizing the photothermal hyperthermia induced by the MGNs (Figure , bottom panel). Because of the high phase-transition temperature of NTSLs, no DOX delivery was achieved (Figure , top panel) at mild hyperthermia.

Figure 4

Confocal fluorescence images of cells incubated with PEG-coated MGNs, illuminated with an 808 nm laser, and treated with either DOX-LTSLs (bottom) or DOX-NTSLs (top). Left panels show the cell nucleus stained with Hoechst, middle panels show DOX fluorescence at ∼570 nm, and right panels are the overlay of left and middle panels, and bright field.

Light activation of MGNs enables highly controlled and localized drug delivery from the liposomes; moreover, the hyperthermia enhances drug uptake and cytotoxicity of anticancer drugs.[15,16] Mild hyperthermia improves cellular membrane permeability and alters the physiological behavior of the cells, making them more susceptible to apoptosis.[43,77−79] We have performed both fluorescence imaging (Figure ) and MTT viability analysis (Figure ) to demonstrate that photothermal hyperthermia induced by MGNs can successfully deliver DOX from LTSLs, resulting in significant cell death. Calcein/PI live/dead cell assays were performed in MDA-MB-231 cells that were illuminated with an 808 nm laser for 15 min. The cells were incubated with PEG-coated MGNs in the presence of either DOX-LTSLs (Figure a) or DOX-NTSLs (Figure b). Because of the high phase-transition temperature (>46 °C) of NTSLs, no cell death was observed with the NTSLs, both with (Figure b) and without (Figure S6b) MGNs. This demonstrates that drug delivery from NTSLs is not achievable under the mild hyperthermia (40–43 °C) generated by MGNs. However, the LTSLs in the presence of PEG-coated MGNs successfully released DOX, resulting in intense cell death (Figure a), where dead cells were stained red with PI. In the absence of MGNs, the DOX-LTSLs did not induce cell death, indicating that the laser itself does not generate enough heat to release DOX from the LTSLs (Figure S6a). We note that unlike photothermal therapy with gold nanoparticles in vitro where necrotic cell death occurs within the laser spot, in photothermal drug delivery in vitro, cell death is localized in the vicinity of the liposomes; but, because of the concentration gradient of DOX, apoptotic cell death extends slightly beyond the laser spot. This process has been observed previously in light-mediated drug delivery in vitro.[80,81] However, this does not imply that photothermal DOX release in vivo will pose a risk of off-target toxicity because DOX remains within the tumor microenvironment, and successful tumor treatment with this approach has been shown previously.[8] Apoptotic cell death by a chemotherapy drug is strongly preferred over necrotic cell death incurred during photothermal therapy by gold nanoparticles because necrosis gives rise to uncontrollable inflammation and undesirable immunogenic responses because of the loss of cellular membrane integrity.[82,83] To confirm apoptotic cell death, Annexin was used to stain the phosphatidylserine of apoptotic cells 16 h post treatment with PEG-coated MGNs + DOX-LTSLs + laser (Figure c). Calcein and PI were also used to indicate live and dead cells. The overlay image (Figure d) of Annexin (blue) with calcein (green) and PI (red) clearly shows cells that were undergoing apoptosis, where the arrows indicate live cells that were in their apoptotic cell cycle. We also performed control experiments where MDA-MB-231 cells were not treated with MGNs or liposomes and were further stained with Annexin (Figure S7). Only a few of these control cells underwent normal cell cycle and apoptosis. We note that the particle density of the MGNs utilized in this study is not significant enough to induce cell death via photothermal heating. Moreover, the MGNs alone (Figure S8a) or laser alone at 5.5 W/cm2 does not induce significant cell death (Figure S8b).

