Preparation, characterization and structural control of MSbNSs.
The schematic illustration of the formation process of MSbNSs is shown in Fig. 1a. Dodecylthiol (DDT) and oleylamine (OLA) were used as capping ligands to control the shape of Sb nanospheres, while octadecene (ODE) was used as solvent, and SbCl3 and tert-butylamine borane were selected as semimetal precursor and reducing agent, respectively. After chemical reduction, numerous small Sb nuclei were formed, which were sensitive to oxygen and could be easily oxidized due to their high surface energy. As the Sb nanocrystals grew, the oxidation product antimony oxide (Sb2O3) was dispersed in Sb nanocrystals, which could further react with DDT in the solution as the selective etching process of Sb2O3 to form mesopores. The pore sizes of MSbNSs are tunable by carefully controlling the oxidative degree of Sb nuclei and the selective etching of the as-formed Sb2O3. Detailed experimental procedures were provided in experimental section. The resultant MSbNSs have shown uniform spherical morphologies in the transmission electron microscopy (TEM, Supplementary Fig. 1), and their high-angle annular dark-field scanning TEM (HADDF STEM) images (denoted as MSbNSs-1, MSbNSs-2, MSbNSs-3 and MSbNSs-4) were shown in Fig. 1b-e. The size of MSbNSs were gradually increased from 45nm, to 48nm, 51nm, and to final 58 nm (Supplementary Fig. 2), respectively. N2 absorption and pore size analysis was performed to demonstrate the mesoporous structures of MSbNSs (Fig. 1f). The obvious hysteresis loop of the absorption/desorption isotherms in all of the cases implied the typical mesoporous structures of the resultant MSbNSs. The Brunauer-Emmett-Teller (BET) surface area of MSbNSs increased from 78 m2g-1 to 231 m2g-1, 235 m2g-1 and to final 234 m2g-1. MSbNSs with different pore sizes were obtained via selective etching of SbNSs by controlling the amount of oxygen (O2) in the reaction precursor. By controlling the pumping time of O2 from 10s to 30s, the mesopore size of MSbNSs has been increased from 2.6 to 3.7 nm. Increase of O2 amount has generated larger pores (~11.5 nm) in MSbNSs-3 besides small mesopores with average sizes of 3.9 nm and 5.7 nm. The nitrogen sorption isotherms of the MSbNSs-3 showed two major capillary condensation steps in the relative pressure ranges 0.1-0.3 and 0.75-0.98, respectively, indicating that at least two sets of pores coexisted in the nanospheres. MSbNSs-1, MSbNSs-2 and MSbNSs-3 can still remain integrated spherical shape. Nevertheless, further increased O2 amount would destroy the nanostructure of MSbNSs and lead to collapsed MSbNSs-4 with larger pores (~18 nm). To further confirm the important role of O2 in the formation of mesopores, a control experiment without the pumping of O2 was performed under the situation that other experimental parameters were kept constant. Mesoporous structures were not observed as shown in Supplementary Fig. 3. MSbNSs with tunable mesopore sizes were successfully fabricated by our selective etching method for the first time to the best of our knowledge. The shape of MSbNSs can also be facilely tuned by regulating the reaction temperature (Supplementary Fig. 4).
Moreover, MSbNSs-1, MSbNSs-2 and MSbNSs-3 showed strong and broad absorption, especially in the NIR-II range (Fig. 1g). From MSbNSs-1 to MSbNSs-3, the absorption peaks were further red-shifted towards NIR-II range, providing the potential for excellent NIR-II photothermal performance. The collapsed MSbNSs-4 lost the strong NIR-II absorption, which is not applicable for the multimodal theranostics.
Formation mechanism of MSbNSs.
