BaTiO₃-core Au-shell Nanoparticles for Photothermal Therapy and Bimodal Imaging

Abstract

We report sub-100 nm metal-shell (Au) dielectric-core (BaTiO₃) nanoparticles with bimodal imaging abilities and enhanced photothermal effects. The nanoparticles efficiently absorb light in the near infrared range of the spectrum and convert it to heat to ablate tumors. Their BaTiO₃ core, a highly ordered non-centrosymmetric material, can be imaged by second harmonic generation, and their Au shell generates two-photon luminescence. The intrinsic dual imaging capability allows investigating the distribution of the nanoparticles in relation to the tumor vasculature morphology during photothermal ablation. Our design enabled in vivo real-time tracking of the BT-Au-NPs and observation of their thermally-induced effect on tumor vessels.

Graphical abstract

1. INTRODUCTION

Photothermal therapy (PTT) induced by plasmonic nanoparticles has emerged as a promising approach to treating cancer.[1–3] Unlike other therapeutic approaches such as chemo- and radiation therapies, PTT is noninvasive with minimal – and local – side effects. PTT is based on the ability of photothermal transducers to efficiently convert light to heat, ablating tumors in a selective and controlled manner. PTT can be used to ablate solid tumors and the two mechanisms responsible for PTT-triggered cell death have been identified — necrosis and apoptosis.[4, 5] However, the study of the role of intratumoral nanoparticle distribution in mediating tumoricidal activity has been hampered by the lack of suitable imaging techniques. Therefore, a plasmonic nanoparticle that has PTT capability and also allows imaging would be highly desirable.

Experimental PTT has often employed plasmonic silica core gold shell nanoparticles[6] (Si-Au-NPs), which have excellent biocompatibility[7], scalability[8], and ease of surface functionalization[9]. They can be easily tuned to absorb NIR light[2, 4], which can penetrate relatively deeply into tissues[10]. One determinant of the effectiveness of PTT with these and other nanoparticles, given the finite deposition of nanoparticles within tumors, is the efficiency with which a given irradiance is converted into heat. This efficiency can be improved by increasing the dielectric constant of the core.[11] Moreover, the ability of plasmonic nanoparticles to absorb light and produce heat decreases as particle size increases; small nanoparticles are predominantly absorptive, generating more heat.[11, 12] Current synthetic procedures for Si-Au-NPs limit them to relatively large sizes (~150 nm).[13, 14] This relatively large size also may not be optimal for penetration of tumors by the enhanced permeation and retention (EPR) effect.[15] The difficulty with making smaller Si-Au-NPs is the ratio of the thickness of the Au layer to the radius of the silica core has to remain constant to maintain the plasmon resonance in the NIR range; growing thinner Au shells on Si nanoparticles is not achievable by current experimental procedures.

We selected barium titanate (BaTiO₃) as the core material to make sub-100 nm gold-shell dielectric-core nanoparticles. BaTiO₃ was selected because its dielectric constant is orders of magnitude higher than that of silica[16] – allowing the fabrication of NIR absorbant nanoshells with relatively thicker gold layers. Recently it was reported that gold shell barium titanate nanoparticles (320 nm diameter) exhibited a photothermal effect in vitro.[17] By using a 50-nm barium titanate core, we were able to synthesize sub-100 nm gold-shell BaTiO₃–core nanoparticles (abbreviated as BT-Au-NP) that have photothermal capability. Here we investigate whether these particles can produce a therapeutically effective PTT in vivo.

These particles were also appealing because of their potential associated imaging modalities. The BaTiO₃ core is a highly ordered non-centrosymmetric material and can be imaged by second harmonic generation (SHG),[18, 19] and the gold shell generates two-photon luminescence (TPL)[20, 21]. These intrinsic imaging properties could allow the nanoparticles to be monitored in real time during the PTT process in vivo.

