Microwave-Synthesized Platinum-Embedded Mesoporous Silica Nanoparticles as Dual-Modality Contrast Agents: Computed Tomography and Optical Imaging

Nanoparticle-based imaging contrast agents have drawn tremendous attention especially in multi-modality imaging. In this study, we developed mesoporous silica nanoparticles (MSNs) for use as dual-modality contrast agents for computed tomography (CT) and near-infrared (NIR) optical imaging (OI). A microwave synthesis for preparing naked platinum nanoparticles (nPtNPs) on MSNs (MSNs-Pt) was developed and characterized with physicochemical analysis and imaging systems. The high density of nPtNPs on the surface of the MSNs could greatly enhance the CT contrast. Inductively coupled plasma mass spectrometry (ICP-MS) revealed the MSNs-Pt compositions to be ~14% Pt by weight and TEM revealed an average particle diameter of ~50 nm and covered with ~3 nm diameter nPtNPs. To enhance the OI contrast, the NIR fluorescent dye Dy800 was conjugated to the MSNs-Pt nanochannels. The fluorescence spectra of MSNs-Pt-Dy800 were very similar to unconjugated Dy800. The CT imaging demonstrated that even modest degrees of Pt labeling could result in substantial X-ray attenuation. In vivo imaging of breast tumor-bearing mice treated with PEGylated MSNs-Pt-Dy800 (PEG-MSNs-Pt-Dy800) showed significantly improved contrasts in both fluorescence and CT imaging and the signal intensity within the tumor retained for 24 h post-injection.