Figure 5

Therapeutic effects of photothermal drug delivery from the liposomes. MDA-MB-231 cells were incubated with PEG-coated MGNs and (a, c, and d) DOX-LTSLs or (b) DOX-NTSLs. The cells were treated with an 808 nm laser at 5.5 W/cm2 for 15 min; laser spot size was ∼3.5 mm2. The cells were stained with calcein (live cells, green) and PI (dead cells, red) 30 h post treatment. (c) The cells were also stained with Annexin V (blue) to evaluate apoptotic cell death 16 h post treatment. (d) Overlay image of Annexin V with calcein/PI, with arrows indicating the live cells that underwent apoptosis (stained with both blue and green).

Figure 6

(a) MTT assay of all samples and (b) MTT assay summarizing the cellular viability of cells treated with DOX-LTSLs, DOX-LTSLs + laser, and DOX-LTSLs + MGNs + laser. Note that data for (b) were simply taken from (a) and shown separately for clarity. All samples for this assay were illuminated with an 808 nm laser at 5.5 W/cm2 for 15 min. The error bars represent n = 4 samples. Data are shown as mean ± standard deviation.

Quantitative assessment of the therapeutic outcome of photothermal drug delivery was achieved with MTT cellular viability assays (Figure a). The MDA-MB-231 cells were incubated with different concentrations of DOX-LTSLs, DOX-NTSLs, and free DOX for 12 h both with and without PEG-coated MGNs. At a low DOX concentration (0.5 or 2 μg/mL), DOX-LTSLs and DOX-NTSLs did not induce significant cell death, showing >90% survival. The cellular viability decreased to ∼77% using both DOX-LTSLs and DOX-NTSLs at a DOX concentration of 20 μg/mL, likely attributed to the increase in liposomes uptake or leakage of DOX at such high concentrations. Cells treated with DOX-LTSLs alone and irradiated with an 808 nm laser did not result in significant cell death because the laser alone does not generate sufficient heat to reach the phase-transition temperature of LTSLs. However, the difference in the cellular viability between cells that were treated with DOX-LTSLs + laser and cells treated with DOX-LTSLs + MGNs + laser at all three DOX concentrations were statistically significant (p = 0.005 for 0.5 μg/mL, 0.003 for 2 μg/mL and 0.003 for 20 μg/mL). The photothermal hyperthermia induced by PEG-coated MGNs initiated rapid drug delivery from DOX-LTSLs, resulting in a cellular viability of 67%, 42%, and 39% at 0.5, 2, and 20 μg/mL DOX concentrations, respectively. Notably, the combination of MGNs with DOX-LTSLs resulted in higher cell death than that caused by free DOX alone at all three concentrations of DOX (84%, 50%, and 43%). For clarity, we also replotted the MTT assay data from Figure a for the DOX-LTSLs samples (DOX-LTSLs, DOX-LTSLs + laser, and DOX-LTSLs + MGNs + laser) to clearly show that a significantly higher cell death is observed because of DOX release from the LTSLs in the presence of MGNs (Figure b). Because NTSLs are not thermally responsive at mild hyperthermia, negligible cell death was observed for all cells incubated with DOX-NTSLs independent of the presence of MGNs. Our results convey that the MGN-mediated mild hyperthermia results in a successful drug delivery from LTSLs with high therapeutic efficacy below the United States Food and Drug Administration (FDA)-approved DOX concentration. FDA standards for DOX treatment in a 60 kg adult human is ∼2 μg/mL every 3 weeks, not to exceed a total of 13.5 μg/mL.[84−86] In summary, whereas free DOX will induce toxicity even in healthy cells when translated to a clinical setting, targeted photothermal drug delivery with MGNs is light-activated, is highly controlled, and can minimize off-target toxicities.