2. MATERIALS AND METHODS

2.1. The synthesis and characterization of BT-Au-NPs

BaTiO₃ nanoparticles (50 nm, from Inframat Advanced Materials, Manchester CT) were first treated with H₂O₂ to hydroxylate the surface. The BT nanoparticles were suspended in H₂O₂ and heated at 80 °C for 6 hours.[22] The H₂O₂ was then discarded and the BT nanoparticles were rinsed 3 times with water and finally suspended in pure ethanol. The hydroxyl groups on the surface of the BT nanoparticles were then converted to amine groups by reaction with (3-aminopropyl)triethoxysilane (APTES, Sigma Aldrich, St. Louis MO) in ethanol.

Small Au colloids (2~3 nm) were synthesized using the method of Duff and Baiker[23] then aged for 2 weeks. Aminated BT nanoparticles were then added to the gold colloid suspension and incubated overnight to allow the formation of gold colloids-decorated BT nanoparticles (BT seeds). BT seeds were then reacted with HAuCl₄ in the presence of formaldehyde to form the BT-Au-NPs.[24] In this process the Au³⁺ ions were reduced to Au metal starting from the Au colloids, which serve as nucleation sites, to grow full shell of Au around the BT nanoparticles. The thickness of the Au shell is proportional to the amount of HAuCl₄ reduced, which was assessed by UV-Vis spectrophotometer, TEM, and SEM.

10 μM of thiolated PEG (poly(ethylene glycol) methyl ether thiol, M_n=5000, Sigma Aldrich, St. Louis MO) was mixed with 2×10¹⁰ BT-Au-NPs per mL of water overnight. The reaction mixture was then centrifuged (500 RCF for 30 min) and resuspended in pure water or PBS as needed.

To synthesize PEGylated BT-NPs, 10 μM silane PEG (mPEG-Silane, M_n=5000, Laysan Bio, Inc, AL) was mixed with 2×10¹⁰ hydroxylated BT nanoparticles overnight. The reaction mixture was then centrifuged and resuspended in water or PBS. The PEGylated NPs did not aggregate in medium containing FBS, as evidenced by the absence of precipitation.

2.2. Calculation of the photothermal conversion efficiency of BT-Au-NPs

The photothermal conversion efficiency η is calculated using equation (1)

$$\mathrm{\eta }=\frac{hs(T_{\mathit{max}}-T_{\mathit{surr}})-Q_{\mathit{dis}}}{I(1-{10}^{-A_{808}})}$$ (1)

where T_max is the equilibrium temperature (77.5°C), T_surr is ambient temperature (23.9°C), Q_dis is the heat dissipated by the container itself and is calculated to be approximately 0 W, I is the incident laser power (1 W cm⁻²), A₈₀₈ is the absorbance of the BT-Au-NP sample solution at 808 nm (measured to be 0.468 by spectroscopy), h is the heat transfer coefficient, s is the surface area of the container. hs can be calculated using equation (2)

$$hs=\frac{m_s{c}_s}{{\tau }_s}$$ (2)

where m_s and c_s are the mass (0.2 g) and heat capacity (4.2 J g ⁻¹) of the solvent, τ_s is the time constant for heat transfer of the system and is determined by linear fitting (Figure S5) of time versus -ln θ during the cooling period of the sample, θ is defined in equation (3)

$$\mathrm{\theta }=\frac{T-T_{\mathit{surr}}}{T_{\mathit{max}}-T_{\mathit{surr}}}$$ (3)

hs is fitted to be 259.35. Thus η= (0.2*4.2/259.35*(77.5-23.9))/(1-10^(−0.468)) = 0.2632.

2.3. In vitro cytoxicity study of BT-Au-NPs

4T1 or HUVEC cells were cultured in 96-well plates inEGM™ −2 endothelial cell growth media in a humidified atmosphere with 5% CO₂ at 37 C. HUVEC Cells were rinsed with PBS and incubated with BT-Au-NPs suspended in fresh media (2×10¹⁰ particles per mL). BT-Au-NP-free control treatments received fresh medium only. After overnight incubation, excess unbound BT-Au-NPs were removed by rinsing with PBS then incubated with fresh culture medium. Cells were exposed to NIR light (RPMC lasers, Inc. O’Fallon, MO, 808 nm, 1W/cm²) for different duration to induce photothermal cell damage. After exposure, cells were incubated for an additional one-hour at 37°C. Cell viability was assessed with MTS test. Slightly different conditions were used to induce photothermal cell damage to 4T1 cells, following a previously published protocol.[25] 4T1 cells were incubated with 0.5 mg/ml BT-Au-NPs for 4 hours, then excess BT-Au-NPs were removed, and fresh medium were added to the wells. Cells were exposed to a 808nm laser for different duration, and then incubated overnight at 37°C, before their viability was assessed with MTS assay.

