Characterization of Mn-N-CNSs
The Mn-N-CNSs was prepared via the high-temperature pyrolysis method using DTPA, MnCl2•4H2O and ethylenediamine as the C, Mn and N sources, respectively. Control experiments showed that the reaction time, temperature, and Mn2+ dosage have a great influence on the fluorescence and MRI properties of Mn-N-CNSs ( Additional file 1: Fig. S1). A optimal reaction condition via the carbonization DTPA and MnCl2•4H2O with the molar ratio of 1:1 in the presence of ethylenediamine reaction at 200oC for 90 min was established to prepare Mn-N-CNSs with high quality. The morphology of as-synthesized Mn-N-CNSs was imaged by HR-TEM and shown in Fig. 1A. The Mn-N-CNSs revealed a highly uniform and monodispese spherical morphology and an mean diameter in a narrow distribution was measured to be 5.2 nm (Fig. 1B). The high-resolved TEM image exhibited the good crystal structure of Mn-N-CNSs with apparent lattice fringes (Fig. 1A inset), which the lattice fringe of 0.22 nm in the HR-TEM graph was consistent with that of graphene nanosheets[38]. A typical energy dispersive X-ray (EDX) pattern of the nanoparticles revealed that the Mn-N-CNSs consist of C, N, O and Mn elements (Additional file 1: Fig.S2). The content of Mn in Mn-N-CNSs was about 10.2 % (wt %), which is higher than that of the previous reported Mn-doped carbon dots [20]. Furthermore, AFM was used to characterize the thickness and morphology of Mn-N-CNSs (Fig.1D and E). The thickness ranged from 0.5 to 1.4 nm, indicating the monolayer or bilayer structure of Mn-N-CNSs.
The composition and structure of Mn-N-CNSs was also investigated by the XRD, FTIR and XPS analysis. The XRD image shown in Fig. 1C indicated a broad diffraction peak at 28o representing the (002) plane, which was in agreement with that of the two-dimensional carbon nanomaterial, graphene[39]. The FTIR spectra shows the peak at 3420 cm-1 for O-H/N-H stretching vibrations, 2914 cm-1 and 1105 cm-1 for C-H stretching vibrations, 1595 cm-1 for C=O stretching vibrations, 1402 cm-1 for C=C stretching vibrations, and 1329 cm-1 for C-N stretching vibrations (Fig. 1F). Besides, the peak at 609 cm-1 corresponding to the Mn-O and Mn-N stretching vibrations represented the coordination absorption peak of Mn2+ ions with the carbon nanosheets.
XPS analysis was performed to analyze the surface elements of Mn-N-CNSs. As exhibited in the XPS spectrum graph (Fig. 2A), the Mn-N-CNSs nanoprobe indicated the presence of C, O, N and Mn on its surface. In the expanded XPS spectra, the C1s peak appeared at 284.7 eV, 285.2 eV and 287.3 eV, corresponded to sp2 C in graphene, sp3 C in C-O and C-N, and C=O from carbonyls and carboxylates, respectively (Fig. 2B). Two typical binding peaks in N1s spectrum were observed at 400.2 eV and 401.7 eV, indicating the existence of (C)3-N and N-H for the nitrogen element, respectively (Fig. 2C). More importantly, two peaks at 640.9 eV and 639.6 eV corresponding to the spin-orbit coupled Mn 2p3/2, suggest that the valence state of Mn in Mn-N-CNSs was Mn2+ and Mn3+, with the Mn2+ in a higher percentage (Fig. 2D). Therefore, these results confirmed the successful synthesis of Mn-N-CNSs from multiple perspectives.
Optical features of Mn-N-CNSs
The UV-vis spectrum of Mn-N-CNSs aqueous solution indicated a peak located at 350 nm (Fig. 3A). The Mn-N-CNSs aqueous solution exhibited pale yellow and transparent in daylight, but turned to strong blue color under UV light (365 nm). The fluorescence quantum yield of Mn-N-CNSs solution was investigated and measured to be 52.53% by using quinine sulfate as standard, which is higher than that of Mn-doped carbon quantum dots (13%)[40]. Besides, the Fig. 3 indicated the fluorescence of Mn-N-CNSs at different excitation wavelengths from 320 nm ~ 520 nm. The PL emission peaks red-shifted from 420 nm to 580 nm at the excitation wavelength from 320 nm to 520 nm with the highest fluorescence intensity (435 nm) at the the excitation wavelength of 360 nm, indicating an excitation-dependent emission and good multi-color emission.
