Cucurbit[7]uril-functionalized magnetic nanoparticles for imaging-guided cancer therapy

Ludan Yue , Chen Sun , Cheryl H. T. Kwong and Ruibing Wang *
State Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences, University of Macau, Taipa, Macau, China. E-mail: rwang@um.edu.mo

Received 4th February 2020 , Accepted 12th March 2020

First published on 12th March 2020


Herein we developed cucurbit[7]uril covalently modified Fe3O4 nanoparticles for facile surface modification via host–guest interactions to realize targeted drug delivery and magnetic resonance imaging of tumors in vivo.


Superparamagnetic iron oxide nanoparticles (Fe3O4 NPs) have drawn increasing interest due to their tunable sizes,1 strong superparamagnetism,2 and biocompatibility,3 allowing a plethora of biological applications including bio-separation and enrichment,4 and bio-imaging5 (magnetic resonance imaging (MRI)). As naked Fe3O4 NPs tend to agglomerate and precipitate in water,6 surface functionalization of Fe3O4 NPs is always necessary, not only to ensure their stability in aqueous solutions, but also to equip Fe3O4 NPs with the capability of targeted delivery, drug/gene loading, and fluorescent dye labelling for diverse biomedical applications. Generally, surface functionalization of Fe3O4 NPs is often achieved first via coordination interactions between the surface of Fe3O4 NPs and carboxylate-containing ligands, followed by subsequent functionalization via laborious covalent conjugation of small functional molecules or polymers such as PEG,7 drug molecules8,9 and/or dyes.10 A facile, modular surface functionalization has been highly desired for Fe3O4 NPs. On the other hand, supramolecular host–guest chemistry may provide a promising strategy for non-covalent conjugation, including loading and release of drugs on demand.11–13 For instance, cucurbit[7]uril (CB[7]) on the surface of NPs may act as drug carriers, and provide an anchor point for non-covalent, modular surface functionalization.14–16 In particular, Trabolsi and co-workers developed a payload delivery platform based on CB[7] stabilized γ-Fe2O3 NPs by directly anchoring CB[7] onto γ-Fe2O3 NPs surface through coordination of carbonyl groups of CB[7] to Fe3+ ions of NPs.16 Despite promising results obtained in vitro, no in vivo studies were performed. In addition, the non-ionic, weak coordination interactions between carboxyl oxygens of CB[7] and γ-Fe2O3 may lead to falling of CB[7] off the surface of γ-Fe2O3, in vivo application would not be feasible.
image file: d0tb00306a-s1.tif
Scheme 1 Schematic design of the OX/FA/CB[7]–Fe3O4 nanoplatform for targeted drug delivery and chemotherapy.

Herein we design and develop CB[7] covalently functionalized Fe3O4 NPs to allow further facile surface functionalization and drug loading/release for image-guided cancer therapy. In this work, mono-allyl-OCB[7] (AO1CB[7]) was covalently anchored onto the Fe3O4 NPs surface through a “click reaction” between the C[double bond, length as m-dash]C of AO1CB[7] and the thiol of the Fe3O4 surface ligand, which allowed non-covalent, tailorable surface modification of the NPs with adamantanamine (Ada)-conjugated functional tags via strong CB[7]–Ada host–guest recognition. In addition, CB[7] also allowed loading of oxaliplatin (OX) via CB[7]–OX host–guest interactions. As a proof of concept, OX loaded folate (FA)-functionalized Fe3O4 NPs demonstrated MRI-guided specific drug release for cancer therapy in vitro and in vivo (Scheme 1).