2.4. In vivo PTT of BT-Au-NPs

Immunodeficient 6 to 8-weeks old nude mice were purchased from Charles River Laboratories (Boston, MA) and maintained under pathogen-free conditions. All animal studies were performed under the guideline of the MIT Animal Care and Use Committee. For subcutaneous 4T1 tumor models: 4T1 mouse breast adenocarcinoma cells (10⁶ cells/0.2 mL in 1:1 (vol/vol) PBS and Matrigel, BD Biosciences San Jose, CA) were inoculated on the mouse mammary fat pad and grown to tumor burden of 100-150 mm³ in volume.

Mice were anesthetized with isoflurane and a suspension of PEGylated BT-Au-NPs in PBS (12 mg/ml, 50 μL) was injected systemically in each mouse. Mice used for control experiments received an injection PBS only. 6 hours after injection, mice bearing tumors were exposed to NIR light (1.2W/cm², 808 nm, 3min). The tumor volume and the mass of mice were recorded every other day for 3 weeks. The Tumor volume was calculated using the previously reported modified ellipsoid formula[26]: tumor volume = (length) × (width) × (width) × (0.5).

Organs including tumor excision and fixation proceeded 7 days after treatment. Histology evaluation was conducted by means of hematoxylin/eosin staining. ICP-MS was performed at the Trace Metal Laboratory at the Harvard T.H Chan School of Public Health.

2.5. Tumor imaging

Mice bearing tumors were anesthetized, intravenously injected with PEGylated BT-Au-NP or PEGylated BT nanoparticles (12 mg/ml, 50 μL) and TAMRA-Dextran (0.1 mL, 2.5 wt%, molecular weight ~70 kDa, Life Technologies, CA, USA), and then a small incision was made in the mouse flank to expose the tumor for imaging. The incision did not interrupt the tumor vasculature. Inverted multiphoton microscope was used for to imaging. Tumors or skin were excited with 820 nm, 45 mW/cm², <100 fs pulse, scanning frame rate of 4s and 3 channels were used: 425 ± 30nm for SHG imaging, 525 ± 45nm for TPL imaging, and 607 ± 70nm for TAMRA-dextran fluorescence imaging. It is worth noting that TAMRA-Dextran is excited via two-photon absorption. Pulsed laser is used as excitation source to increase the probability of SHG and TPL as both are nonlinear optical processes. The images were then processed with ImageJ.

Animal study protocol was reviewed and approved by the BCH Committee on Animal Care.

2.6. Statistical analysis

Data that conformed to a normal distribution are described with means and SDs. Otherwise, variables are presented as median ± quartiles. Comparisons between groups for PTT efficacy were done using two-way repeated-measures analysis of variance (ANOVA) with the interaction of group-by-time assessed by the F-test to compare change in tumor volume over time between the treatment groups. Bonferroni-Holm adjusted two-tailed p < 0.05 were considered statistically significant to protect against Type I errors due to multiple group comparisons. Statistical analysis was conducted using the Stata software package (release 14, Stata Corporation, College Station, TX).