T1 longitudinal relaxivity
Mn-related nanomaterials have emerged as a novel T1 weighted contrast agents for MR imaging owing to favorable electronic configuration and enriched biomedical features[41,42]. As MRI nanoprobe, the MRI behavior of Mn-N-CNSs was investigated compared with the commercial agent Gd-DTPA by the previous 3.0T MR system (Fig. 4). The relaxivity values of Mn-N-CNSs and Gd-DTPA were detected and compared by measuring the longitudinal relaxation time (T1) at various Mn/Gd concentrations. As indicated in Fig. 4B, Mn-N-CNSs clearly induced a concentration dependent brightening effect to the T1-weighted MR images, which the bright signal could be enhanced with the increasing concentration of the nanoparticles. The longitudinal relaxation rate (r1), obtained by measuring the relaxation time as a function of Mn concentration, was found to be 10.30 mM-1s-1, which was 2.3-fold that of the commercial contrast agent Gd-DTPA (4.45 mM-1s-1), and much higher than Mn-carbon dots hybrid (3.26 mM-1s-1) that we have reported [20]. Such high r1 relaxivity value may come from high loading concentration of Mn ions on the surface of carbon nanosheets to shorten the longitunidal relaxation of water protons, thus increasing the signal intensity of T1-weighted MR images.
Stability of Mn-N-CNSs
The stability of Mn-N-CNSs in different solutions including DI water, PBS, FBS and RPMI-1640 medium were investigated. The as-prepared Mn-N-CNSs showed excellent colloidal dispersity in the above mentioned media(Additional file 1: Fig. S3A). The results of dynamic light scattering (DLS) measurement illustrated that the particles had negligible aggregation in the biological fluids (Additional file 1: Fig. S3B). Moreover, the Mn-N-CNSs also displayed stable T1-weighted MR signal in water and different biological fluids, indicating their high stability as MR contrast agent in various media (Additional file 1: Fig. S3C).
The fluorescent stability of Mn-N-CNSs was also investigated under various conditions. The results showed that Mn-N-CNSs maintained high fluorescent stability for a long period (2 month) (Additional file 1: Fig. S4A). The fluorescence anti-photobleaching test also confirmed the excellent fluorescent stability of Mn-N-CNSs under long-time irradiation by UV light (365 nm) (Additional file 1: Fig. S4B). In addtion, the fluorescence intensity of Mn-N-CNSs almost did not change in NaCl solution with high concentration (1 M) (Additional file 1: Fig. S5A) and relatively neutral pH (pH=5~9) (Additional file 1: Fig. S5B). All these features make the as-prepared Mn-N-CNSs excellent candidates as a powerful molecular imaging probe for biological application.
In vitro cytotoxicity and cellular uptake assays
The cytotoxicity evaluation of the Mn-N-CNSs is critical to ensure the biocompatibility of the nanoprobe to be deployed in the tumor imaging. The in vitro cytotoxicity of Mn-N-CNSs at various concentrations was investigated by performing MTT assays for 24 h against HO-8910 ovarian carcinoma cells and NIH3T3 fibroblast cells (Fig. 5A). The cell viability remained more than 90% at all treated concentrations for two cell lines. Even at the highest concentration of 1 mg/mL (3.2 mM Mn, measured by ICP-MS), 90.1% and 91% of the cells survived after 24h incubation of the Mn-N-CNSs for the HO-8910 and NIH3T3 cells, respectively, demonstrating the good biocompatibility of Mn-N-CNSs as MRI nanoprobe.
To observe the cellular uptake of Mn-N-CNSs@Anti-HE4 by NIH3T3 and HO-8910 cells, a biological TEM (Bio-TEM) analysis was performed. As Fig. 5B shows, large amounts of Mn-N-CNSs@Anti-HE4 nanoparticles existed outside of the NIH3T3 cell membrane indicating the failed uptake by NIH3T3 cells. However, for the HO-8910 cells, the Mn-N-CNSs@Anti-HE4 nanoprobes were largely phagocytized inside the cells and existed in the intracellular endosomes and the cytoplasm. The result indicated that the Mn-N-CNSs@Anti-HE4 nanoprobes could be selectively taken up and internalized by the HO-8910 ovarian cancer cells.
