Ultra-sensitive diagnosis of orthotopic patient derived hepatocellular carcinoma by Fe@graphene nanoparticles in MRI

Abstract

We synthesized core–shell structured Fe/G-PEG NPs composed of an Fe core and a PEGylated graphene, which exhibited superior stability against oxidation, good water dispersity, low cytotoxicity and high magnetic performance. These advantages ensure them as candidates for ultra-sensitive contrast enhanced magnetic resonance imaging agents in vivo.

---

MRI has been thought of as a vital imaging tool in diagnosis owing to its non-invasive nature,¹ excellent resolution of soft tissue,² multi-parameter imaging,³ tomographic capabilities⁴ and no ionizing radiation damage.⁵ However, most magnetic resonance imaging results are too vague and too insensitive to provide a detailed diagnostic message.^6,7 So far, MRI contrast agents are commonly used to enhance MRI sensitivity and accuracy.^8–10 Fe₃O₄ NPs are widely used MRI contrast agents in clinics because of their low cost, easy fabrication, and good water stability.^11–14 Unfortunately, the highest saturated mass magnetization of Fe₃O₄ NPs is too low to provide an ideal imaging result. Thus, high injection dosage of Fe₃O₄ NPs is needed to provide practical sensitivity in the MRI diagnosis, which may cause an overdose of the Fe₃O₄ NPs and furthermore lead to toxicity and immune responses. Therefore, searching for a new MRI contrast agent, which has a higher-performance magnetization, is vital to precisely monitor early-stage diseases.

Magnetic Fe NPs have a powerful magnetization and their highest mass magnetization is larger than that of Fe₃O₄ NPs (Fe, 218 emu g⁻¹, Fe₃O₄, 73.7 emu g⁻¹).^15,16 But, they do not get a wide range of application as contrast agents in MRI owing to their instability. Once exposed to air, they are quickly oxidized into various oxide NPs with much reduced magnetization.^17,18 Hence, there is a great need to fabricate Fe NPs with excellent stability and water dispersity to be used as MRI agents in clinic. Unfortunately, there is little work related to the fabrication of stale Fe NPs, which can be used as MRI contrast agents. Herein, we reported a kind of mPEG-DSPE modified Fe/graphene core–shell NPs (Fe/G-PEG NPs), which not only possessed superior stability against oxidation, but also good dispersity in water. The highest saturated magnetization of Fe/G-PEG NPs was 215 emu g⁻¹, which was much higher than that of Fe₃O₄ NPs. To verify the MRI ability of Fe/G-PEG NPs, we adopted a patient derived xenograft (PDX) tumour and transplanted it mouse models. Different from the cell line-derived xenografts, which usually have a poor differentiated histology and lack of tissue organization, the PDX maintains the cell differentiation and morphology of the original patient tumours. PDX bearing mice, having the similar vasculature, central necrosis, molecular signatures and peripheral growth with that of the patient's tumor,^19,20 can act as an emerging platform for preclinical assays owing to its human tumour development characteristics.^21–23 Here, owing to their strong magnetization, MRI images with high resolution were obtained by intravenously injecting low dosage of Fe/G-PEG NPs (0.5 mg kg⁻¹ weight) into the PDX bearing nude mice, which would result in low toxicity against the normal tissue.