3. RESULTS

3.1. Synthesis and Characterization of the BT-Au-NPs

BT-Au-NPs as synthesized (Fig. 1a; See Methods) had an average hydrodynamic diameter of 82.9 ± 21.6 nm and polydispersity index of 0.124 by dynamic light scattering (DLS). TEM (Fig. 1b) confirmed the approximate size and showed the core-shell structure with an Au shell thickness ~ 10 nm and a core diameter of ~ 80 nm. The sub-100 nm size, spherical morphology, uniformity, and lack of aggregation of the BT-Au-NPs were further confirmed by SEM (Fig. 1c). The maximum light absorption of BT-Au-NPs was at around 800 nm (Fig. 1d). The photothermal capability of the BT-Au-NPs was seen in the greater heating of an aqueous suspension of BT-Au-NPs compared with an equal volume of water when irradiated with a NIR laser (808 nm, 1W/cm², 10 min) (Fig. 1e). The BT-Au-NPs were PEGylated by reaction with thiolated PEG (poly(ethylene glycol) methyl ether thiol, M_n=5000) to prevent particle aggregation and increase circulation time. All BT-Au-NPs described below are PEGylated. The extinction coefficient of BT-Au-NP in H₂O at 808 nm was measured to be 3.2 L g⁻¹ cm⁻¹. BT-Au-NPs do not undergo photothermal reshaping – TEM images of irradiated and unirradiated BT-Au-NPs were indistinguishable. The photothermal conversion efficiency η of BT-Au-NPs for 808 nm laser was calculated to be 26.32%, following a previously reported method[27] (also see Methods, section 2.2).

Figure 1.

BT-Au-NPs fabrication and characterization. (a) Scheme of BT-Au-NP fabrication. Representative (b) TEM and (c) SEM images of BT-Au-NPs. (d) Absorption spectrum of BT-Au-NPs. (e) Rate of temperature increase of aqueous suspension of BT-Au-NPs at optical density of 0.468, mass concentration at 0.145 mg/mL. (808 nm CW laser, 1 W/cm², n = 4)

3.2. In vitro cytotoxicity of BT-Au-NPs

Human umbilical endothelial cells (HUVEC) incubated overnight with BT-Au-NPs were exposed to NIR light (808 nm CW laser, 1 W/cm²) and cell viability was assessed over time by MTS test assay. After 90 seconds of irradiation, cell viability was reduced by 95%. Cells irradiated in the absence of BT-Au-NP for the same duration, showed a 30% decrease in viability, which was attributed to heat generated by the NIR light. The cytotoxicity of BT-Au-NPs following overnight incubation and without subsequent irradiation was negligible (zero time point, Fig. 2a). 4T1 cell viability was decreased by 89% after 180 seconds of irradiation.

Figure 2.

In vitro cellular uptake and viability. (a) Viability of HUVECs irradiated with NIR laser (808 nm CW laser, 1 W/cm²) with or without incubation overnight with BT-Au-NPs. Data are mean ± SD, (n = 4). (b) Viability of 4T1 cells irradiated with NIR laser (808 nm CW laser, 1 W/cm²) with or without exposure to BT-Au-NPs. At time 0 (beginning of irradiation), cells had been exposed to BT-Au-NPs for 4 hours. Data are means ± SD, n = 4. (c) Representative fluorescence (Ex/Em= 488/520±20 nm) and SHG (Ex/Em = 800/400±15 nm) images of U87 cells incubated overnight with and without BT-Au-NPs. Cells were stained with calcein-AM.

3.3. In vitro cellular uptake of BT-Au-NPs

The cellular uptake of BT-Au-NPs was assessed by fluorescence and SHG imaging. In this study, human glioblastoma U87 cells were used instead of HUVECs due to their higher resistance to heat. (HUVECs were photothermally damaged when imaged, data not shown). U87 cells were incubated with BT-Au-NPs overnight (see Methods), stained with calcein-AM, then imaged with a two-photon confocal microscope where fluorescence (Ex/Em = 488/520±20 nm, to depict cell morphology) and SHG (Ex/Em= 800/400±15 nm, to depict BT-Au-NPs) images could be acquired simultaneously. SHG images (Fig. 2b) showed that BT-Au-NPs were taken up and distributed in the cellular cytoplasm. No SHG signal was recorded in U87 cells that were not incubated with BT-Au-NPs (Fig. 2b).