Introduction
Nanoparticles (NPs)-based imaging contrast agents have emerged as a vigorous research area with agents such as iron oxide for magnetic resonance imaging (MRI) [1][2][3], quantum dots for optical imaging [4,5], and gold NPs for X-ray computed tomography (CT) [6][7][8] being developed. Although NPs can be modified with specific targeting molecules to enhance their delivery to tumors and, therefore, increase tumor contrast, a high percentage of administered NPs will nonetheless accumulate in tissues/organs other than the targeted tumor. Substantial amounts of NPs residing in tissues/organs that do not undergo further biodegradation or excretion could cause unwanted cytotoxicity and other side effects [9][10][11][12][13][14][15]. Thus, defining the characteristics of synthetic NPs that govern their biodegradation and/or excretion in vivo is a prerequisite for the successful application of such NPs as imaging contrast agents in clinical applications. In this study, we aimed to develop dual-modality NP-based Large-pore MSNs were synthesized; a TEM image of these MSNs is shown in Figure 1a. Using n-octane as a swelling agent to expand the inner micelle space, we observed that the as-synthesized MSNs possessed both large pore diameters (~5.0 nm) and well-ordered pore structures. The large-pore MSNs had sphere-like shapes and small sizes with average particle diameters of approximately 50 nm. We modified the primary amine group in the channels and on the surface of MSNs. Before the pores were extracted, the adsorption of Pt 4+ ions to the surfaces of the aminosilane-modified silica nanoparticles was achieved using electrostatic attractions between the positively charged surface of the amine-modified silica and the negative charge of the PtCl 6− ions. Using microwave irradiation, the PtCl 6− ions were fully converted to Pt 0 NPs on the MSN surface. After microwave reduction, TEM images of the nPtNP-embedded MSNs were collected, as shown in Figure 1b. The size of the nPtNPs was approximately 3 nm, and the weight percentage of Pt on the outer surface of MSNs was calculated to be 14 % using ICP-MS. The naked nPtNPs on the MSNs were easily modified by thiol groups. Thiol-terminated polyethylene glycol (PEG) could be conjugated to the nPtNPs to enhance their dispersibility in bio-media and to prevent protein adsorption onto the MSNs-Pt NPs. Moreover, many target peptides that have thiol side chains, such as cysteine, can also be attached to the MSNs-Pt. The NIR fluorescence dye used for the OI modality, Dy800, was encapsulated into the nanochannels of the MSNs-Pt. The process involved using the Dy800-NHS ester to react with aminosilane, followed by mixing with the pore-extracted MSNs-Pt to yield MSNs-Pt-Dy800. Figure 2a presents the energy-dispersive X-ray spectrum (EDX) of the MSNs-Pt. We observed three peaks at 2, 9.2, and 11 KeV that represent the composition of Pt and the peak from 1.9 KeV represents the Si in the spectrum. Pt elements were clearly observed in the spectrum, providing evidence for the incorporation of the nPtNPs onto the MSNs. Figure 2b presents the zeta potentials of the amine-modified MSNs (MSNs-NH2), MSNs-NH2-Pt, and PEG-modified MSNs-Pt (PEG-MSNs-Pt) at pH values ranging from 2 to 10. The zeta potential of the MSN-NH2-Pt was lower than that of the MSNs-NH2 in the pH range from 7~10 because of the positively charged functional groups conjugated to the surface and because the Pt:aminopropyl groups on the surfaces of the MSNs were bonded with the Pt NPs. After PEG modification, the MSNs-Pt-PEG possessed a neutral surface charge. Figure 2a presents the energy-dispersive X-ray spectrum (EDX) of the MSNs-Pt. We observed three peaks at 2, 9.2, and 11 KeV that represent the composition of Pt and the peak from 1.9 KeV represents the Si in the spectrum. Pt elements were clearly observed in the spectrum, providing evidence for the incorporation of the nPtNPs onto the MSNs. Figure 2b presents the zeta potentials of the amine-modified MSNs (MSNs-NH 2 ), MSNs-NH 2 -Pt, and PEG-modified MSNs-Pt (PEG-MSNs-Pt) at pH values ranging from 2 to 10. The zeta potential of the MSN-NH 2 -Pt was lower than that of the MSNs-NH 2 in the pH range from 7~10 because of the positively charged functional groups conjugated to the surface and because the Pt:aminopropyl groups on the surfaces of the MSNs were bonded with the Pt NPs. After PEG modification, the MSNs-Pt-PEG possessed a neutral surface charge.  The surface areas and pore size distributions of the MSNs were characterized using the Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods. The nitrogen adsorption-desorption plot of the MSNs (Figure 3a-i) exhibited a typical type IV isotherm and a pore size distribution centered at 3.6 nm. The surface area and pore volume were calculated to be 1073.63 m 2 g −1 and 0.88 cm 3 g −1 , respectively. The surface areas of the MSNs-NH2 and MSNs-Pt decreased to 844.15 m 2 g −1 and 649.77 m 2 g −1 , respectively (as shown in Figure 3a ii and iii). The majority of the Pt NPs were reduced on the surface of the MSNs to leave sufficient surface area for drug delivery, catalysis, and other applications. Thermogravimetric analysis (TGA) was employed to identify the surface modification. Figure 3b indicates that PEG accounted for 24.5% of the total mass of the PEG-MSNs-Pt NPs. This high percentage of PEG coating the MSNs can provide a longer circulation time for in vivo optical/CT imaging.  The surface areas and pore size distributions of the MSNs were characterized using the Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods. The nitrogen adsorption-desorption plot of the MSNs (Figure 3a-i) exhibited a typical type IV isotherm and a pore size distribution centered at 3.6 nm. The surface area and pore volume were calculated to be 1073.63 m 2 g −1 and 0.88 cm 3 g −1 , respectively. The surface areas of the MSNs-NH 2 and MSNs-Pt decreased to 844.15 m 2 g −1 and 649.77 m 2 g −1 , respectively (as shown in Figure 3a-ii,a-iii). The majority of the Pt NPs were reduced on the surface of the MSNs to leave sufficient surface area for drug delivery, catalysis, and other applications. Thermogravimetric analysis (TGA) was employed to identify the surface modification. Figure 3b indicates that PEG accounted for 24.5% of the total mass of the PEG-MSNs-Pt NPs. This high percentage of PEG coating the MSNs can provide a longer circulation time for in vivo optical/CT imaging.  The surface areas and pore size distributions of the MSNs were characterized using the Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods. The nitrogen adsorption-desorption plot of the MSNs (Figure 3a-i) exhibited a typical type IV isotherm and a pore size distribution centered at 3.6 nm. The surface area and pore volume were calculated to be 1073.63 m 2 g −1 and 0.88 cm 3 g −1 , respectively. The surface areas of the MSNs-NH2 and MSNs-Pt decreased to 844.15 m 2 g −1 and 649.77 m 2 g −1 , respectively (as shown in Figure 3a ii and iii). The majority of the Pt NPs were reduced on the surface of the MSNs to leave sufficient surface area for drug delivery, catalysis, and other applications. Thermogravimetric analysis (TGA) was employed to identify the surface modification. Figure 3b indicates that PEG accounted for 24.5% of the total mass of the PEG-MSNs-Pt NPs. This high percentage of PEG coating the MSNs can provide a longer circulation time for in vivo optical/CT imaging.