In vitro specific cell fluorescence imaging
To investigate the feasibility of Mn-N-CNSs@Anti-HE4 nanoprobe for targeted cellular imaging, HE4 positive HO-8910 cells and negative NIH3T3 cells were co-cultured with Mn-N-CNSs@Anti-HE4 and Mn-N-CNSs nanoprobes. As illustrated in Fig. 6, the NIH3T3 cells exhibited relatively weak multi-color fluorescence (blue, green, red) after the co-incubation with Mn-N-CNSs@Anti-HE4 and Mn-N-CNSs and there was no obvious difference between the two groups, indicating that the Anti-HE4 mAb didnot result in any facilitation in the cellular uptake of nanoprobe by NIH3T3 cells owing to the low HE4 expression. On the other hand, the HO-8910 cells also revealed slightly weak fluorescence signal under the incubation of Mn-N-CNSs nanoparticles. However, in the fluorescence image, the fluorescence signals were more intense and significantly enhanced in the HO-8910 cells treated by Mn-N-CNSs@Anti-HE4 compared to the Mn-N-CNSs treated group. It confirmed the targeting of Anti-HE4 mAb on the surface of Mn-N-CNSs@Anti-HE4 to the HO-8910 cells, therefore, facilitating the cellular fluorescence imaging to the ovarian cancer cells. Particularly, the cells remained their living morphology during the test period, indicating the low toxicity of Mn-N-CNSs@Anti-HE4 to the cells. Therefore, the Mn-N-CNSs@Anti-HE4 nanoprobe could effectively and selectively label the ovarian cancer cells with multi-color fluorescence.
The quantitative cellular uptake of the Mn-N-CNSs@Anti-HE4 by NIH3T3 and HO-8910 cells upon measuring the fluorescent intensity was shown in Fig. 7. It is obvious that the fluorescent intensity of the HO-8910 cells was 10-fold higher than that of NIH3T3 cells in multi-colors. In addition, the Mn-N-CNSs@Anti-HE4 displayed considerably higher intensity than that of Mn-N-CNSs group. These findings confirmed the assistance of Anti-HE4 in the cellular uptake of Mn-N-CNSs by HO-8910 ovarian cancer cells, indicating the role of Anti-HE4 antibody in cellular uptake through active targeting via receptor-mediated endocytosis.
In vitro enhanced MR imaging
Due to the superior MR contrast of the nanoprobes, an in vitro targeted MRI was performed. The HO-8910 and NIH3T3 cells were co-cultured with Mn-N-CNSs@Anti-HE4, Mn-N-CNSs, Gd-DTPA and PBS. As shown in Fig. 4C (a), the nanoprobe imaged cells could be identified on the basis of the bright signal at the tube bottom. Owing to the nonspecific absorption, the nanoprobes induced slight MR signal enhancement to the NIH3T3 cells compared with the control group. For NIH3T3 cells, the Anti-HE4 mAb didnot result in any enhancement in MR signal compared with Mn-N-CNSs attributed to the low affinity of Anti-HE4 mAb to NIH3T3 normal cells. On the other hand, the HO-8910 cells also revealed slightly weak MR signal under the incubation of Mn-N-CNSs. However, although the nonspecific absorption existed in the NIH3T3 cells and Mn-N-CNSs nanoprobe incubated HO-8910 cells, a notable brightening effect in the T1 MRI signal could still be distinguished for the Mn-N-CNSs@Anti-HE4 treated HO-8910 cells. The MR intensity values of different groups were presented in Fig. 4C (b). The MR signal intensity value of Mn-N-CNSs@Anti-HE4 in HO-8910 cells was higher than that of the NIH3T3 cells and Mn-N-CNSs in HO-8910 cells. It suggested that the Anti-HE4 mAb may favor the Mn-N-CNSs@Anti-HE4 to be internalized by the HE4 high expressed HO-8910 ovarian cancer cells, thus inducing a higher T1 signal. In addition, as shown in Fig. 8, the Mn-N-CNSs@Anti-HE4 exhibited overwhelming high MR signal in the HO-8910 cells compared with the commercial Gd-DTPA in the same Mn/Gd concentration, indicating the superiority of Mn-N-CNSs@Anti-HE4 as a MRI nanoprobe. These results demonstrated that the Mn-N-CNSs@Anti-HE4 could specifically and efficiently label the HO-8910 ovarian cancer cells both by fluorescence imaging and MR imaging. Therefore, the proposed Mn-N-CNSs@Anti-HE4 nanoprobe would favor targeted FL/MR dual-modal imaging for specific and accurate ovarian carcinoma diagnosis.