Conclusions

In this study, we synergistically combined the therapeutic effects of PEG-coated MGNs with DOX-LTSLs to eliminate aggressive TNBC cells with a highly controlled, light-activated approach. The unique nanoantenna-like geometry of MGNs, consisting of a core with multiple sharp protrusions, enables strong light-to-heat conversion, allowing rapid and highly efficient drug delivery from LTSLs. We chose a model TNBC cell line, MDA-MB-231, to assess both qualitatively (calcein/PI live/dead cell assay and apoptosis assay) and quantitatively (MTT assay) the therapeutic outcome of MGNs combined with DOX-LTSLs. Our results provide strong evidence that photothermal heating by the MGNs gives rise to mild hyperthermia sufficient to reach the phase-transition temperature of LTSLs, resulting in 58% cell death at a DOX concentration of 2 μg/mL. We envision that the co-delivery of LTSLs with MGNs will ultimately facilitate a clinically applicable technology, enabling the delivery of dose-controlled treatment and minimizing off-target toxicities that are associated with chemotherapy. Future integration of this approach to include imaging capabilities would enable longitudinal tracking to further improve detection, treatment, and assessment of therapeutic outcomes with a platform that can be externally triggered by near-infrared light. In addition, LTSLs can ultimately be loaded with a range of drugs, and when combined with MGNs, this platform will facilitate targeted treatment of multiple diseases, including cancer, multiple sclerosis, and many infectious diseases.

Materials and Methods

Synthesis of PEG-MGNs

All reagents were purchased from Sigma-Aldrich except methoxy-polyethylene glycol-thiol (mPEG-SH, Mw 5000, JenKem Technology). The MGNs were synthesized through the one-step HEPES-mediated growth method.[21] First, 18 mL of Milli-Q water (18 MΩ) was added to 12 mL of 270 mM HEPES (pH 7.40 ± 0.2). Then, 300 μL of 20 mM chloroauric acid (HAuCl4) was mixed with the HEPES solution through gentle inversion and left to react at room temperature for 75 min. To PEGylate the MGNs, 3 mL of 50 μM mPEG-SH was added to the MGNs and reacted at 4 °C for 24 h.[87,88] After the reaction, the MGN solution was centrifuged at 6000 rpm for 20 min while discarding the supernatant and resuspending the pellet in 2 mL of Milli-Q water. To ensure complete surface PEGylation, a second addition of mPEG-SH was performed (100 μL of 5 μM mPPEG-SH) for 10 min with constant stirring. Lastly, the PEG-coated MGNs were washed with Milli-Q water twice to remove excess chemicals.

Characterization of PEG-Coated MGNs

The PEG-coated MGNs were imaged with an Osiris transmission electron microscope at 200 keV to evaluate their size and morphology. The absorbance properties of the MGNs and the PEG-coated MGNs were characterized using a Varian Cary 5000 UV–vis near-infrared spectrophotometer. A Malvern Nano ZS dynamic light-scattering apparatus was used to measure the zeta potential of the MGNs before and after PEGylation.

Photothermal Ability of PEG-Coated MGNs

Laser optics used for this project were purchased from Thorlabs. The optics included an 808 nm laser diode (L808P1WJ), a current controller (LDC240C), a thermoelectric controller (TEDC300C), a collimating lens (C230TME-B), and a silver-polished mirror (PF-10-03-P01). Calipers and a power meter (Thorlabs, PM100D) were used to measure the laser spot size and power density.

PEG-MGNs at four different concentrations, 80, 110, 140, and 170 μg/mL, were mixed with a complete cellular media, Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS, ATCC). The solution was preheated to 37 °C, added to a 35 mm cell culture dish, and illuminated with the laser at 4, 5.5, and 7 W/cm2 for 20 min. To mimic the biological environment, the temperature was kept constant by using a temperature controller and a heating mantle. The culture dishes with the PEG-coated MGNs suspensions were placed on top of an aluminum block wrapped with antireflecting black films atop the heating mantle. The temperature of the solution was monitored using an infrared camera (FLIR T400), taking images incrementally over time.

Synthesis of DOX-Loaded Liposomes

All lipids, dipalmitoylphosphatidylcholine (DPPC), 1-stearoyl-2-hydroxy-sn-glycero-3-phosphocholine (MSPC), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000), cholesterol, and hydrogenated-l-α-phosphatidylcholine (HSPC) were purchased from Avanti Polar Lipid. Doxorubicin hydrochloride was purchased from Fisher Scientific.