To shed light on the novel selective etching mechanism, X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) were carried out to analyze the reaction process. XRD patterns revealed that MSbNSs-1 and MSbNSs-2 remained pure rhombohedral phase of Sb corresponding with the standard card JCPDS 085-1324, while a mixed phase of rhombohedral phase of Sb and cubic phase of Sb2O3 corresponding with JCPDS 05-0534 were detected in MSbNSs-3 and MSbNSs-4 (Fig. 2a). XPS results further illustrated that lower amount of O2 first oxidized Sb to form Sb2O3 as measured in MSbNSs-1, while higher amount of O2 would further generate Sb2S3 that derived from the reacting Sb2O3 with DDT, since Sb2S3 first appeared in MSbNSs-2 and gradually increased in MSbNSs-3 and MSbNSs-4 (Fig. 2b). Sb2O3 and Sb2S3 generated in the early oxidation process were not crystalized and Sb2S3 can be dissolved in reaction solvent to form the mesoporous structures, thus not detectable in XRD patterns. Moreover, Sb2S3 could react with O2 to form Sb2O3 and SO2, leading to a dynamic reaction cycle to accelerate the etching process. A higher amount of O2 could even further oxidize Sb2O3 to Sb2O5 as Sb3d5/2 only appeared in MSbNSs-4. Therefore, we proposed a step-oxidization reaction mechanism to form the mesoporous structure in the following three steps as shown in Fig. 2c.
Sb+O2→Sb2O3 (1)
Sb2O3+C12H25SH→Sb2S3+C12H25OH (2)
Sb2S3+O2→Sb2O3+SO2 (3)
Sb2O3+O2→Sb2O5 (4)
Step (2) and Step (3) were further confirmed by measuring the Fourier-transform infrared spectroscopy (FTIR) spectra of the reaction solutions (Fig. 2d). O-H stretching of alcohol (3550-3200 cm-1) and S=O stretching of SO2 (1350-1300 cm-1) were tested to be stronger as the increase of O2 amount, especially in the solvent of MSbNSs-4. Elemental mapping using energy-dispersive X-ray spectroscopy (EDS) under STEM mode further confirmed the existence of O and S in MSbNSs (Fig. 2e, Supplementary Fig. 5). The corresponding mass percentages of Sb, O and S elements in MSbNSs-3 are 94.5%, 3.3% and 2.2%, respectively (Supplementary Fig. 6). From MSbNSs-2 to MSbNSs-4, although the amount of O2 in the reaction precursor were increasing, the residual ratios of O elements in the final products were decreasing, while the ratios of S element were increasing with a significant increase in MSbNSs-4, which further proved our proposed reaction mechanism (Fig. 2f).
NIR-II photothermal, degradation and drug loading/release performance of MSbNSs.
Different from the good photothermal stability of Sb nanopolyhedrons reported in our previous work,37 the as-synthesized MSbNSs showed varying degrees of photothermal degradation under 1210 nm laser irradiation (Fig. 3a). In particular, MSbNSs-3 showed the most obvious photothermal degradation among MSbNSs-1/2/3. The detailed degradation behavior of MSbNSs-3 was first studied by TEM (Fig. 3b). Obvious degradation of the mesopores in MSbNSs-3 and numerous soluble Sb species could be observed after two cycles of irradiation. Almost all the spherical structures were destroyed after three successive cycles of irradiation, indicating that MSbNSs-3 gradually collapsed after laser irradiation. The color of the black MSbNSs-3 solution also gradually faded to transparent after three cycles of irradiation. The photothermal stability of MSbNSs-2 decreased to some extent, but the degradation of MSbNSs-2 proceeded slowly (Supplementary Fig. 7). The degradation of the mesopores in MSbNSs-2 can also be observed after three successive cycles of irradiation, but the collapse of mesoporous nanostructures would not appear. By comparison, MSbNSs-1 showed the best photothermal stability (Supplementary Fig. 8). The photothermal degradation mechanism of MSbNSs was also analyzed based on the FDTD simulations as Sb-based nanostructures possess localized surface plasmon resonance (LSPR) effect.41 The detailed fitting process is described as shown in supporting information. The results showed that electric fields at the joints and gaps inside the mesoporous structure were amplified by more than several orders of magnitude, resulting in stronger localization of heat generation in mesopores. The simulation results indicated that the mesopores of both MSbNSs-2 (Fig. 3c) and MSbNSs-3 (Fig. 3d) could generate localized “hot channels” and similar heat power densities were generated in MSbNSs-2 and MSbNSs-3, while “hot channels” would gradually disappear as mesopores grew larger (~11.5 nm), which is consistent with the change of the photothermal performance of MSbNSs-2 and MSbNSs-3. During the photothermal process, the increased temperature would promote the following reaction (Step (5)) of surface Sb atoms with water molecules to form Sb-H and Sb-OH, leading to the dissolution of Sb nanocrystals in aqueous solution.42
Sb+2H2O→Sb-OH2+Sb-HOH→Sb-H+Sb-OH (5)
Since MSbNSs-3 have larger pore sizes, the amount of accessible water within the NSs would be higher. The photothermal degradation process was further promoted in MSbNSs-3. The looser mesoporous structure and lower crystallinity of MSbNSs-3 could also accelerate the dissolution and result in the enhanced photothermal degradation property in MSbNSs-3.