3.4. Efficacy of PTT using BT-Au-NPs

The in vivo efficacy of BT-Au-NPs as PTT agents was evaluated in a mouse 4T1 breast adenocarcinoma cell tumor model (Fig. 3a). 4T1 cells were inoculated in mouse mammary fat pads and grown to 100-150 mm³. To demonstrate that irradiation of BT-Au-NPs in tumors could result in a photothermal effect, animals were then injected with BT-Au-NPs (12 mg/mL, 50 μL), and irradiated with collimated light (CW, 808 nm, 1.2 W/cm², 3 min) 6 hours afterwards. The temperature of the tumor increased by 26.4°C to 58.4°C (measured by a photothermal camera), which is sufficient to induce tumor destruction. [28] (Fig S1a). Similar irradiation after injection of PBS only increased the local temperature by 10.5°C. Subsequently, mice were injected intravenously with the following regimens (N=5 each): PBS with or without irradiation, or BT-Au-NPs with or without irradiation at the tumor site. Tumors were irradiated with collimated light 6 hours after BT-Au-NPs administration (CW, 808 nm, 1.2 W/cm², 3 min; a duration of irradiation that did not cause tissue injury). The mean tumor volume exceeded 500 mm³ (tumor diameter over ca. 1 cm) after 5 days in the groups treated with PBS (with and without irradiation) and BT-Au-NPs without irradiation. All mice were euthanized when the tumor size reached around 1000 mm³. Irradiation of the BT-Au-NPs-injected mice greatly reduced tumor size. The lack of tumor growth persisted for at least 2 weeks after a single treatment of BT-Au-NP injection and irradiation at the tumor site. There was no weight loss in any group of animals (Fig. S1b).

Figure 3.

Effect of BT-Au-NP in a murine 4T1 breast tumor model. (a) Tumor volumes in mice treated with BT-Au-NPs or PBS with or without subsequent irradiation (808 nm CW laser, 1.2 W/cm², 3 min). Data are mean ± SEM (n=4). The asterisk denotes highly significant differences vs. each of the other groups in the slope or time-related change in volume (all p<0.001). (b) Biodistribution 24 hours after injection of titanium expressed as percent injected dose per gram tissue (%ID/g) of mice treated with BT-Au-NPs only. Data are means ± SD (n=5).

Two-way repeated-measures ANOVA indicated significant group differences in average tumor volume (F = 4.45, p = 0.021) averaged across the time period post injection. Post-hoc Bonferroni-Holm testing confirmed significantly smaller mean volume in mice treated with BT-Au-NPs+light compared to PBS (p = 0.002), PBS+light (p = 0.012) and BT-Au-NPs (p = 0.03). No significant differences in mean volume were observed between PBS versus BT-Au-NPs (p = 0.208) or PBS+light (p = 0.41) and no difference between PBS+light versus BT-Au-NPs (p = 0.64). Additionally, assessing the group-by-time interaction test in repeated measures ANOVA (i.e., F-test for comparing slopes over time) indicated significantly less change in volume per day for mice treated with BT-Au-NPs+light compared to PBS (F = 27.47, p < 0.001), PBS+light (F = 33.39, p < 0.001) and BT-Au-NPs (F = 44.52, p < 0.001).

3.5. Biodistribution and in vivo toxicity of BT-Au-NPs

Mice were injected intravenously with 50 μL of 12 mg/ml BT-Au-NPs. After 24 hours, the concentrations of Ti in their organs were measured by inductively coupled plasma mass spectroscopy (ICP-MS, Fig. 3b). BT-Au-NPs primarily accumulated in the spleen and liver, with much lower concentrations detected in the kidney, heart, lung and blood. The accumulation of BT-Au-NPs in the liver and spleen is consistent with previous reports for gold nanoparticles of comparable size.[29, 30]

Blood chemistries (Table S1) obtained 7 days after administration of BT-Au-NPs in healthy mice showed no or minimal deviations from the normal range, suggesting no or minimal systemic toxicity. (Here “toxicity” refers to the harmful side effects of drugs, rather than tumoricidal activity.) Hematoxylin-eosin stained histological sections of organs from the same mice were normal (Fig. S2a). The kidney tubules, spleen white pulp and liver parenchyma do not show significant changes after treatment with BT-Au-NP.