Fe/G NPs were synthesized by the combination of the hydrothermal and reduction reactions.²⁴ After the conjugation of PEG to the Fe/G NPs, Fe/G-PEG NPs were obtained and their structure was verified by XRD analysis as shown in Fig. 1A, which confirmed that these as-synthesized NPs were Fe NPs and all the peaks could be well indexed to that of α-Fe (BCC phase, Im3̄m space group, JCPDS (6-0696)) with a lattice constant of 2.886 Å. As we known, Fe nanoparticles are not stable and easily oxidized in air even at room temperature. However, in this work, there were no peaks of iron oxides in the XRD curve of Fe/G-PEG NPs, indicating that the coating layer of graphene prevented the inner core of Fe NPs from being oxidized during the conjugation procedure. The hydrodynamic size of as-prepared Fe/G-PEG NPs was about 253 ± 12 nm (Fig. S1†). After 2 months, their size slightly rose to 260 ± 7 nm (Fig. S2†), indicating that these Fe/G-PEG NPs were quite stable in the aqueous solution, which was quite valuable when they were used as MRI enhanced contrast agents in vivo. The size and morphology of the Fe/G-PEG NPs were also measured by TEM. As shown in Fig. 1B, it was clearly observed that the size of Fe/G-PEG NPs ranged from 75 to 125 nm and the Fe NPs were embedded in the corrugated graphene sheet. From the enlarged image shown in Fig. S3,† a layered structure together with an interface between Fe and graphene was clearly observed and the interplanar d-spacing equalling 0.32 nm could be ascribed to the 110 lattice of Fe NPs the surface modification with mPEG-DSPE was also confirmed by FT-IR analysis (Fig. 1C). There is no typical absorption peak in the spectrum of original Fe/G NPs. However, remarkable peaks centred at 1108, 2885, 3461 cm⁻¹ were clearly observed in the spectrum of Fe/G-PEG NPs, which could be attributed to the typical vibration peaks of –C–O–C–, –CH and –OH from mPEG-DSPE respectively, consequently proving the existence of mPEG-DSPE on the surface of Fe/G-PEG NPs. From Fig. S4,† we found that these Fe/G NPs would be deposited on the bottle of the tube in the PBS even after several minutes, whereas these Fe/G-PEG NPs could be homogeneously dispersed in the PBS solution for several days. These results also proved that the modification of Fe/G NPs with mPEG-DSPE greatly improved the water stability of Fe/G-PEG NPs. As a precondition for these Fe/G-PEG NPs to be used as MRI agents, they should possess super magnetic property. Therefore, the magnetic hysteresis loops for Fe/G-PEG NPs was measured at room temperature. As shown in Fig. 1D, a typical ferromagnetic curve with small coercivity was observed in the loop of Fe/G-PEG NPs. According to the mass percent of Fe in Fe/G-PEG NPs, the calculated magnetization of the Fe/G-PEG NPs was 215 emu g⁻¹, which was very close to the saturated magnetization of bulk Fe (217 emu g⁻¹). Fe₃O₄ NPs are the most common magnetic material and have been used as MRI enhanced contrast agents in clinic. So in this work, Fe₃O₄ NPs were used as positive control and their magnetic hysteresis loop was also measured (Fig. 1D). We observed that the saturated magnetic magnetization of Fe₃O₄ NPs was only 65 emu g⁻¹, which only accounted one third of the value of Fe/G-PEG NPs. All these above mentioned results inspired us that these Fe/G-PEG NPs could be used as MRI enhanced contrast agents due to their excellent water stability and super magnetic property.

Fig. 1 (A) XRD pattern of Fe/G-PEG NPs. (B) TEM images of Fe/G-PEG NPs. (C) FT-IR spectral of Fe/G NPs with and without mPEG-DSPE surface modification. (D) The magnetic hysteresis loop for Fe/G-PEG NPs and Fe₃O₄ NPs.

Before moving to the in vivo MRI experiment, the in vitro MRI ability of these Fe/G-PEG NPs was investigated. Fig. 2A shows the in vitro MRI images of Fe/G-PEG NPs and Fe₃O₄ NPs. Clearly, Fe/G-PEG NPs showed darker MRI images than that of Fe₃O₄ NPs with the same concentration ranging from 0 to 0.4 mM. Fig. 2B reveals the inverse relaxation times 1/T₂ of Fe/G-PEG NPs and Fe₃O₄ NPs at different concentrations. The value of 1/T₂ increases with the increase of iron concentration for both samples, which is reasonable because higher amount of magnetic NPs will increase proton relaxation rate and shorten the transverse relaxation time (T₂).^25,26 Besides, the transverse relaxivity (r₂) is 405 and 132 mM⁻¹ s⁻¹ for Fe/G-PEG NPs and Fe₃O₄ NPs respectively and Fe/G-PEG NPs have higher values of r₂ than that of Fe₃O₄ NPs. These results obviously illustrate that Fe/G-PEG NPs have a better MRI efficacy than that of Fe₃O₄ NPs, and can be used as T₂ contrast agents for MRI diagnosis. To verify the suitability of these Fe/G-PEG NPs to be used as in vivo MRI contrast agents, the cytotoxicity of Fe/G-PEG NPs was evaluated. Results of MTT assay indicated that Fe/G-PEG NPs exerted negligible cytotoxic effect on cell viability with concentration ranging from 0.1 to 5000 μg mL⁻¹ after 24 hours incubation (Fig. 2C). A slight reduction in cell viability (17.2%) post 24 hours incubation with 5000 μg mL⁻¹ of Fe was observed, revealing that the Fe/G-PEG NPs featured low cytotoxicity and were suitable for in vivo applications.

