2.1 Synthesis and Characterization of GdIO
The GdIO was synthesized through high temperature of solvent thermal decomposition with Fe(acac)3 and Gd(acac)3 as a precursor. As shown in Fig. 1a-d, the original MIO consisted of abundance of tiny crystals and presented typical spherical shape with a size of about 150 nm. The high-resolution transmission electron microscope (HRTEM) image of MIO showed significant lattice fringe, and the interplanar spacing was 0.294 nm, which was assigned to (220) plane of spinel structure of magnetite. Nevertheless, after engineered by gadolinium species, the magnetic particles were etched, and showed the hollow structure, which provided enough capacity to load anticancer drug. In addition, it could be seen that the edge of GdIO transformed from spherical shape to cube, and its HRTEM images showed that there were two types of lattice fringes. The interplanar spacing in cube edge had no significant variation and was also 0.294 nm compared to MIO, indicating that the GdIO still contained the spinel structure of magnetite. It was surprising that the lattice fringe in the inner of GdIO was different with the cube edge, and its lattice fringe was 0.281 nm, which were assigned to (302) plane of Gd(OH)CO3 (PDF:24–0421), respectively. These results indicated that the GdIO was consisted of MIO and Gd species. Subsequently, the energy-dispersive X-ray (EDX) element mapping and line scanning analysis of GdIO confirmed the presence of Gd, Fe, O, N, and C elements, and showed that Gd and Fe species uniformly distributed into MIO. Besides, the signal of all elements in lumen region was weaker than that of edges.
DLS results indicated that the hydrodynamic size of GdIO was approximate 357 nm, which was far larger than TEM observation. This result demonstrated that the pure GdIO might be aggregation in the distilled water. Remarkably, after iRGD-PEG-ss-PEG modification, the hydrodynamic size of ipGdIO significantly decreased and had only about 160 nm, indicated that the ipGdIO had excellent colloidal stability. The X-ray diffraction (XRD) pattern was used to analyze the structure and composition of the GdIO, and the result showed that the GdIO particles mainly consisted of Fe3O4 and Gd(OH)CO3, which was consistent with HRTEM observation. In addition, the full spectra of XPS further confirmed the presence of Gd, Fe, C, N, and O elements. Fe2p spectra indicated that Fe1/2p and Fe3/2p peaks appeared at 710.5eV and 723.3eV, respectively, which were assigned to Fe(III) in the GdIO.25 In addition, Gd4d peaks appeared at 142.6 and 157.9 eV, further confirming the presence of Gd(III) in GdIO.26
2.2 Magnetic Property and Catalytic Activity of ipGdIO
As we known, the magnetism of particle played a key role for achieving excellent MRI performance. Based on this, the magnetic property of the GdIO was investigated through superconducting quantum interference device (SQUID). The field-dependent magnetization curves indicated that GdIO showed high saturation magnetization with about 42.5 emu/g at 300K, implying that the GdIO had strong ability to shorten T2 relaxation time of water molecules. Notably, when at 3K, the magnetization of GdIO gradually increased with magnetic field increase, and cannot reach to saturation, this might be closely associated with the presence of Gd species. In addition, there had no significant coercive force and remanence when at 300 K, indicating that the GdIO showed excellent superparamagnetism at room temperature. These results demonstrated that the GdIO had excellent magnetic property and could be a potential candidate as MRI contrast agent.
As previously described,27–29 Fenton-catalytic efficacy mainly depended on the concentration of iron ions. Therefore, the Fe release behaviors of the GdIO were investigated under different pH conditions. As shown in Fig. 2f, the release amount of iron ions from GdIO gradually increased with pH decrease, and the maximal release amount reached to approximate 15.2%. It was noted that the Fe release rate of GdIO was significantly quicker than that of MIO, suggesting that the GdIO had stronger response ability to ATME. Based on this, it was worth expecting that the GdIO showed high performance of Fenton-catalytic activity. Next, the Fenton-catalytic activity of the GdIO was analyzed through electron spin resonance (ESR) spectra (Fig. 2h and 2i). After the GdIO treatment, there were significant peaks (1:2:2:1) assigned to hydroxyl radicals (·OH). Moreover, the amount of ·OH gradually increased with the GdIO concentration increase, implying that the Fenton-catalytic activity of the GdIO was dosage-dependent. In addition, the GdIO showed stronger ability to catalyze H2O2 into ·OH when compared to traditional MIO, indicating that the GdIO might be a better ferroptosis agent than traditional MIO. Because of abundance of pore structures, the Dox could be effectively loaded into GdIO (GdIO-Dox). Subsequently, the GSH-responsive polymer (iRGD-PEG-ss-PEG) was coated onto the surface of the GdIO-Dox, and then fabricated ipGdIO-Dox. In Fig. 2e, the release amount of Dox in ipGdIO-Dox gradually increased with GSH concentration increase, indicating that the Dox release of ipGdIO-Dox was GSH dependent. This result demonstrated that the iRGD-PEG-ss-PEG chain on the surface of ipGdIO-Dox not only targeted tumor cells, but effectively depleted GSH, which would be beneficial to promote Dox release and up-regulate cellular ROS.