LTSLs and NTSLs were synthesized by a reverse-phase evaporation method.[56−58] Briefly, 5 mg of lipid (for LTSLs: DPPC–MSPC–DSPE-PEG2000 = 90:10:4; for NTSLs: HSPC–CHOL–DSPE-PEG2000 = 75:50:3 in molar ratio)[59,60,64,65] was dissolved in 1 mL of chloroform, then dried using nitrogen air and subsequent vacuum. To form the liposomes, the lipid film was hydrated with a 300 mM citrate buffer (pH 4.0) for 60 min (at 55 °C for LTSLs and 60 °C for NTSLs). The liposomes were then extruded 10 times through a 400 nm and a 100 nm polycarbonate membrane to obtain the desired size. The outside pH of the liposome solution was titrated to pH 7.5 using 0.5 M sodium carbonate. As a result, a pH gradient was generated across the lipid bilayer.[61−63] DOX was then added to the liposome solution at a 1:20 DOX/lipid weight ratio and mixed for 60 min (at 37 °C for LTSLs and 60 °C for NTSLs). The final product was passed through a PD10 column to remove excess DOX.

Characterization of DOX-Loaded Liposomes

A Malvern Nano ZS dynamic light-scattering apparatus was used to measure the hydrodynamic size of DOX-LTSLs and DOX-NTSLs. The encapsulation efficiency (EE) of DOX was monitored at 480 nm by using UV–vis spectroscopy.[89] Triton X-100 (2% wt/v) was added to the DOX-LTSLs and NTSLs solution to fully lyse the liposomes.

Temperature-Triggered DOX Release Measurement

DOX release from the LTSLs and NTSLs was monitored at three different temperatures, 37, 42, and 45 °C. First, 0.9 mL of 25 mM HEPES buffer saline (HBS) was prewarmed in an incubator to the above-listed temperatures. At time zero, 100 μL of DOX-LTSLs or DOX-NTSLs was added to the HBS and incubated for a given time point. Dequenching of DOX fluorescence was monitored at 480 nm using a UV–vis spectrophotometer.[89] Cumulative drug release was calculated bywhere A, A0, and A∞ are the absorbances at a given time point, at time zero, and when the liposomes were fully lysed with Triton X-100, respectively.

Confocal Fluorescence Imaging of Cellular Internalization of PEG-Coated MGNs

Alexa Fluor 647 NHS Ester conjugate (ThermoFisher Scientific) was attached to the PEG-coated MGNs through an OPSS-PEG-amine (JenKem Technology) linker, which creates a covalent amide bond between the NHS ester and amine. Alexa Fluor was mixed with the OPSS-PEG-amine linker in a 3:1 molar ratio at room temperature for 2 h. The product was then added to the MGN solution and reacted at 4 °C for 3 h. An additional mPEG-SH layer was then added to the fluorophore-conjugated MGNs for further stability. The final product was centrifuged at 4000 rpm for 10 min. The Alexa-conjugated MGNs were added to the cells and incubated for 24 h. Calcein was used to stain the live cells. Z-stack images were taken at both time zero (when the MGNs were just added to the cells) and after 24 h of MGNs incubation. A Zeiss LSM 710 confocal microscope was used for the imaging.

TEM Imaging of Cellular Uptake of PEG-Coated MGNs

MDA-MB-231 cells were incubated with the PEG-coated MGNs for 24 h. The cells were then washed with a buffer, fixed in 2.5% gluteraldehyde in 0.1 M cacodylate buffer at room temperature for 1 h and 4 °C overnight. The samples were processed and imaged in the Vanderbilt Cell Imaging Shared Resource Research Electron Microscope facility. The cell samples were post-fixed in 1% osmium tetraoxide at room temperature and then washed with a 0.1 M cacodylate buffer 3 times. Following fixation, the samples were dehydrated through a graded ethanol series followed by incubation in 100% ethanol and propylene oxide (PO) and 2 exchanges of pure PO. Lastly, the samples were embedded in epoxy resin and polymerized at 60 °C for 48 h. To obtain thin sections of the sample, 70–80 nm ultra-thin sections were cut from the block and collected on copper grids. The copper grids were stained with 2% uranyl acetate and lead citrate. The samples were imaged on a Philips/FEI Tecnai T12 electron microscope at different magnifications.