In Vitro Optical and CT Images of Dual-Modality MSNs
The optical properties of the MSNs-Pt-Dy800 are shown in Figure 4a. The fluorescence spectrum of MSNs-Pt-Dy800 was similar to that of Dy800, with the maximum emission at 783 nm. The optical image of MSNs-Pt-Dy800 sample was collected at an excitation of 770 nm, and pronounced NIR fluorescence was observed, as shown in the inset of Figure 4a. For the second modality, the CT imaging of the tube array composed of various MSNs samples is presented in Figure 4b. Compared to that of MSNs ( The optical properties of the MSNs-Pt-Dy800 are shown in Figure 4a. The fluorescence spectrum of MSNs-Pt-Dy800 was similar to that of Dy800, with the maximum emission at 783 nm. The optical image of MSNs-Pt-Dy800 sample was collected at an excitation of 770 nm, and pronounced NIR fluorescence was observed, as shown in the inset of Figure 4a. For the second modality, the CT imaging of the tube array composed of various MSNs samples is presented in Figure 4b. Compared to that of MSNs (  The in vitro CT imaging of the tube array is illustrated in Figure 5a. As expected, the scans demonstrated a positive contrast enhancement in a dose-dependent manner. Appreciable enhancement for CT contrast was observed using 1.25 mg mL −1 of PEG-MSNs-Pt. The calculated CT signal contrast intensities within the range of 1.25~20 mg mL −1 of PEG-MSNs-Pt are plotted in Figure  5b. The CT signal using 20 mg mL −1 of PEG-MSNs-Pt (corresponding to 2.8 mg of Pt NPs) exhibited an intensity equivalent to that of the commercial iodine-containing CT contrast agent Visipaque TM at 20 mg mL −1 (360 HU). The in vitro CT imaging of the tube array is illustrated in Figure 5a. As expected, the scans demonstrated a positive contrast enhancement in a dose-dependent manner. Appreciable enhancement for CT contrast was observed using 1.25 mg mL −1 of PEG-MSNs-Pt. The calculated CT signal contrast intensities within the range of 1.25~20 mg mL −1 of PEG-MSNs-Pt are plotted in Figure 5b. The CT signal using 20 mg mL −1 of PEG-MSNs-Pt (corresponding to 2.8 mg of Pt NPs) exhibited an intensity equivalent to that of the commercial iodine-containing CT contrast agent Visipaque TM at 20 mg mL −1 (360 HU).

Biocompatibility and Cellular Uptake of MSNs-Pt NPs
The in vitro cytotoxicity of the PEG-MSNs-Pt NPs was evaluated in MDA-MB-231 breast cancer cells using an MTT assay. The water-soluble PEG-MSNs-Pt NPs were tested at concentrations of 25, 50, and 100 µg mL −1 for incubation times of 1, 2, 6, and 24 h. The MTT assay (Figure 6a) showed that only minor cytotoxic responses were detected at concentrations below 50 µg mL −1 after 24 h of exposure. Furthermore, at the highest concentration of 100 µg mL −1 , the cell viability decreased to 65% at 6 and 24 h. With regard to the cellular uptake, the TEM image of the cell section showed that the MSNs-Pt NPs were located in the endosomes after 24 h of incubation (Figure 6b). Some of the MSNs-Pt began to escape from the endosome to the cytoplasm (inset Figure 6b). Our results revealed high cell viability and cellular uptake of the MSNs-Pt NPs, which suggest that they possess excellent in vitro biocompatibility.

Biocompatibility and Cellular Uptake of MSNs-Pt NPs
The in vitro cytotoxicity of the PEG-MSNs-Pt NPs was evaluated in MDA-MB-231 breast cancer cells using an MTT assay. The water-soluble PEG-MSNs-Pt NPs were tested at concentrations of 25, 50, and 100 µg mL −1 for incubation times of 1, 2, 6, and 24 h. The MTT assay (Figure 6a) showed that only minor cytotoxic responses were detected at concentrations below 50 µg mL −1 after 24 h of exposure. Furthermore, at the highest concentration of 100 µg mL −1 , the cell viability decreased to 65% at 6 and 24 h. With regard to the cellular uptake, the TEM image of the cell section showed that the MSNs-Pt NPs were located in the endosomes after 24 h of incubation (Figure 6b). Some of the MSNs-Pt began to escape from the endosome to the cytoplasm (inset Figure 6b). Our results revealed high cell viability and cellular uptake of the MSNs-Pt NPs, which suggest that they possess excellent in vitro biocompatibility.