Fig. 2 (A) T₂-Weighted MR images of Fe/G-PEG NPs and Fe₃O₄ NPs at different concentration. (B) T₂ relaxation rates as a function of concentration (mM) for Fe/G-PEG NPs and Fe₃O₄ NPs. (C) Cytotoxicity profiles of HCC cells post 24 h incubation with Fe/G-PEG NPs via MTT assay.

Fig. 3A is the MRI images obtained from the normal mice. All experiments are ethical and passed by the relevant departments, and all the protocols for animal test were reviewed and approved by committee on animals in Nanjing University (China) and carried on according to guidelines given by National Institute of animal care the first row was tested without injection while the second row and the third row were tested 6 hours and 24 hours post the intravenous injection of Fe/G-PEG NPs at a dose of (0.5 mg kg⁻¹ weight) respectively. Every row showed 5 different layers from the same mouse. From the picture, it is found that the brightness of images obtained from the normal mice without the injection of Fe/G-PEG NPs keeps constantly, whereas a remarkably enhanced darkness was observed throughout the images of liver in the normal mice 6 hours post the intravenous injection of Fe/G-PEG NPs. These results are reasonable that the reticuloendothelial system of liver contains lots of Kupffer cells which can selectively engulf these Fe/G-PEG NPs.^27,28 More amounts of Fe/G-PEG NPs will be accumulated inside the liver as the time is extended. Because Fe/G-PEG NPs is a kind of T₂ weighted MRI contrast agent, it provides the negative signal in the MRI diagnostic measurement. When more Fe/G-PEG NPs were in the liver, darker images were observed. However, after 24 hours, the MRI images of normal mice are brighter than those obtained at 6 hours, which might be due to the metabolism of Fe/G-PEG NPs, resulting in lower concentration of Fe/G-PEG NPs in the liver. It is no doubt that these Fe/G-PEG NPs have a great clinical potential owing to their wonderful MRI efficacy and their small administered dosage (usually 2–5 mg kg⁻¹ for Fe₃O₄ in mice^29,30). The in vivo bio-distribution of Fe in mice is shown in Fig. 3C. At 6 h post-injection, the Fe content in liver reaches maximum of about 16.8% dose per g tissues and 24 later most Fe/G-PEG NPs have been metabolised and Fe content in liver reaches minimum of about 2.3% dose per g tissues. These results are consistent with the qualitative data of in vivo MRI. Then the ability of these Fe/G-PEG NPs to diagnose the tumour bearing mice was investigated by intravenously injecting them into PDX bearing nude mice. As shown in Fig. 3B, before injection, there is no distinct discrimination between the normal tissue and tumour tissue in the MRI images. 6 hours after the injection, the tumour tissue is bright with distinct margin and no negative signal was observed, while a remarkably decreased signal was observed in the normal liver tissue (the tumour was depicted clearly with the arrows in Fig. 3B). It was because the liver tissue contained Kupffer cells which could take in these Fe/G-PEG NPs while the tumour tissue did not have the Kupffer cells, and less amounts of Fe/G-PEG NPs were entrapped in the tumour, resulting in a brighter image. This experiment convincingly demonstrated that Fe/G-PEG NPs showed enhanced MRI sensitivity for the detection of tumour in vivo. Based on above results, we believe that such kind of high-performance MRI probe can be potentially used in the clinic to diagnose the biological events effectively.

Fig. 3 (A) In vivo liver MRI of the healthy BALB/c mice without injection or 6 h and 24 h post the intravenous injection of Fe/G-PEG NPs at a dose of (0.5 mg kg⁻¹ weight) respectively. (B) In vivo liver MRI of the PDX bearing nude mice without injection or 6 h post the intravenous injection of Fe/G-PEG NPs at a dose of (0.5 mg kg⁻¹ weight) respectively. The arrows depict the carcinoma area, (C) in vivo bio-distribution of Fe/G-PEG NPs in mice.

