Gd2O3-mesoporous silica/gold nanoshells: A potential dual T1/ T2 contrast agent for MRI-guided localized near-IR photothermal therapy

Abstract

A promising clinical trial utilizing gold-silica core-shell nanostructures coated with polyethylene glycol (PEG) has been reported for near-infrared (NIR) photothermal therapy (PTT) of prostate cancer. The next critical step for PTT is the visualization of therapeutically relevant nanoshell (NS) concentrations at the tumor site. Here we report the synthesis of PEGylated Gd₂O₃-mesoporous silica/gold core/shell NSs (Gd₂O₃-MS NSs) with NIR photothermal properties that also supply sufficient MRI contrast to be visualized at therapeutic doses (≥10⁸ NSs per milliliter). The nanoparticles have r₁ relaxivities more than three times larger than those of conventional T₁ contrast agents, requiring less concentration of Gd³⁺ to observe an equivalent signal enhancement in T₁-weighted MR images. Furthermore, Gd₂O₃-MS NS nanoparticles have r₂ relaxivities comparable to those of existing T₂ contrast agents, observed in agarose phantoms. This highly unusual combination of simultaneous T₁ and T₂ contrast allows for MRI enhancement through different approaches. As a rudimentary example, we demonstrate T₁/T₂ ratio MR images with sixfold contrast signal enhancement relative to its T₁ MRI and induced temperature increases of 20 to 55 °C under clinical illumination conditions. These nanoparticles facilitate MRI-guided PTT while providing real-time temperature feedback through thermal MRI mapping.

Keywords: MRI contrast; gadolinium oxide; gold nanoshells; mesoporous silica; photothermal.

Fig. 1.

Synthesis and characterization of 2k-PEGylated Gd₂O₃-MS NSs. (A) A schematic representation of the steps involved in synthesizing 2k-PEGylated MRI active Gd₂O₃-MS NS. A 95 ± 16-nm diameter of MS silica particles was incubated with ∼2 nm Gd₂O₃ NPs and sealed with Au shell to form Gd₂O₃-MS NSs. The Gd₂O₃-MS NS nanoparticle was stabilized by adsorbing 2 kDa methoxy PEG molecules onto its surface through their thiol group. (B) TEM images corresponding to each step of the 2k-PEGylated Gd₂O₃-MS NS synthesis: (i and ii) cores of silica (i) before and (ii) after CTAC removal through 4 h of baking at 500 °C, (iii) Gd₂O₃-MS, (iv) Au-seeded Gd₂O₃-MS, and (v) 2k-PEGylated Gd₂O₃-MS NS. (Scale bar: 50 nm.) (C) (i) A high-resolution STEM image of a 2k-PEGylated Gd₂O₃-MS NS sample with STEM-HAADF imaging mode and its elemental mapping images correspond to (ii) Si, (iii) Gd, and (iv) Au atoms present in the sample. (Scale bar: 200 nm.) (D) Gold-shell thickness distributions of three samples of 2k-PEGylated Gd₂O₃-MS NSs determined by high resolution TEM image analysis. The distributions were fitted with a Gauss function utilizing OriginPro 2021b, giving an average shell thickness ± SE of (i) 22.3 ± 0.1 nm (305 NSs), (ii) 27.5 ± 0.1 nm (339 NSs), and (iii) 31.0 ± 0.2 nm (297 NSs). (E) Their corresponding extinction spectra in Milli-Q water with the same color code (blue, 22.3 nm; green, 27.5 nm; and red, 31.0 nm).

Fig. 2.