Fe3O4 NPs were prepared via a solvothermal method17 at high temperature (about 300 °C). As shown in Fig. 1A, the resulting hydrophobic Fe3O4 NPs had an average size of 7.40 nm as visualized by TEM. Fe3O4 NPs were subsequently transferred into an aqueous solution via a ligand exchange reaction to introduce sulfhydryl on the surface to yield hydrophilic Fe3O4–DMSA NPs, with a diameter of 7.45 nm and a negative surface potential of −24.03 mV (Fig. S1A and B, ESI). AO1CB[7] was prepared according to the literature method,18,19 and its structure was confirmed by 1H NMR (Fig. S2, ESI). CB[7]–Fe3O4 NPs were obtained via a “click reaction” under 254 nm UV irradiation. As shown in Fig. 1B, the morphology of CB[7]–Fe3O4 NPs was rather similar to that of Fe3O4 NPs, with moderate, ca. 3 nm, increase in size as visualized by TEM. The surface zeta potential increased to −7.65 mV (Fig. S1B, ESI), due to the modification of CB[7]. Subsequently, DLS analysis of both Fe3O4–DMSA NPs and CB[7]–Fe3O4 NPs revealed much larger size ranges for both NPs (Fig. S1C, ESI), attributed to the magnetism-induced agglomeration, similar to the previous observations on magnetic NPs.20,21 The amount of CB[7] on Fe3O4 NPs surface was determined to be 9.88 × 10−7 mol CB[7] per mg Fe3O4, by fluorescence assay14 (Fig. S3, ESI). To further demonstrate the presence of CB[7], FT-IR spectroscopy and thermo-gravimetric analysis (TGA) were performed. As shown in Fig. 1C, the stretching band at 1750 cm−1 corresponding to C[double bond, length as m-dash]O of CB[7] observed in the CB[7]–Fe3O4 NP spectrum was attributed to the conjugation of AO1CB[7] with Fe3O4 NPs. The stretching band at 1680 cm−1, corresponding to C[double bond, length as m-dash]C, disappeared due to the “click reaction”. In TGA analysis, as shown in Fig. 1D, Fe3O4 NPs showed no obvious weight loss between 370 to 480 °C, while AO1CB[7] exhibited a dramatic weight loss with the junction points at 370 and 480 °C, corresponding to the structural disruption of CB[7] in this temperature range. The junction points could also be observed in the mixture of AO1CB[7] and Fe3O4 NPs at 370 and 480 °C due to the thermolysis of CB[7]. CB[7]–Fe3O4 NPs exhibited no obvious junction points in the mentioned temperature range, instead, a smooth curve with a smaller slope was observed below 480 °C, indicating the slow endothermic process of breaking down of CB[7] on Fe3O4.


image file: d0tb00306a-f1.tif
Fig. 1 Materials characterization. Transmission electron microscopy (TEM) images and size distribution of (A) Fe3O4 NPs and (B) CB[7]–Fe3O4 NPs (scale bar: 20 nm). (C) FT-IR spectra of AO1CB[7] and CB[7]–Fe3O4 NPs. (Inset showing the zoomed view of the C[double bond, length as m-dash]C at about 1680 cm−1.) (D) The TGA (thermogravimetric analysis) curves of Fe3O4 NPs, AO1CB[7], CB[7]–Fe3O4 NPs, and the mixture of Fe3O4 and AO1CB[7].

Furthermore, the magnetic properties of CB[7]–Fe3O4 NPs, including superparamagnetism and MR imaging potential, were evaluated by hysteresis loop and T2 transverse relaxivity tests. As shown in Fig. 2A and B, the hysteresis intensities of Fe3O4 NPs and CB[7]–Fe3O4 NPs were 27.8 and 22.3 emu g−1, respectively, suggesting a moderately decreased hysteresis of Fe3O4 NPs upon surface modification with CB[7], owing to the presence of nonmagnetic CB[7]. Similarly, the T2 relaxation rates (1/T2) of Fe3O4 NPs and CB[7]–Fe3O4 NPs were calculated to be 9.788 and 9.152 s−1 mM−1, respectively, implying a promising potential for MRI imaging.


image file: d0tb00306a-f2.tif
Fig. 2 (A) Hysteresis loops of Fe3O4 NPs and CB[7]–Fe3O4 NPs. (B) 1/T2 (T2 relaxation rates) of Fe3O4 NPs and CB[7]–Fe3O4 NPs with linear correlation. (Inset: T2-Weighted MR images of Fe3O4 NPs (upper) and CB[7]–Fe3O4 NPs (lower), with the same Fe concentration and relaxivity values). (C) FT-IR spectra of OX, FA-Ada and OX/FA/CB[7]–Fe3O4 NPs. (D) Release profile of OX in the presence of 0, 0.1 and 1.0 mM SPM against incubation time.