To confirm that the high magnetic performance of Fe/G-PEG NPs in the MRI diagnosis related to the existence of Fe, the bio-distribution of Fe in the liver in tumour bearing mice was detected by the Prussian blue staining experiment. Fig. 4A provides a representative H&E staining image of liver section observed under a light microscope, and a remarkable boundary is found in the image. The left side of the boundary is the zone of hepatic carcinoma while the other side is the normal liver tissue. The H&E staining image indicates that the transplanted PDX can grow well in the liver in the nude mice, and ultrastructure and differentiation of HCC can be clearly distinguished. The presence of Fe in the normal liver tissue was probed by staining the Fe with Prussian blue. From Fig. 4B we can easily find the characteristic blue from the complex of Fe and Prussian blue, which confirms the presence of Fe in the liver. However, no obvious blue colour is observed in the area of HCC. The Prussian blue staining experiment also proved that the liver tissue contained Kupffer cells can take in Fe/G-PEG NPs.

Fig. 4 (A) Sections from the liver were stained with haematoxylin and eosin (H&E), (B) sections from the liver were stained with Prussian blue.

In summary, we synthesized a core–shell structure Fe/G-PEG NPs composed of a Fe core and a PEGylated graphene. These Fe/G-PEG NPs are stable at normal condition and well dispersed in PBS solution for a long time. Compared to the normally used Fe₃O₄ NPs, these Fe/G-PEG NPs had a higher magnetization and exhibited a more preferable MRI contrast ability in vitro. They featured low cytotoxicity and could distinguish the tumour and normal tissue in the in vivo MRI diagnostic procedure in tumour bearing mice. Due to their high sensitivity and low injection dosage, these Fe/G-PEG NPs could be a promising MRI enhanced contrast agent in clinic.

Associated content

Experimental methods for fabrication and characterization of Fe/G-PEG NPs, in vitro MRI study, in vitro cytotoxicity assay (MTT assay), in vivo MRI imaging, Fe bio-distribution in liver.

Author contributions

C. Zhang did the experiment, collected the data and wrote the paper; J. Ren and Y. T. Yang did the cell experiments; D. H. Wang provided the Fe/G NPs; J. He and Z. Y. Zhou perform the in vitro and in vivo MRI experiments and analyzed the data. D. Huo and Y. Hu designed the experiments, supervised the students. All authors contributed to discussion, data analysis, and the manuscript writing.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Nos. 21474047), the Natural Science Foundation of Jiangsu Province (BK20141225).