To study the tissue toxicity of the treatment regimen used below in vivo, mice were injected with BT-Au-NPs and irradiated with NIR light (808 nm, 1.2 W/cm², 3 min) at the tumor site 6 hours after injection, and tissues were obtained after 7 days. Histology of the tumor revealed central necrosis (Fig. S2). No necrosis was observed in mice injected with BT-Au-NPs but not irradiated. Histology elsewhere (kidney, spleen, liver) was benign.

3.6. In vivo SHG and TPL imaging

For in vivo multi-photon microscopy, nude mice bearing 4T1 breast tumors (~ 100 mm³) on their mammary fat pads were anesthetized then BT-Au-NPs and tetramethylrhodamine (TAMRA)-dextran (a fluorescent dye used to delineate the vasculature, but which can extravasate) were administered intravenously. The BT-Au-NPs would allow SHG imaging (of tumor collagen and BT-Au-NPs), and the TPL imaging (of BT-Au-NPs). Then an incision was made in the mouse flank to expose the tumor for imaging without interrupting the tumor vasculature. The tumor site was irradiated with an 820 nm pulsed laser (45 mW/cm², <100 fs pulse). (Pulsed lasers are used as excitation sources in SHG and TPL imaging, as both are non-linear processes requiring high-energy excitation.) Images were acquired simultaneously in three different emission channels: 425 ± 30nm for SHG imaging, 525 ± 45nm for TPL imaging, and 607 ± 70nm for TAMRA-dextran fluorescence imaging. Importantly, irradiation of the BT-Au-NPs during imaging also caused heating by the photothermal effect (as seen in Fig. 1e). The same procedure was followed in separate mice to study effects on mouse skin in the absence of tumor.

The target site was irradiated continuously (820 nm, 45 mW/cm², <100 fs pulse) for 15-20 min. Images were acquired by intravital microscopy (Fig. 4a, movie S1) and the real-time data were extracted for analysis (Fig. 4b-d). We evaluated the time courses of the permeation of TAMRA-dextran which was determined by measuring the area of TAMRA-dextran (red color) in the images (Fig. 4b), the accumulation of BT-Au-NPs by measuring the area of BT-Au-NPs (yellow color) in the images (Fig. 4c), and vessel dilation by measuring vessel diameters in the images (Fig. 4d). In the tumor tissue, we observed enhanced TAMRA-dextran (over 10% increase in average area after 800 s [~13 min], Fig. 4b) along with increased accumulation of BT-Au-NPs (~ 10% of the whole image area after 800 s, Fig. 4c). The TAMRA-dextran permeation area began to increase around the same time when BT-Au-NPs began to accumulate, which precedes the time when vessel diameter began to increase. The increase of TAMRA-dextran permeation area is attributed to both vessel dilation and extravasation, with the latter being the dominant cause before 500 s, during which time TAMRA permeation area increases much faster than vessel diameter. However, there was considerable heterogeneity in the time course of all three phenomena observed in Fig. 4 b-d, consistent with the known heterogeneous nature of tumors[31].

Figure 4.

BT-Au-NPs imaging and microvascular distribution. Time-lapse microscopy images from mouse 4T1 tumors after intravenous injection with BT-Au-NPs or BT-NPs (laser for SHG: 820 nm, 45 mW/cm², <100 fs pulse). Images are the merger of 3 separate channels, red: TAMRA-dextran used to demarcate blood vessels, blue: SHG imaging, and green: two photon luminescence. BT-Au-NPs (arrows) in blood are yellow (a combination of green, blue and red) and are easily distinguishable from collagen (blue only). (b) - (d) show the temporal evolution under continuous NIR irradiation of (b) TAMRA-dextran permeation (by measuring the area of TAMRA-dextran in the image), (c) accumulation of BT-Au-NPs or BT-NPs accumulation (by measuring the area of NPs [yellow color]) and (c) vessel dilation (by measuring the vessel diameters). Data are median ± interquartile range (n=4).