To demonstrate that CB[7]–Fe3O4 NPs may get surface-functionalized via a noncovalent approach by inserting an Ada-conjugated functional tag, based on the strong binding affinity (Ka = 108 M−1)22 between Ada and CB[7], FA-Ada was prepared as a targeting tag. Using the same principle, a classic chemotherapeutic anti-tumor drug, oxaliplatin that has a strong binding affinity with CB[7] (Ka = 2 × 105 M−1)23 was selected as the model chemotherapeutic drug for loading onto the CB[7]–Fe3O4 NP surface in this work. The successful loading of FA-Ada and OX has been confirmed by the FT-IR spectrum. As shown in Fig. 2C, the appearance of a stretching vibration peak at 1570 cm−1, corresponding to C[double bond, length as m-dash]N of FA, indicated the existence of FA-Ada in OX/FA/CB[7]–Fe3O4 NPs. The stretching vibration peak at 3080 cm−1, corresponding to the cyclohexane group of OX, was observed in the OX/FA/CB[7]–Fe3O4 NP spectrum, indicating the existence of OX in OX/FA/CB[7]–Fe3O4 NPs. Additionally, the loading efficiency of OX was quantified to be 8.02% via ICP-MS analysis. Subsequently, the anchoring efficiency of FA-Ada was determined to be ca. 11.97%, calculated from the amount of CB[7] available on the surface of nanoparticles. The size of OX/FA/CB[7]–Fe3O4 NPs was visualized and found to be 12.64 nm by TEM (Fig. S4A, ESI), moderately larger than that of CB[7]–Fe3O4 NPs. The surface charge was also similar to that of CB[7]–Fe3O4 NPs (Fig. S4B, ESI). DLS analysis (Fig. S4C, ESI) also suggested strong magnetism-induced agglomeration,20 similar to that observed for CB[7]–Fe3O4 NPs. Due to the high concentration of spermine (SPM) in several types of tumor cells, OX can be competitively released from CB[7] in these cells.24 To quantitatively investigate the OX release profile in the presence of SPM, OX/FA/CB[7]–Fe3O4 NPs (at OX concentration of 50 μg mL−1) were placed in a dialysis tube (MW 3500 kDa) in PBS containing 0.1 mM (intracellular level in normal cells) or 1.0 mM (intracellular level in tumor cells) SPM at room temperature. The release profile was measured by ICP-MS according to the concentration of Pt. As shown in Fig. 2D, in the presence of 1.0 mM SPM, a rapid release of OX was observed, with 70.75% accumulated release after incubation for 72 h. In contrast, only 36.67% OX was released after incubation in 0.1 mM SPM solution for 3 days. Meanwhile, in the absence of SPM, only 16.62% OX leakage was observed, suggesting that OX may be released in a specific manner in these cancer cells.

The biocompatibility of CB[7]–Fe3O4 NPs was initially evaluated in human breast cancer (MCF-7) cells, mice breast cancer (4T1) cells and human hepatocyte line (L02) cells via MTT assays. As shown in Fig. S5 (ESI), CB[7]–Fe3O4 NPs (up to 250 μg mL−1) exhibited negligible toxicities against all of these three cell lines, which was also supported by the results of flow cytometry analysis (for details of the study see Fig. S6 in the ESI). Subsequently, we prepared Ada-conjugated FITC (FITC-Ada) to noncovalently functionalize CB[7]–Fe3O4 NPs with the fluorescent FITC for intracellular uptake study. MCF-7 cells, 4T1 cells and L02 cells were used respectively as tumor cell models and noncancerous cell model to investigate the cellular internalization behaviours of these NPs by confocal microscopy (CLSM). Each of the cell lines was cultured with free FITC, FITC/CB[7]–Fe3O4 NPs, and FITC/FA/CB[7]–Fe3O4 NPs (at a FITC concentration of 0.0067 mg mL−1), respectively, for 12 h before examination under a microscope. As shown in Fig. 3, strong green fluorescence signal was observed in the FITC/FA/CB[7]–Fe3O4 NP treated cancer cells (MCF-7 and 4T1), in a sharp contrast to the weak fluorescence observed in the FITC/CB[7]–Fe3O4 NP and free FITC-treated cells, confirming the active-targeting effects of FA. Conversely, no obvious fluorescence was seen in L02 cells in all the three groups of cells, likely owing to the low level of FA-receptors on the L02 cell membrane. Furthermore, the cellular internalization mechanism was further confirmed by pre-incubating MCF-7 and 4T1 cells with FA, before incubation with FITC/FA/CB[7]–Fe3O4 NPs, such that the FA-receptors would be pre-occupied by free FA. As shown in Fig. S7 (ESI), green fluorescence in FA pre-treated group was very weak in sharp contrast to the strong fluorescence in the group of cells without FA pre-treatment. These results suggest that active-targeting played an important role in the intracellular uptake of these NPs by the cancer cells, supporting the feasibility of noncovalent surface functionalization and suggesting a significant potential of CB[7]–Fe3O4 NPs in targeted drug delivery.