MRI characterization of 2k-PEGylated Gd₂O₃-MS NSs in Milli-Q water at 4.7 T, 20 °C. (A–D) Longitudinal magnetization recovery (T₁), transverse magnetization decay (T₂), and their T₁w and T₂w MR images, respectively, at various gadolinium concentrations of Gd₂O₃-MS NSs with a gold shell thickness of 22.3 ± 0.1 nm. Dots are experimental data ± SD and solid lines are the fit functions using Eq. 1. The data in A and B were obtained with a RARE spin–echo sequence that loops across a range of repetition time values and MSME spin–echo scanning, respectively. T₁w MR imaging parameters: TR = 400 ms, TE = 9.9 ms, flip angle (FA) = 180°. T₂w MR imaging parameters: TR = 1,750 ms, TE = 423 ms, FA = 180°. (E) The inverse of T₁ (R₁, open red circles) and the inverse of T₂ (R₂, open green triangle) relaxivity rate constants at various gadolinium concentrations gave slope values of r₁ = 15.8 ± 0.6 mM_Gd⁻¹·s⁻¹ and r₂ = 120 ± 5 mM_Gd⁻¹·s⁻¹ (mean ± SE), respectively, for Gd₂O₃-MS NSs (22.3 ± 0.1 nm Au). (F) r₁ (red) and r₂ (green) relaxivity rates, and their r₂/r₁ ratio (blue) for an aqueous suspension of 2k-PEGylated Gd₂O₃ as synthesized (15.8 ± 0.7 mM_Gd⁻¹·s⁻¹, 30.6 ± 0.3 mM_Gd⁻¹·s⁻¹, 1.94 ± 0.09), Au-seeded Gd₂O₃-MS (19 ± 1 mM_Gd⁻¹·s⁻¹, 132 ± 6 mM_Gd⁻¹·s⁻¹, 6.9 ± 0.5), and Gd₂O₃-MS NSs with gold shell thicknesses of 22.3 ± 0.1 nm (15.8 ± 0.6 mM_Gd⁻¹·s⁻¹, 120 ± 5 mM_Gd⁻¹·s⁻¹, 7.6 ± 0.4), 27.5 ± 0.1 nm (8.3 ± 0.2 mM_Gd⁻¹·s⁻¹, 66 ± 3 mM_Gd⁻¹·s⁻¹, 8.0 ± 0.4), and 31.0 ± 0.2 nm (3.6 ± 0.1 mM_Gd⁻¹·s⁻¹, 41 ± 3 mM_Gd⁻¹·s⁻¹, 11.4 ± 0.9) in comparison with a T₁ contrast agent (Magnevist: 4.7 ± 0.1 mM_Gd⁻¹·s⁻¹, 5.34 ± 0.04 mM_Gd⁻¹·s⁻¹, 1.14 ± 0.03) and T₂ contrast agents (Resovist, 2.8 ± 0.1 mM_Gd⁻¹·s⁻¹, 176 ± 9 mM_Gd⁻¹·s⁻¹, 63 ± 4; Ferumoxide, 2.3 ± 0.1 mM_Gd⁻¹·s⁻¹, 105 ± 5 mM_Gd⁻¹·s⁻¹, 46 ± 3). The blue dashed line is r₂/r₁ ratio = 10. Results show that Gd₂O₃, Au-seeded Gd₂O₃, and Gd₂O₃-MS NSs (Au shell thickness: 22.3 ± 0.1 nm and 27.5 ± 0.1 nm) have statistically higher r₁ relaxivity values than Magnevist (*P < 0.001 vs. Magnevist by ANOVA with Tukey’s honestly significant difference post hoc test). (G) Enhancement factors r₁ (red) and r₂ (green) relative to Magnevist. Relaxivity values Resovist and Ferumoxide were taken from reference (41). (H) r₁ relaxivity values versus gold shell thickness. The SBM theoretical data (black solid line) with 3.3 × 10⁶ Gd³⁺ per Gd₂O₃-MS NS is in good agreement with the experimental data (red symbols), r² = 0.74. Error bars are represented in E–H as ± SE.

Fig. 3.

MR imaging processing of T₁w/T₂w intensity ratio with 2k-PEGylated Gd₂O₃-MS NSs (27.5 ± 0.1 nm gold shell) in 0.48% agarose phantoms performed under 4.7 T magnetic strength. (A and B) T₁ and T₂ signal contrast for agarose phantoms consisting of 1.1 × 10⁹ (C₄, red), 2.1 × 10⁹ (C₃, green), 4.2 × 10⁹ (C₂, blue), and 8.4 × 10⁹ (C₁, cyan) Gd₂O₃-MS NSs per milliliter. The profiles were obtained by subtracting the fitted longitudinal recovery and the fitted transverse decay, respectively, for each Gd₂O₃-MS NS dilution from Milli-Q water. Orange dashed lines indicate the repetition time (TR = 1,580 ms) and the echo time (TE = 250 ms) at which the optimum T₁ and T₂ signal contrast, respectively, can be achieved. (C and D) T₂w MR images (Upper, gray scale; Lower, color scale) and their corresponding signal intensities at echo times = 83 ms (red), 250 ms (green), and 425 ms (purple) for various concentrations of 2k-PEGylated Gd₂O₃-MS NSs relative to that for Milli-Q water. (E) MR images acquired with TE = 9.9 ms and TR = 1,580 ms (i) and images of the processed T₁w/T₂w intensity ratio with T₂w MRI at TR = 3,857 ms and TE = 83 ms (ii), 250 ms (iii), and 425 ms (iv) for 2k-PEGylated Gd₂O₃-MS NSs at various concentrations. (F) Their corresponding MRI signal intensities relative to water signal (= 1). The values and the error bars are the mean ± SD. The data for T₁w MR images and T₂w MR images were obtained utilizing RARE sequence (TR = 1,580 ms, TE = 9.9 ms, flip angle = 180°) and MSME (TR = 3,857 ms; TE = 83 ms, 250 ms, or 425 ms; and flip angle = 180°), respectively.