Biocompatibility and Cellular Uptake of MSNs-Pt NPs
The in vitro cytotoxicity of the PEG-MSNs-Pt NPs was evaluated in MDA-MB-231 breast cancer cells using an MTT assay. The water-soluble PEG-MSNs-Pt NPs were tested at concentrations of 25, 50, and 100 µg mL −1 for incubation times of 1, 2, 6, and 24 h. The MTT assay (Figure 6a) showed that only minor cytotoxic responses were detected at concentrations below 50 µg mL −1 after 24 h of exposure. Furthermore, at the highest concentration of 100 µg mL −1 , the cell viability decreased to 65% at 6 and 24 h. With regard to the cellular uptake, the TEM image of the cell section showed that the MSNs-Pt NPs were located in the endosomes after 24 h of incubation (Figure 6b). Some of the MSNs-Pt began to escape from the endosome to the cytoplasm (inset Figure 6b). Our results revealed high cell viability and cellular uptake of the MSNs-Pt NPs, which suggest that they possess excellent in vitro biocompatibility.

In Vivo Optical and CT Images of Dual-Modality MSNs
In vivo molecular imaging was performed in the MDA-MB-231 breast tumor animal model. PEG-MSNs-Pt (1.8 mg mL −1 Pt concentration, 150 µL) was injected into mice intratumorally. An OI image was acquired after injection (24 h), as shown in Figure 7a. Significant optical contrast was observed, and the NPs aggregated in the tumor site 24 h post-injection. On the other hand, significant CT contrast enhancement in the tumor was observed at 24 h, which might be due to the enhanced permeability and retention effect (red circle in Figure 7b). The CT contrast generated by the PEG-MSNs-Pt was similar to that of bone. The CT contrast of the PEG-MSNs-Pt was retained at 24 h post-injection. The PEG-MSNs-Pt exhibited great potential as a dual-modality molecular imaging contrast agent and can allow drug loading for therapeutics in future applications.

In Vivo Optical and CT Images of Dual-Modality MSNs
In vivo molecular imaging was performed in the MDA-MB-231 breast tumor animal model. PEG-MSNs-Pt (1.8 mg mL −1 Pt concentration, 150 µL) was injected into mice intratumorally. An OI image was acquired after injection (24 h), as shown in Figure 7a. Significant optical contrast was observed, and the NPs aggregated in the tumor site 24 h post-injection. On the other hand, significant CT contrast enhancement in the tumor was observed at 24 h, which might be due to the enhanced permeability and retention effect (red circle in Figure 7b). The CT contrast generated by the PEG-MSNs-Pt was similar to that of bone. The CT contrast of the PEG-MSNs-Pt was retained at 24 h post-injection. The PEG-MSNs-Pt exhibited great potential as a dual-modality molecular imaging contrast agent and can allow drug loading for therapeutics in future applications.

Discussion
PEG-MSNs-Pt-Dy800 showed a great contrast efficacy in OI/CT. Similarly, in previous reports, different sized FePt NPs were discussed, and their CT/Magnetic Resonance Imaging (MRI) contrast efficacies were compared. Smaller FePt NPs presented a higher concentration than large ones in mouse brains. The large FePt NPs exhibited longer circulation time in the blood, as well as a better contrast effect in CT/MRI [35]. These kinds of NPs possess great potential for translation of clinical diagnosis. Herein, we developed a dual-modality nanoplatform with enhanced contrasts for OI and CT to achieve the precision diagnosis of tumors in vivo. Such a nanoplatform consisted of nPtNPs reduced on the surface of MSNs and Dy800 loaded in the nanochannels of MSNs; further, PEG modification improved the circulation time in blood providing favorable pharmacokinetics. It allows us to efficiently perform fast tracing of NPs using optical imaging systems, followed by confirmation of orientation using CT.
It has been indicated that different hybrid NPs containing PtNPs possess contradictory effects, such as cell protection and cytotoxicity. It is worth noting that PtNPs exhibit high catalytic activity, scavenging the singlet oxygen and superoxide to protect cells [36][37][38]. On the other hand, nanoparticulate Pt (nano-Pt) showed cisplatin-like function to inhibit cancer cell progression, cytotoxicity and less drug resistance [41]. Previous reports also used microwave-assisted methods to synthesize mesoporous silica aerogel-supported PtNPs (SAP) for a proton exchange membrane fuel cell [42]. Similar to our design, a group reported nPtNPs within the nanochannel of mesoporous silica nanospheres as a catalyst for hydrolysis [43]. In this regard, we used MSNs with nPtNPs on the surface, which were much smaller than those reported previously. We suggest that