In Vitro Temperature-Triggered Release of DOX

MDA-MB-231 cells were purchased from ATCC (HTB-26) and cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. The cells were maintained at 37 °C and 5% CO2. Forty-eight hours before the experiments, 330 000 cells were seeded on a 35 mm cell culture dish. On the day of the experiment, fresh media containing DOX-LTSLs, DOX-NTSLs, or DOX at 2 μg DOX/mL was added to the cells. The cells were incubated at either 37 or 42 °C for 15 min. After 12 h, the media was changed to fresh cDMEM and cultured for an additional 18 h. The cells were stained with calcein (5 μM) and PI (1 μM) for 20 min. A Zeiss LSM 710 confocal microscope was used for fluorescence imaging of the cells.

In Vitro Photothermal Drug Delivery and Fluorescence Imaging

MDA-MB-231 cells were passaged for at least 14 days before seeding for the experiment. Approximately, 11 000 cells per well were seeded on a 96-well plate and allowed to adhere and grow. After 48 h, 65 μL of PEG-coated MGNs (170 μg/mL) in DMEM were added to the cells and incubated at 37 °C for 24 h. The media was then removed, and the cells were washed with cDMEM twice. A Zeiss Observer Z1 microscope was used to visualize the cellular uptake of PEG-coated MGNs.

To show the DOX release by LTSLs or NTSLs (Figure ), 24 h after MGNs incubation, DOX-LTSLs at 2 μg/mL were added to the cells. The cells were then allowed an hour to equilibrate. After an hour, an 808 nm laser was applied to the cells and illuminated at 5.5 ± 0.1 W/cm2 for 15 min with a laser spot size of approximately 3.5 mm2. An infrared camera was used to monitor the temperature changes during laser irradiation (see Figure S5). The culture media was removed 3 h post laser treatment and washed with PBS twice. This process removed any LTSLs (or NTSLs) from the cells, but DOX that was already released from the LTSLs (or NTSLs) remained internalized in the cells. The cells were then stained with Hoechst and imaged with a confocal microscope.

For in vitro photothermal therapeutic study (Figures and 6), a similar procedure was followed as stated above. However, the culture media was changed to fresh cDMEM 12 h after laser irradiation instead of 3 h after laser treatment. The cells were left undisturbed (with the fresh media) for an additional 18 h. Thus, after a total of 30 h post laser treatment, the cells were stained with calcein and PI and imaged using the confocal microscope to evaluate the cellular viability qualitatively. The timeline of the experiment is shown in Scheme S1. For apoptosis study, Annexin V and Pacific Blue (ThermoFisher Scientific) at 25/100 μL were used to stain the cells 16 h post laser treatment.

Cellular Viability Assay

MDA-MB-231 cells were passaged and seeded identically on 96-well plates as stated previously. The cells were incubated with 170 μg/mL PEG-coated MGNs for 24 h and washed with the cell medium twice. DOX-LTSLs or DOX-NTSLs at DOX concentrations of 0.5, 2, and 20 μg/mL were then added to the cells. An 808 nm laser at 5.5 W/cm2 was applied for 15 min. The drug was removed, and the culture media was refreshed after 12 h of laser irradiation. After an additional 18 h of incubation (30 h total), the old media was removed and replaced with 100 μL of fresh media mixed with 10 μL of 12 mM 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide solution (MTT, ThermoFisher Scientific). After 2 h of MTT incubation, 85 μL of media was removed and 50 μL of dimethyl sulfoxide (DMSO) was added to solubilize and dissolve the formazan. A Biotek Synergy H1 plate reader was used to read the absorbance at 540 nm.

Statistical Analysis

All data are presented as mean ± standard deviation. Student’s t-tests were performed to evaluate the statistical differences. Statistical significance is considered when p < 0.05.