2.3 The internalization, ROS Generation of ipGdIO in vitro
The internalization process of ipGdIO was investigated through CLSM observation. First, FITC molecules were rationally labeled into ipGdIO, and then was used to incubate cancer cells. As shown in Fig. 3a and 3c, MDA-MB-231 cells treated with pGdIO showed strong green fluorescence, indicating that pGdIO could be effectively internalized by cancer cells. Moreover, the ipGdIO-treated cells showed stronger green fluorescence than that of ipGdIO, demonstrating that the targeting ligand promoted the uptake of ipGdIO. In addition, after the co-incubation with ipGdIO and amiloride, the green fluorescence of cell showed a slight decrease, indicating that the internalization of ipGdIO were partly inhibited. Remarkably, when the incubation with ipGdIO at 4 oC, the green fluorescence of cells significantly decreased. These results demonstrated that the internalization pathway of ipGdIO mainly energy- and clathrin-mediated endocytosis. Subsequently, the internalization process of ipGdIO were quantitatively analyzed through assessing Fe content in cells using ICP-MS. In Fig. 3b, it could be seen that the cell uptake of particle was dosage- and time-dependent. Notably, when treatment for 3h, the Fe content of MDA-MB-231 cells treated with ipGdIO (93.8 ng/10000cells) was higher 3-fold than that with pGdIO (23.2ng/10000cells), further confirming that ipGdIO had excellent targeting ability to MDA-MB-231 cells. In addition, after co-incubation with ipGdIO and amiloride, cellular Fe content decreased from 93.8 ng/10000cells to 85.6 ng/10000cells. Moreover, after incubation at 4 oC, the cellular Fe content further decreased to be only 60.3ng/10000cells. These results further demonstrated that the internalization of ipGdIO was realized through energy- and clathrin-mediated endocytosis, which was consistent with CLSM and flow cytometry result.
After entered cancer cells, it was very critical that whether ipGdIO activated cellular ROS. Based on this, we investigated cellular ROS level before and after ipGdIO treatment. As shown in Fig. 3d and 3e, the control group had no green fluorescence. The ultra-small iron oxide (USIO) was a classical ferroptotic agent, and was chosen as positive control. MDA-MB-231 cells treated with USIO only showed weak green fluorescence, but pGdIO had stronger green fluorescence, indicating that pGdIO had stronger ability to activate cellular ROS. In addition, when compared to pGdIO, the green fluorescence of MDA-MB-231 cells treated with ipGdIO was further strengthened, this was because the targeted ipGdIO could be more effective internalized by cancer cells, then result in more ROS generation. Remarkably, the use of either NAC (N-acetyl-L-cysteine, ROS scavenger) or DFO (Deferoxamine, ferroptosis inhibitor) alleviated cellular ROS production at different levels, suggesting that the Fe-activated ROS was closely associated with ferroptosis process.
2.4 Cytotoxicity, and Anticancer Mechanism of ipGdIO-Dox
The cellular ipGdIO-Dox not only activated excessive ROS, but released abundance of Dox. Then, they damaged mitochondria and inhibited DNA replication, resulting in cancer cell death. The cytotoxicity of ipGdIO-Dox was investigated through CCK-8 assay. As shown in Fig. 4a, all groups showed concentration-dependent effect on cell viability. In addition, the Dox seemed to have stronger inhibition ability on cell viability than pGdIO-Dox when concentration was less than 1 µg/mL. However, with concentration increase, the cytotoxicity of pGdIO-Dox was significantly stronger than Dox, this might be associated with the sustainable release of Dox from pGdIO-Dox. The ipGdIO-Dox showed the strongest cytotoxicity when compared to pGdIO-Dox and Dox, indicating that ipGdIO-Dox had the strongest anticancer ability. Notably, the use of NAC significantly decreased the cytotoxicity of ipGdIO-Dox, further suggesting that ROS-induced cancer cell death was an important pathway. In order to verify mechanism of cancer cell death, various inhibitors were used to treat cancer cells. As shown in Fig. 4b, both DFO and NAC partly recovered cell viability, confirming the presence of ROS-induced ferroptosis. In addition, the use of z-VAD also increased on cell viability, indicating that the ipGdIO-Dox induced apoptosis was also a vital approach for cancer cell death. In addition, the cycles of MDA-MB-231 cells treated with different samples were analyzed through flow cytometry. Compared to saline, all groups significantly inhibited S phase of MDA-MB-231, this was main because Dox effectively inhibited DNA replication in MDA-MB-231 cells.30 Interestingly, the ipGdIO-Dox showed strongest inhibition ability on G2-M phase of MDA-MB-231, and the use of NAC could dramatically recover G2-M phase. This result indicated that high level of ROS might block G2-M phase,31 and then induce cancer cell death.