In vivo targeted MR imaging
Encouraged by the excellent performance of in vitro MR imaging, we next investigated the ability of molecular MRI with the Mn-N-CNSs@Anti-HE4 nanoprobe in specifically imaging the ovarian carcinoma in HO-8910-tumor-bearing mice. The in vivo T1-weighted MR images of the mice were obtained pre- and post-injection of the Mn-N-CNSs@Anti-HE4 and Mn-N-CNSs at various time points. As shown in Fig. 9, it was clearly observed that the subcutaneous tumor area gradually brighter than the surrounding tissues after intravenous injection of Mn-N-CNSs@Anti-HE4 and Mn-N-CNSs nanoprobe. The T1 MR signal gradually increased within 2 hours, then turned weaker over time. In comparison, the T1 signal in tumor region was significantly enhanced at the same time points after administration of Mn-N-CNSs@Anti-HE4. It indicated that the Anti-HE4 mAb could improve the targeting ability of the nanoprobe to the HO-8910 tumor to enhance the MR contrast effect in vivo. Besides, as shown in Fig. 9D, bright T1 MR signal was found in the gall bladder within 1h administration of Mn-N-CNSs@Anti-HE4 nanoprobe and the signal then disappeared after 24h, indicating the prepared nanoprobe was mainly cleared from body by the urinary system. The urine samples from the mice after the injection were collected for further investigation. The nanoparticles were harvested by centrifugation, and analysis on the fragment found strong T1-weighted signals, along with strong FL that is characteristic of Mn-N-CNSs (Additional file 1: Fig. S6). These results suggest that the Mn was still well incorporated into nanoparticles during excretion, thus ensuring the stability of the MR signal and low toxicity.
In vivo biocompatibility analysis
In vivo biocompatibility assays of Mn-N-CNSs@Anti-HE4 were conducted on healthy Kunming mice model for 22 days. Mn-N-CNSs@Anti-HE4 were intravenously injected into the mice and the saline injected mice were refered as control group. At determined time points, blood was collected for the complete blood count and serum biochemistry tests. For the complete blood count assay, several important standard markers including red blood cells (RBC), white blood cells (WBC), hematocrit (HCT), hemoglobin (HGB) platelet (PTL) and mean corpuscular volume (MCV) were selected to investigate the effect to immune system. As indicated in Additional file 1: Fig. S7, slight variations could be observed in the values of various hematological markers after 1d administrations. However, the values returned back to the normal level of control animals after 8 days. Though the values of hematological markers fluctuated in a short period after administration of Mn-N-CNSs@Anti-HE4 nanoprobe, the values were still within the normal ranges for each markers. The serum biochemistry study was also carried out to monitor the potential toxic effect of Mn-N-CNSs@Anti-HE4 nanoprobe. Indicators of heart, kidney and liver functions were evaluated including aspartate aminotransferase (AST), alanine aminotransferase (ALT), total protein (TP), indicators-albumin (ALB), total protein (TP) and creatinine. As shown in Fig. 10A, the values of various biochemical markers revealed a higher level for the animals treated by Mn-N-CNSs@Anti-HE4 nanoprobe compared with the saline treated group at 1 day post-injection. Then, the values lowered back to the normal level after 8 day post-injection, which was consistent with the fluctuation behavior of hematological markers. These results demonstrated that the Mn-N-CNSs@Anti-HE4 nanoprobe injection may slightly affect the biological conditions of the mice in a short period, but without acute damage in a long term.
Besides, the biosafety of the nanoprobe was further investigated by the histological analysis. Major organs (heart, kidney, liver, spleen and lung) were sliced and stained for the detailed microscopic evaluation of the interaction of tissues and Mn-N-CNSs@Anti-HE4 nanoprobe. H&E-stained images of organ sections displayed no apparent inflammation, histopathological abnormalities or lesions after treatment with the probe. The body weights of the mice were also monitored during the whole test period. The mice upon Mn-N-CNSs@Anti-HE4 administration exhibited no notable variation compared to those of the control group in a long term (Additional file 1: Fig. S8). These results demonstrated the low toxicity and good biocompatability of the Mn-N-CNSs@Anti-HE4 nanoprobe, unveiling its further applications in biomedical fields.