Discussion
PEG-MSNs-Pt-Dy800 showed a great contrast efficacy in OI/CT. Similarly, in previous reports, different sized FePt NPs were discussed, and their CT/Magnetic Resonance Imaging (MRI) contrast efficacies were compared. Smaller FePt NPs presented a higher concentration than large ones in mouse brains. The large FePt NPs exhibited longer circulation time in the blood, as well as a better contrast effect in CT/MRI [35]. These kinds of NPs possess great potential for translation of clinical diagnosis. Herein, we developed a dual-modality nanoplatform with enhanced contrasts for OI and CT to achieve the precision diagnosis of tumors in vivo. Such a nanoplatform consisted of nPtNPs reduced on the surface of MSNs and Dy800 loaded in the nanochannels of MSNs; further, PEG modification improved the circulation time in blood providing favorable pharmacokinetics. It allows us to efficiently perform fast tracing of NPs using optical imaging systems, followed by confirmation of orientation using CT.
It has been indicated that different hybrid NPs containing PtNPs possess contradictory effects, such as cell protection and cytotoxicity. It is worth noting that PtNPs exhibit high catalytic activity, scavenging the singlet oxygen and superoxide to protect cells [36][37][38]. On the other hand, nanoparticulate Pt (nano-Pt) showed cisplatin-like function to inhibit cancer cell progression, cytotoxicity and less drug resistance [41]. Previous reports also used microwave-assisted methods to synthesize mesoporous silica aerogel-supported PtNPs (SAP) for a proton exchange membrane fuel cell [42]. Similar to our design, a group reported nPtNPs within the nanochannel of mesoporous silica nanospheres as a catalyst for hydrolysis [43]. In this regard, we used MSNs with nPtNPs on the surface, which were much smaller than those reported previously. We suggest that nPtNPs on MSNs display a mechanism similar to anti-cancer platinum drugs, such as cisplatin. One important mechanism of cisplatin toxicity is increased oxidative stress. Cisplatin generates reactive oxygen species that induce cell death [44]. Among the series cell death pathway, the intracellular glutathione plays a crucial role in these mechanisms, as cisplatin-glutathione conjugates reduce the platinum binding to DNA and transport out of the cells protects dividing cells from cisplatin toxicity [45]. Besides the pronounced contrasts of dual-modality imaging, in our study, we also attempted to monitor the induced cytotoxic effects involving intracellular glutathione (GSH) and related enzymes' variations after treatment with MSNs-Pt. Our preliminary results implicated that MSNs-Pt have much higher therapeutic efficacy than PEG-MSNs-Pt in both in vitro and in vivo experiments. Likewise, it was demonstrated that nano-Pt triggered the autophagy and immunological responses in the tumor area [41]. These preclinical experiments are currently ongoing; we are focusing on the underlying molecular mechanisms of therapeutic effects rendered by MSNs-Pt with the nPtNP distributions on the different topological domains of MSNs, i.e., nPtNPs reduced on the MSN surface and surface/channels, etc.
In summary, we reported on the microwave synthesis of nPtNPs on the surface of NIR dye-doped MSNs for dual-modality imaging of OI and CT. The PEG-stabilized NPs were very stable in aqueous solution and did not affect cell viability; furthermore, they provided an active surface zone for further conjugation of the targeting motif. The CT contrast was significantly enhanced following the integration of naked nPtNPs on MSNs. The composite NPs have great potential for use as probes in optical and CT imaging and for use as drug delivery carriers due to the features of MSNs. These dual-modality NPs provide significant CT and optical contrast, and they remained in the tumor for at least 24 h. The novel microwave synthesis of MSNs-Pt provides a simple and rapid approach for the production of nanoparticle-based OI/CT dual-modality contrast agents.