Considering the synergistic effect of chemo- and ferroptosis therapy, some key proteins including Bcl-2, Bax, GPX4, and cleaved Caspase-3 were investigated through western blot analysis. The expression of Bcl-2 and Bax families were closely associated with the mitochondria-mediated apoptosis.32,33 As shown in Fig. 4c, the expression of Bcl-2 in cells treated with pGdIO-Dox significantly reduced compared to saline group, and corresponding Bax expression showed opposite result, demonstrating the presence of mitochondrial damage induced by ROS generation. After ipGdIO-Dox treatment, the expression of Bax was further up-regulated, and Bcl-2 was down-regulated, indicating that ipGdIO-Dox had the stronger ability to damage mitochondria. In addition, the expression of cleaved caspase-3 in cells dramatically increased after ipGdIO-Dox treatment, further confirming that ipGdIO-Dox had the strongest ability to kill cancer cells. In addition, GPX4 protein could effectively eliminate cellular ROS through converting GSH into GSSG, and then avoid ferroptosis.34,35 Therefore, the expression of GPX4 was considered to be an important ferroptotic marker. In Fig. 4c, the expression of GPX4 significantly decreased after pGdIO-Dox treatment, confirming the presence of ferroptosis. Moreover, ipGdIO-Dox-treated cells showed the lowest GPX4 expression, indicating that the targeting ligand could further accelerate ferroptosis therapy. On the basis of above analysis, it could be concluded that ipGdIO-Dox effectively induced cancer cell death through the synergistic effect of chemo- and ferroptosis therapy.
2.5 Antitumor Activity, Biodistribution, and Pharmacokinetics of ipGdIO-Dox
The excellent performance of ipGdIO-Dox in vitro encouraged us to further explore its anticancer ability in vivo. Cancer-bearing mice were intravenously injected with ipGdIO-Dox at a dosage of 2 mg/kg, and then tumor volumes of mice were recorded. As shown in Fig. 5a and 5b, Dox showed significant inhibition on tumor volume when compared to saline group. In addition, the anticancer ability of pGdIO-Dox had a slight enhancement when compared to free Dox, which could be attributed to the synergistic effect of chemo- and ferroptosis therapy. It was noted that the ipGdIO-Dox showed the strongest tumor inhibition effect, this was because targeting ligand accelerated the ipGdIO-Dox accumulation in tumor site and then strengthened its anticancer activity. Remarkably, the use of NAC significantly weakened the anticancer effect of ipGdIO-Dox, confirming the involvement of ROS-mediated ferroptosis therapy. After treatment, the tumors of mice were excised, weighed, and then photographed. As shown in Fig. 5c, the tumor weight of mice injected with ipGdIO-Dox was the smallest, and had only 0.23 g when compared to other groups. In addition, the tumor weight of mice significant increased after NAC and ipGdIO-Dox co-injection. Meanwhile, the photos of tumor also showed the consistent result with tumor volume and tumor weight (Fig. 5e). These results demonstrated that ipGdIO-Dox could effectively inhibit tumor growth through synergistic action of chemo- and ferroptosis therapy, and be a promising candidate in clinical use.