Fig. 4.

Experimental setup for 2k-PEGylated Gd₂O₃-MS NSs-assisted MRI-guided PTT. (A, Lower) A schematic representation of the MRI-guided photothermal illumination setup showing the sample placed in the active field of the 72-mm ID MRI coil and water sample as the reference. The samples were illuminated with a CW-GaAlAs Laser diode (810 ± 20 nm) coupled with Visualase laser diffusing fiber. (A, Upper) Set of snapshot images of (i) a diffusing tip illumination profile of the 635 to 655 nm aiming beam illumination and (ii) a superimposed image of the sample with a cartoon image of the diffuser tip illustrating the way it was set up during the thermal MRI measurements. (B, Lower) Coronal T₁w MR image obtained before PTT with a fast-spoiled gradient-echo (fSPGR) sequence, TR = 275 ms, TE = 5.1 ms, number of averages = 2, FA = 90°, FOV = 35 mm, and matrix size = 256 × 256. (B, Upper) Axial T₁w MR images at the (i) 2k-PEGylated Gd₂O₃-MS NSs location (slice 2), consisting of 1.1 × 10⁹ (C₄), 2.1 × 10⁹ (C₃), 4.2 × 10⁹ (C₂), and 8.4 × 10⁹ (C₁) Gd₂O₃-MS NSs per milliliter and (ii) NS-free agarose (slice 4) were taken during PTT with a FLASH sequence. TR = 577 ms, TE = 6 ms, number of averages = 1, FA = 35°, FOV = 35 mm, matrix size = 128 × 128, slice thickness = 1.5 mm, and spacing between slices = 1.75 mm (scale bar = 2.6 mm).

Fig. 5.

Temperature change measurements with thermal MRI method at 4.7 T of various concentrations of 2k-PEGylated Gd₂O₃-MS NSs (27.5-nm gold shell) in 0.48% agarose phantoms under 3 min of 810-nm illumination at 4 W/1 W. (A) Maximum temperature changes across the four NMR tubes at the 2k-PEGylated Gd₂O₃-MS NS locations (slice 2 in Fig. 4B, XY plane), consisting of 1.1 × 10⁹ (C₄), 2.1 × 10⁹ (C₃), 4.2 × 10⁹ (C₂), and 8.4 × 10⁹ (C₁) Gd₂O₃-MS NSs per milliliter. The thermal MRI mapping was obtained using a MATLAB code by processing the last T₁w MRI collected before the laser illumination was turned OFF. (B) Temperature change profile during the treatment of three areas of interest: 2k-PEGylated Gd₂O₃-MS NSs with 8.4 × 10⁹ NSs per milliliter (red, slice 2), ∼2 mm below the 2k-PEGylated Gd₂O₃-MS NSs where the agarose medium was NS-free (green, slice 4), and the controlled measurements–Milli-Q water (black, slice 2). Each data point at time x represents the maximum temperature change calculated based on the phase change in T₁w MR image collected at time x relative to the first T₁w MR image. A two-dimensional fast-low-angle-shot (FLASH) gradient-echo scanning sequence was used to acquire time series of 14 T₁w images, 1 min and 13 s per scan. Solid lines are a guide to the eye. The red shaded area represents the illumination phase; its left side is the baseline phase, and its right side is the recovery phase. (C and D) Maximum temperature changes reached at the 2k-PEGylated Gd₂O₃-MS NS locations with various concentrations and NP-free agarose (∼2 mm below the 2k-PEGylated Gd₂O₃-MS NSs band) after 3 min of NIR illumination at 4- and 1-W laser powers, respectively (***P < 0.0001, **P < 0.001, *P < 0.05 utilizing ANOVA with Tukey’s honestly significant difference post hoc test). Error bars represent ± SE.