image file: d0tb00306a-f3.tif
Fig. 3 CLSM images of cellular uptake of FITC/FA/CB[7]–Fe3O4 NPs, FITC/CB[7]–Fe3O4 NPs and free FITC at a Fe concentration of 25 μg mL−1 and a FITC concentration of 0.0067 mg mL−1, respectively, by MCF-7, 4T1 and L02 cells upon 12 h incubation. Scale bar: 25 μm.

With excellent biocompatibility and high specific cellular uptake efficiency, the cytotoxicity of OX loaded NPs, OX/FA/CB[7]–Fe3O4 NPs, was further evaluated against MCF-7, 4T1 and L02 cells as an in vitro anticancer efficacy and safety study. Free OX was also investigated for comparison. As shown in Fig. 4, as expected, dose-dependent cytotoxicity was observed in all the three OX treated cell lines. The toxicity of OX/FA/CB[7]–Fe3O4 NPs was dramatically enhanced in cancer cell lines including MCF-7 and 4T1 cells, likely due to the improved cellular uptake and selective, and intracellular drug release. In contrast, the cytotoxicity of OX/FA/CB[7]–Fe3O4 NPs in L02 cells was attenuated in comparison with free OX, suggesting a good safety profile of this platform against non-cancerous cells. The results were supported by apoptosis rates obtained by flow cytometry assays (Fig. S8, ESI) against MCF-7, 4T1 and L02 cells. Collectively, these results demonstrated that OX/FA/CB[7]–Fe3O4 NPs, as a potential specific drug delivery platform, may show attenuate toxicity towards noncancerous cells and enhance targeted cytotoxicity to cancer cells.


image file: d0tb00306a-f4.tif
Fig. 4 Cytotoxicity of free OX and OX/FA/CB[7]–Fe3O4 NPs against MCF-7 cells (A and D), 4T1 cells (B and E) and L02 cells (C and F), at Pt concentrations of 0.34, 1.71, 3.42, 17.11, and 34.23 μM, respectively, after incubation for 24 and 48 h, respectively. The statistical significance levels of *p < 0.05, **p < 0.01 and ***p < 0.005 were calculated via two-way ANOVA.

To further evaluate the targeting efficiency and imaging potential, FA/CB[7]–Fe3O4 NPs were firstly studied for MRI imaging in vivo. As shown in Fig. 5 and Fig. S9 (ESI), 4T1 tumor-bearing balb/c mice were intravenously injected with FA/CB[7]–Fe3O4 NPs and CB[7]–Fe3O4 NPs, respectively, and were subsequently imaged using MRI at different time points post-injection. As expected, the FA/CB[7]–Fe3O4 NP treated mice exhibited the strongest MR signal in tumor (highlighted with red circles) at 2 h followed by a steady decrease and the signal was still observable at 12 h time point post-administration. In contrast, the MR signal in the mice injected with CB[7]–Fe3O4 NPs was too weak to be observed clearly, confirming again that the noncovalent functionalization of the NPs with a targeting tag may effectively increase targeted delivery efficiency and retention of the NPs at the tumor site. As the distribution and retention is mostly driven by the nanosize and the targeting molecule of NPs, it is reasonable to postulate that even with a therapeutic payload such as OX encapsulated into surface CB[7], the NPs would exhibit a rather similar, if not completely the same, tumor retention and clearance kinetics.


image file: d0tb00306a-f5.tif
Fig. 5 MRI images of mice post intravenous injection of CB[7]–Fe3O4 NPs and FA/CB[7]–Fe3O4 NPs (at 1 mg mL−1 Fe concentration) at 0, 2, 6 and 12 h, respectively.