To investigate the role of the heat generated from BT-Au-NPs, BT-NPs (particles without Au shells, that do not generate heat with irradiation) were injected intravenously and the tumor was imaged under the same conditions as with the BT-Au-NPs. No vessel dilation or particle accumulation was observed (Fig. 4b-d, movie S1, S2). The heat generated from the irradiation itself did not increase the permeation of TAMRA-dextran (< 2.5 % after 15 min), and did not dilate tumor vessels (Fig. S3, movie S3).

In the skin, permeation of TAMRA-dextran was enhanced to a much lesser degree and no obvious accumulation of BT-Au-NPs was observed (Fig. S4, movie S4), presumably because the vasculature in the skin was not leaky. Blood vessel dilation did not correlate with BT-Au-NP accumulation or TAMRA extravasation.

4. DISCUSSION AND CONCLUSIONS

We have demonstrated that BT-Au-NP can provide therapeutically effective PTT in vivo, and provide imaging by SHG and TPL that can be used to monitor the process in vivo.

The size reduction of photothermal transducers is crucial in PTT as it enhances light absorption and nanoparticle heating. The size reduction is also important for enhancing nanoparticle infiltration into tumors.[32, 33] Although the BT-Au-NPs used here were small, the maintained the dielectric-core Au-shell structure of conventional silica-core-gold-nanoshells, which provides ease of surface functionalization. Compared with other theranostic agents that require attaching an external imaging probe to the therapeutic carrier, BT-Au-NPs have the advantage of exploiting the intrinsic properties of the PTT agent, which can better preserve the integrity of the nanostructure and yield a simpler system.

The combination of SHG from the BaTiO₃ core and TPL from the Au shell makes BT-Au-NPs appealing contrast agents for real-time in vivo imaging, as it produces intense signal at relatively low optical excitation. The dual imaging modalities allowed us to track the BT-Au-NPs in real time and explore their thermally-induced effect on tumors. We observed a significant accumulation of the BT-Au-NPs within a few minutes. We speculate that the increased accumulation of NPs in irradiated tumor tissue was due to a combination of photothermal processes and the intrinsic properties of the tumor vasculature. The photothermal effects could include heat-induced vasodilation and/or thermal effects on the blood vessels themselves; vascular injury can enhance extravasation by a phenomenon comparable to EPR[34]. The fact that particle accumulation and extravasation of particle or TAMRA-dextran were not seen in skin – where the irradiance and blood BT-Au-NP concentration were presumably the same as in tumor - suggest that photothermal events were necessary but not sufficient for the observed events to occur; underlying abnormality of the tumor vasculature had to contribute.[35, 36] Nanoparticles travel more slowly in tortuous tumor vessels[37], which could result in more heat dissipation in the local environment, which in turn could lead to more particle accumulation and vessel dilation.

We have shown that irradiation of tumors in animals injected intravenously with BaTiO₃ nanoparticles can result in a therapeutically efficacious photothermal effect at irradiances the did not result in thermal injury (1.2 W/cm² for 3 min). These particles could also be a useful tool for in vivo imaging to dissect the local performance of nanomaterials under PTT or other therapies.

Statement of Significance.

Photothermal therapy induced by plasmonic nanoparticles has emerged as a promising approach to treating cancer. However, the study of the role of intratumoral nanoparticle distribution in mediating tumoricidal activity has been hampered by the lack of suitable imaging techniques. This work describes metal-shell (Au) dielectric-core (BaTiO₃) nanoparticles (abbreviated as BT-Au-NP) for photothermal therapy and bimodal imaging. We demonstrated that sub-100 nm BT-Au-NP can efficiently absorb near infrared light and convert it to heat to ablate tumors. The intrinsic dual imaging capability allowed us to investigate the distribution of the nanoparticles in relation to the tumor vasculature morphology during photothermal ablation, enabling in vivo real-time tracking of the BT-Au-NPs and observation of their thermally-induced effect on tumor vessels.