Preparation of MSNs-NH 2 NPs
The large-pore MSNs-NH 2 sample with a sphere-like, well-ordered pore structure was synthesized using a low concentration of TEOS, a surfactant, and a basic (NH 4 OH) catalyst in a two-step preparation. The sol-gel process for the co-condensation of TEOS and the synthesis of the MSNs-NH 2 sample is described as follows. First, cetyltrimethylammonium bromide (CTAB) (0.58 g) was dissolved in NH 4 OH (0.51 M, 300 mL) at 50 • C, and then 4.5 g of n-octane was added under vigorous stirring.

Synthesis of Platinum NPs on the MSNs
The adsorption of Pt 4+ ions on the surfaces of the aminosilane-modified silica nanoparticles was achieved by harnessing electrostatic attractions between the positively charged surface of the amine-modified silica and the negative charge of the PtCl 6− ions. The strong attractions between the ions in each ion pair (-NH 3+ PtCl 6− -) provided stability for the further reduction of the nPtNPs on the surfaces of the silica nanoparticles. The loading solution was mixed by adding 50 mg of MSNs-NH 2 with 50 µL of H 2 PtCl 6 solution (1 mg mL −1 ) in the presence of 50 mL of water. The mixture was stirred at room temperature for 0.5 h. The resulting MSNs-NH 3+ PtCl 6− complexes were washed twice with pure water and then centrifuged at 11000 rpm for 20 min. The MSNs-NH 3+ PtCl 6complexes were reduced by re-suspending the above MSNs-NH 3+ PtCl 6− complexes in 5 mL of H 2 O; subsequently, the solution was exposed to a 300 W fixed-mode microwave synthesis system (Discover SP-D, CEM Tech., Matthews, NC, USA) for 10 min at 120 • C. The MSNs-Pt were collected by centrifuging at 12000 rpm for 20 min, washed, and re-dispersed with deionized water and ethanol several times. The platinum concentration was measured by inductively coupled plasma mass spectrometry.

Preparation of MSNs-Pt-Dy800 Samples
The covalent conjugation of the near-infrared fluorophore DyLight in the MSNs-Pt nanochannels to obtain MSNs-Pt-Dy800 was conducted as follows. First, a Dy800-NHS ester stock solution was prepared by mixing Dy800-NHS ester powder (1 mg) in Dimethylformamide (DMF) (1 mL). Next, 200 µL of the DMF/Dy800 stock solution was added to an extracted MSNs-Pt sample (15 mg), and the mixture was further diluted in 50 mL of DMF. After stirring at room temperature for 2 h, the MSNs-Pt-Dy800 complex was washed twice with Ringer's solution, followed by centrifuging for 20 min at 12000 rpm. The amount of Dy800 loaded in the individual MSNs-Pt samples was determined by measuring the decrease in optical absorption of the residual solution at 777 nm.

PEG Conjugation
First, 100 mg of MSNs-Pt-Dy800 nanoparticles was added to a flask in the presence of 20 mL of a mPEG5000-SH solution (1 mg mL −1 ). After stirring the solution for 20 min at room temperature, the solution was centrifuged, washed three times with water, and suspended in 10 mL of water for in vitro experiments.

Characterization
The ζ-potentials of all the samples were measured using a Malvern Zetasizer 3000 NANOZS (Westborough, MA, USA). The zeta potential distribution was obtained from the average of ten measurements. The morphologies of the samples were characterized by TEM (Hitachi H-7650 operating at an accelerating voltage of 80 kV, Berkshire, UK). Thermogravimetric analysis (TGA) data were obtained with a NETZCH TG209-F3 thermogravimetric analyzer (NETZCH, Germany). The samples were heated from 30 to 650 • C at a heating rate of 15 • C per minute under nitrogen. The Pt content in the samples was measured using inductively coupled plasma mass spectrometry (ICP-MS, PerkinElmer, Akron, Ohio, USA). The sample preparation was as follows: 100 µL of MSN-Pt NP samples was placed in a TFM digestion vessel (CEM Corporation, Matthews, NC, USA). Then, a mixture of 0.5 ml of HNO 3 and 1.5 ml of HCl was added. The sample was heated to 200 • C via microwave irradiation for 20 min. The samples were then transferred to polypropylene (PP) vials and diluted 200 times.