References

1. Y. Itai, K. Ohtomo, S. Furui, T. Yamauchi, M. Minami and N. Yashirou, Am. Journalism Rev., 1985, 145, 1195–1199.
2. Y. S. Kim, S. H. Lim, K. W. Oh, J. Y. Kim, S. H. Koh, J. H. Kim, S. H. Heo, D. Chang, Y. J. Lee and H. Y. Kim, Atherosclerosis, 2012, 224, 411–416.
3. H. Jadvar and P. M. Colletti, Eur. J. Radiol., 2014, 83, 84–94.
4. K. M. Verbeeten, K. L. Hermann, M. Hasselqvist, G. S. Lausten, P. Joergensen, C. M. Jensen and H. S. Thomsen, Eur. J. Radiol., 2005, 15, 165–169.
5. W. A. Willinek and H. H. Schild, Eur. J. Radiol., 2008, 65, 2–14.
6. D. A. Benaron, Cancer Metastasis Rev., 2002, 21, 45–78.
7. V. Marks, Chem. Eng. News, 2005, 83, 25–34.
8. D. Zhao, X. T. Li, X. D. Shi, K. Ye, W. Y. Liu, G. Qiu and X. H. Lu, RSC Adv., 2016, 6, 61001–61005.
9. C. T. Yang, P. Padmanabhan and B. Z. Gulyása, RSC Adv., 2016, 6, 60945–60966.
10. Y. X. Tang, Y. Liu, W. Li, Y. Xie, Y. J. Li, J. Wu, S. J. Wang, Y. Tian, W. Tian, Z. G. Teng and G. M. Lu, RSC Adv., 2016, 6, 62550–62555.
11. X. Wang, M. Q. Tu, K. Yan, P. H. Li, L. Pang, Y. Gong, L. Li, R. Q. Liu, Z. S. Xu and H. B. Xu, ACS Appl. Mater. Interfaces, 2015, 7, 24523–24532.
12. S. J. Xu, F. Yang, X. Zhou, Y. P. Zhuang, B. X. Liu, Y. Mu, X. Wang, H. Shen, G. Zhi and D. C. Wu, ACS Appl. Mater. Interfaces, 2015, 7, 20460–20468.
13. W. Feng, X. J. Zhou, W. Nie, L. Chen, K. X. Qiu, Y. Z. Zhang and C. L. He, ACS Appl. Mater. Interfaces, 2015, 7, 4354–4367.
14. X. D. Wang, H. Y. Liu, D. Chen, X. W. Meng, T. L. Liu, C. H. Fu, N. J. Hao, Y. Q. Zhang, X. L. Wu and J. Ren, ACS Appl. Mater. Interfaces, 2013, 5, 4966–4971.
15. S. Peng, C. Wang, J. Xie and S. H. Sun, J. Am. Chem. Soc., 2006, 128, 10676–10677.
16. P. F. Su, G. Chen and J. Zhao, Chem. J. Chin. Univ., 2011, 7, 1472–1477.
17. L. M. Lacroix, S. Lachaize, A. Falqui, M. Respaud and B. Chaudret, J. Am. Chem. Soc., 2009, 131, 549–557.
18. E. Snoeck and C. Gatel, Nano Lett., 2008, 8, 4294–4298.
19. E. Marangoni, A. Vincent-Salomon, N. Auger, A. Degeorges, F. Assayag, P. D. Cremoux, L. D. Plater, C. Guyader, G. D. Pinieux, J. G. Judde, M. Rebucci, C. T. Perennou, X. Sastre-Garau, B. Sigal-Zafrani, O. Delattre, V. Dieras and M. F. Poupon, Clin. Cancer Res., 2007, 13, 3989–3998.
20. D. Siolas and G. J. Hannon, Cancer Res., 2013, 73, 5315–5319.
21. M. Hidalgo, F. Amant, A. V. Biankin, E. Budinská, A. T. Byrne, C. Caldas, R. B. Clarke, S. D. Jong, J. Jonkers, G. M. Mælandsmo, S. R. Roman, J. Seoane, L. Trusolino and A. Villanueva, Cancer Discovery, 2014, 4, 998–1013.
22. J. J. Tentler, A. C. Tan, C. D. Weekes, A. Jimeno, S. Leong, T. M. Pitts, J. J. Arcaroli, W. A. Messersmith and S. G. Eckhardt, Nat. Rev. Clin. Oncol., 2012, 9, 338–350.
23. K. T. Jin, L. S. Teng, Y. P. Shen, K. F. He, Z. Z. Xu and G. L. Li, Clin. Transl. Oncol., 2010, 12, 473–480.
24. X. C. Zhao, Z. M. Zhang, L. Y. Wang, K. Xi, Q. Q. Cao, D. H. Wang, Y. Yang and Y. W. Du, Sci. Rep., 2013, 3, 3421–3424.
25. N. Nasongkla, E. Bey, J. Ren, H. Ai, C. Khemtong, J. S. Guthi, S. Chin, A. D. Sherry, D. A. Boothman and J. Gao, Nano Lett., 2006, 6, 2427–2430.
26. J. M. Zhao, J. Liu, Y. B. Lin, G. L. Zong, L. Z. Wu, H. S. Qian and J. X. You, Chin. J. Magn. Reson. Imaging, 2014, 5, 138–143.
27. S. W. Park, J. W. Kang and S. M. Lee, J. Surg. Res., 2015, 194, 91–100.
28. L. J. Dixon, M. Barnes, H. Tang, M. T. Pritchard and L. E. Nagy, Comprehensive Physiology, 2013, 3, 785–797.
29. H. Yang, Y. M. Zhuang, Y. Sun, A. T. Dai, X. Y. Shi, D. M. Wu, F. Y. Li, H. Hu and S. P. Yang, Biomaterials, 2011, 32, 4584–4593.
30. S. Cheong, P. Ferguson, K. W. Feindel, I. F. Hermans, P. T. Callaghan, C. Meyer, A. Slocombe, C. H. Su, F. Y. Cheng, C. S. Yeh, B. Ingham, M. F. Toney and R. D. Tilley, Angew. Chem., Int. Ed., 2011, 50, 4206–4209.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section. See DOI: 10.1039/c6ra23511e

---