The photothermal degradation property of MSbNSs can further boost the release efficiency as nanocarriers for different types of drugs. We evaluated their capabilities of drug delivery and NIR-II triggered on-demand release of drugs using DOX as a typical anticancer model drug (Fig. 3e). The drug loading capacity and efficiency and of MSbNSs-2 and MSbNs-3 were studied as shown in Fig. 3f. A burst in the release of DOX (~53%) can be observed with 1210 nm irradiation, which is about 6 times higher than that of the group without NIR-II light irradiation (Fig. 3g). The on-demand release properties under NIR-II laser irradiation could be ascribed to the collapse of the MSbNSs and local photothermal effect to accelerate the diffusion of drugs. All of the above results reveal that the MSbNSs can be used simultaneously as agents for PTT, PAI and novel drug carriers for NIR-II controlled drug release.
Considering the appropriate pore size and NIR-II absorption wavelength as well as photothermal degradation, we have chosen MSbNSs-3 to perform multimodal theranostics in the following work. MSbNSs-3 were PEGylated and the absorption properties of PEGylated MSbNSs-3 in aqueous solution still exhibited broad absorption from visible to NIR wavelengths and the strongest absorption peak was 1480 nm, located within the NIR-II window (Fig. 4a). The strong absorption of PEGylated MSbNSs-3 in the NIR-II region suggest their potential in NIR-II PAI and PTT. The photothermal effect of PEGylated MSbNSs-3 was studied by the irradiation of 1210 nm laser. This provides another efficient NIR-II excitation wavelength besides the typical 1064 nm. As shown in Fig. 4b, 4c, the temperature variation of PEGylated MSbNSs-3 displayed a concentration-dependent behavior. We also investigated the photothermal performances of PEGylated MSbNSs-3 irradiated under different power densities (Fig. 4d). The PTCE of PEGylated MSbNSs-3 was calculated to be ~39% (Fig. 4e), which is superior to the majority of current NIR-II PTAs. By comparison, PEGylated MSbNsS-1 and MSbNSs-2 was calculated to be ~44% and ~41% (Supplementary Fig. 9). Moreover, a strong concentration-related photoacoustic signal of PEGylated MSbNSs-3 was observed (Fig. 4f), implying that PEGylated MSbNSs-3 could be used as bright NIR-II photoacoustic contrast agents.
In vitro synergistic PTT and chemotherapy.
To study the synergistic therapeutic outcomes of DOX-loaded PEGylated MSbNSs-3 (PEGylated MSbNSs-3/DOX), the cellular internalization process was first checked in panc02 cancer cells using confocal microscopy. The fluorescence of DOX overlapped with the fluorescence of Lyso-tracker (Fig. 5a), indicating that the uptake process was mainly achieved by endocytosis. The cytotoxicity and treatment efficacy of PEGylated MSbNSs-3/DOX were further tested by Live/dead assays. PEGylated MSbNSs-3 with/without DOX were incubated with panc02 cancer cells under different concentrations from 0 to 100 μg mL-1 (Fig. 5b). Without 1210 nm irradiation, PEGylated MSbNSs-3 showed a high cell viability indicating a negligible cytotoxicity, meanwhile PEGylated MSbNSs-3/DOX would kill cancer cells by releasing DOX. The in vitro chemotherapy efficacy is limited compared with NIR-II PTT. The 1210 nm irradiation on PEGylated MSbNSs-3 could generate heat to kill cancer cells, leading to a cell viability reduced to 20% at 100 μg mL-1. The synergistic PTT and chemotherapy showed an improved in vitro therapeutic capability with a minimal cell viability of 8% (Supplementary Fig 10).