Cell Viability Assay
The cell viability in the presence of the NPs was evaluated using a 3-[4-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT, Sigma, MO, USA) assay. For the MTT assay, MDA-MB-231 cells were seeded in 96-well plates at a density of 1 × 10 4 per well in 200 µL of media and grown overnight. The cells were then incubated with various concentrations of PEG-MSNs-Pt (0, 25, 50, and 100 µg mL −1 ) for 1, 2, 6, and 24 h. Following this incubation, the cells were incubated in media containing 0.1 mg mL −1 of MTT for 1 h. Then, the MTT solution was removed, and the precipitated violet crystals were dissolved in 200 µL of dimethyl sulfoxide (DMSO). The absorbance was measured at 560 nm using a microplate reader.

Cellular Uptake and TEM Imaging
To determine the cellular uptake of nanoparticles, MDA-MB-231 cells were plated in 12-well plates pre-coated with poly-D-lysine containing complete media and 10% FBS at a density 1 × 10 4 cells/well. The dishes were then placed in an incubator overnight at 37 • C and 5% CO 2 to permit cell attachment. The next day, the cells were carefully washed with phosphate-buffered saline (PBS) and refreshed with 3 mL of serum-free medium. Nanoparticles were added to the well and mixed gently at a final concentration of 50 µg mL −1 . The treated cells were then incubated at 37 • C and 5% CO 2 for 6 h. After incubation, the cells were rinsed three times with PBS.
For cell ultra-sections, the treated cells were pre-fixed in 2.5% glutaraldehyde for 1 h and then washed twice with 0.1 M PBS (pH 7.0). Osmium tetroxide (2%) was used for 1 h for post-fixation. Dehydration was then performed using an ascending series of ethanol concentrations at 30%, 50%, 70%, 80%, and 95% for 10 min each, followed by three 100% ethanol dehydrations for 10 min each. The dehydrated samples were then embedded in Spurr's resin and allowed to polymerize for 15 h at 68 • C. Ultrathin sections were cut with a Leica UC6 ultramicrotome and examined with a Hitachi H-7650 TEM at an accelerating voltage of 80 kV without further staining.

In Vivo Optical Imaging
In vivo fluorescence imaging was carried out on a commercial Xenogen/Caliper IVIS-200 In Vivo Imaging System (with isoflurane anesthesia, PerkinElmer, Akron, OH, USA) and on a home-made in vivo optical imaging system (with urethane anesthesia). For the latter, light from a 100 W Hamamatsu xenon lamp was passed through fluorophore-dependent bandpass interference filters (Model 545AF75, Omega Optical Inc., Brattleboro, VT, USA) to induce Dy800 fluorescence. The imaging sensor was a charge-coupled device (CCD) camera (DW436, Andor Technology Ltd, Belfast, Northern Ireland) that was thermoelectrically cooled to −90 • C. Bandpass interference filters (FF495 Ex 02-25, Smerok, New York, NY, USA) were placed over a 50 mm f/1.2 lens (Nikon, Tokyo, Japan) to allow fluorescence to pass but to block the excitation light. The animal body temperatures were maintained at 37 • C via a heating pad. Animals used in these studies were approved under the guidelines of the Institutional Animal Care and Use Committee (IACUC) of the National Health Research Institutes (Animal protocol number: NHRI-IACUC-104012A, 6 February 2015).

CT Imaging
For the in vitro studies, 10 mg mL −1 of the PEG-MSNs-Pt NPs were prepared. The prepared solutions were placed in 0.25 mL tubes to make concentrations of 1.25, 5, 10, and 20 mg mL −1 . Images were then obtained using an animal CT scanner (Trumph X, Gamma Medica Ideas, Northridge, CA, USA) with the following parameters: tube voltage, 70 kV; current intensity, 180 µA; and field of view, 512 × 512 mm 2 . For the animal studies, 1 × 10 6 MDA-MB-231 cells were subcutaneously injected into female nude mice. During the imaging, the mice were anesthetized with 2% isoflurane, and a pre-contrast CT scan was performed using the same setting as in the phantom study. Then, the MSNs-Pt NPs (10 mg mL −1 , 100 µL) were injected intratumorally, and images were sequentially obtained from 30 min to 24 h post-injection.