At the end of treatment, the part of tumor tissues were cut into slices and then further analyzed through H&E staining and IHC staining. As shown in Fig. 5h, the necrotic areas of tumor tissues were as follows: Saline < ipGdIO-Dox + NAC < Dox < pGdIO-Dox < ipGdIO-Dox. In addition, IHC staining analysis showed the expression of apoptotic protein, and the order was as follows: Saline < ipGdIO-Dox + NAC < Dox < pGdIO-Dox < ipGdIO-Dox. These result further demonstrated that ipGdIO-Dox had the strongest anticancer ability. The body weight of mice were recorded, and the results indicated that all groups maintained a normal increase on body weight, and no significant difference was observed (Fig. 5d). The vital organs of mice in all groups including the heart, liver, spleen, lung, and kidney were also cut, and then analyzed through H&E staining. All groups had no significant pathological change (Figure S2). These results demonstrated that the ipGdIO-Dox had low side effects on the body. Besides, the biosafety of ipGdIO-Dox was also an important consideration for further clinical use. Therefore, the biodistribution and pharmacokinetics of ipGdIO were further investigated as shown in Fig. 5f and 5g. It could be seen that pGdIO-Dox and ipGdIO-Dox mainly accumulated in liver and kidney tissue, this might be because the liver and kidney had abundance of reticuloendothelial system. Notably, the tumor accumulation of ipGdIO-Dox was significantly higher than that of pGdIO-Dox, confirming that ipGdIO-Dox had effective targeting ability to tumor. The pharmacokinetics curves indicated that the blood circulation time of Dox was very short and the half-time had only approximate 0.82 h. Nevertheless, the blood circulation time of pGdIO-Dox significantly increased when compared to free Dox. Notably, the ipGdIO-Dox showed the longest blood circulation time, which would be beneficial to promote the tumor accumulation of particles. Meanwhile, at post-injection 48 h, the ipGdIO-Dox could be completely excreted out from body, which decreased the potential risk for body because of long-term of accumulation. Based on above analysis, it was concluded that the ipGdIO-Dox had excellent biosafety and might be a promising candidate for further clinical use.
2.6 In Vivo Tumor MRI Diagnosis
As we mentioned earlier, the GdIO showed excellent magnetic property, which encouraged us to explore its MRI contrast performance. The T1 and T2 relaxation rate were very key indicator to assess the MRI contrast performance of particles, and it could be calculated via the slope of linear 1/Ti (i = 1 or 2) to metal ion concentration. As shown in Figure a, the T1 relaxation rate of ipGdIO was 4.87 mM− 1s− 1 and 2.44 mM− 1s− 1 at 3.0 T and 7.0 T, respectively. Subsequently, the T2 relaxation rate of ipGdIO was also measured and showed 279.2 mM− 1s− 1 and 384.5 mM− 1s− 1 at 3.0 T and 7.0 T, respectively. In addition, the T1 images of ipGdIO significantly brightened with concentration increasing, and corresponding T2 images also gradually darkened. These results indicated that ipGdIO had strong T1 and T2 contrast performance and had great potential as T1-T2 dual-modal MRI contrast agent. Next, we further assessed tumor diagnostic ability of ipGdIO-Dox through 7.0 T MRI scanner. The T1- and T2-weighted images of mice at axial plane were acquired at different time after injection of pGdIO-Dox and ipGdIO-Dox with a dose of 10 mg/kg. The T1-weighted images (T1WI) of tumor tissue gradually brightened at post-injection (p.i.) of pGdIO-Dox and the brightest image appeared at p.i. 2 h. After that, the T1WI of tumor began to weaken, this might be because pGdIO-Dox began to be metabolized from tumor. Compared to pGdIO-Dox, the T1WI of tumor in mice injected with ipGdIO-Dox was significantly brighter, which could be attributed to excellent targeting ability. In addition, the MRI signal to-noise ratio change (ΔSNR) of tumor region was analyzed through MRIcro software.The greatest ΔSNR of ipGdIO-Dox appeared at p.i. 3h and was up to more than 65%, which was more 2-folds than that of pGdIO-Dox. In addition, the T2WI of tumor treated with pGdIO-Dox significantly darkened with extend time, and the darkest images also appeared at p.i. 2 h, which was consistent with its T1WI result. Remarkably, the tumor region seemed to be very bright at pre-injection of ipGdIO-Dox, this might be because the tumor of mice contained abundance of adipose tissues. However, after injection of ipGdIO-Dox, the T2WI of tumor gradually darkened, and the darkest image appeared at p.i. 3 h. In addition, the analysis of ΔSNR indicated that the maximal ΔSNR of ipGdIO-Dox was significantly higher than that of pGdIO-Dox, and it was about 1.5-folds of pGdIO-Dox. These results demonstrated that the systemic delivery of ipGdIO-Dox effectively enhanced the ΔSNR of T1WI and T2WI in tumor, which can accelerate the early diagnosis and accurate therapy of tumors. Therefore, it could be concluded that ipGdIO-Dox might provide a promising candidate to achieve a low toxicity, effective, and accurate cancer theranostics.