In vivo NIR-II PAI-guided PTT and chemotherapy.
In vivo 1210 nm PAI was conducted to guide the optimal systemic administration window of PEGylated MSbNSs-3/DOX. The PA signal in the tumor increased gradually after the intravenous injection of PEGylated MSbNSs-3/DOX, indicating the gradual accumulation of these nanomedicines in the tumor site (Fig. 5c). The PA signal achieved the highest at 8 h post-injection, with a 9-fold higher PA amplitude than the pre-injection tumor (Supplementary Fig.11). With the guidance of PAI, the photoirradiation of 1210 nm was conducted at 8 h post-injection to prove the therapeutic efficacy of the synergistic NIR-II PTT and chemotherapy on 4T1 tumor-bearing mice. The tumor temperature of PEGylated MSbNSs-3/DOX treated mice increased rapidly to ~46 ℃ within 3 min and finally arrived to ~50 ℃ at 7 min, which is significantly higher than that of PBS-treated tumors (Fig. 5d). To further investigate the in vivo photothermal triggered degradation of PEGylated MSbNSs-3, PA images before and after laser irradiation were studied as shown in Fig.5e. An obvious photothermal degradation of PEGylated MSbNSs-3 was observed after the irradiation of 1210 nm laser for 10 min, which was indicated by the dramatic drop of PA intensity of tumor by 70% (Supplementary Fig. 12). To assess the clearance of PEGylated MSbNSs-3, the biodistribution of MSbNSs-3 in mice were studied. After the PEGylated MSbNSs-3 were injected into mice and without irradiation of 1210 nm laser, the distribution of PEGylated MSbNSs-3 in major organs were measured at 1, 7, 14 days. PEGylated MSbNSs-3 were mainly cumulated in the liver, spleen and kidney (Fig. 5f).
The tumor sizes in different groups were measured and compared (Fig. 5g). Mono chemotherapy by PEGylated MSbNSs-3/DOX showed a mild tumor inhibition rate. Meanwhile, the NIR-II PTT by PEGylated MSbNSs-3 showed excellent therapeutic outcomes with eradication of tumors owing to the excellent photothermal performance of MSbNSs-3. Tumors treated by PEGylated MSbNSs-3/DOX for synergistic NIR-II PTT and chemotherapy were also eradicated, with an even faster speed. The body weights of mice in all the groups were not influenced by different treatments in all the groups (Fig. 5h). The survival rates of mice were monitored to further prove the treatment efficacy. The treatment of NIR-II PTT and chemotherapy resulted in a highest survival rate (Fig. 5i). Moreover, there were also no obvious histopathological abnormalities found in major organs of the mice (Supplementary Fig. 13), indicating the good biocompatibility of the MSbNSs nanoplatforms. Hematoxylin and eosin (H&E) staining and TdT-mediated Dutp-biotin nick end-labeling (TUNEL) staining of tumor slices evaluated the necrosis and apoptosis of tumor cells in all the groups (Fig. 5j). No apparent necrosis or apoptosis was found in control groups, PEGylated MSbNSs-3/DOX caused moderate necrosis and apoptosis by releasing DOX. Large areas of necrosis and apoptosis were found in 1210 nm irradiated PEGylated MSbNSs-3 and PEGylated MSbNSs-3/DOX groups. The synergistic NIR-II PTT and chemotherapy caused the most necrosis and especially apoptosis, which is also consistent with the tumor growth curves. Altogether, these data demonstrated that our novel PEGylated MSbNSs-3/DOX nanoplatform showed high potential for multimodal NIR-II theranostics with excellent therapeutic efficacy, good biocompatibility and photothermal degradability.