To further investigate the therapeutic efficacy of OX/FA/CB[7]–Fe3O4 NPs in vivo, 4T1 tumor-bearing balb/c mice (tumor volume ≈ 50 mm3) were randomly divided into four groups (n = 3), and intravenously injected with: (A) OX/FA/CB[7]–Fe3O4 NPs, (B) FA/CB[7]–Fe3O4 NPs, (C) free OX, and (D) PBS control, at a Fe concentration of 100 μg mL−1 and a Pt concentration of 68.45 μM for 200 μL, respectively. As shown in Fig. 6A, the mice were treated once on Day 0 and subsequently monitored for 14 days, and the tumor size and body weight of all mice were recorded every two days and normalized for comparison. On the 14th day, all mice were sacrificed to collect the tumors. The representative tumors shown in Fig. 6B clearly demonstrated a significantly improved therapeutic outcome against 4T1 cancer in mice treated with OX/FA/CB[7]–Fe3O4 NPs, in comparison with other treatment groups. Indeed, as shown in Fig. 6C, the tumors grew rapidly in both PBS and FA/CB[7]–Fe3O4 NP treated groups, in a sharp contrast to the OX/FA/CB[7]–Fe3O4 NP treated group, where the tumor was effectively suppressed during the 14-day follow-up, confirming the excellent therapeutic effects of the platform against cancer in vivo. The growth of tumors in the OX treated mice slowed down to some extent; however, the mean size still showed an upward trend enduring the 14-day follow-up. In consideration of the potential of this platform for clinical application, we further investigated the side effects of these CB[7]–Fe3O4 NPs from the body weight evolution and histological analysis of major organs (Fig. 6D and 7). A modest increase of the body weight was observed in all of PBS, OX/FA/CB[7]–Fe3O4 NP and FA/CB[7]–Fe3O4 NP treated groups whereas an obvious decrease of the body weight was observed in the OX treated group, implying the excellent biocompatibility of these NPs and the obvious systemic toxicity of free OX in vivo. As shown in Fig. 7, the H&E stained tissues of major organs (including the heart, liver, spleen, lungs and kidneys) exhibited obvious spleen damage in free OX and PBS treated group due to tumor metastasis,25 while no obvious tissue damages or inflammation infiltration occurred in OX/FA/CB[7]–Fe3O4 NP or FA/CB[7]–Fe3O4 NP treated group, confirming the excellent biocompatibility and safety of CB[7]–Fe3O4 NP nanoplatform in vivo. On the other hand, the H&E stained tumor tissues of free OX treated group exhibited moderate level of nuclear chromatin condensation, as a result of tumor necrosis. Obvious nuclear fragmentation as well as cell shrinkage was observed in the group of mice treated with OX/FA/CB[7]–Fe3O4 NPs, suggesting a higher level of apoptosis than those treated with free OX alone. Conversely, the tumor tissues of mice treated with FA/CB[7]–Fe3O4 NPs showed morphologies similar to those of the PBS treated group, implying normal tumor growth without therapeutic effects in this group.


image file: d0tb00306a-f6.tif
Fig. 6 In vivo therapeutic effect. (A) balb/c mice bearing 4T1 tumor were divided into 4 groups for treatment under different conditions (PBS, OX, FA/CB[7]–Fe3O4 NPs, and OX/FA/CB[7]–Fe3O4 NPs, at a Fe concentration of 100 μg mL−1 and a Pt concentration of 68.45 μM for 200 μL, respectively) and were followed up for 14 days. (B) Representative harvested tumors from each group of mice. (C) Tumor volume changes and (D) body weight changes of the mice in each group during the 14 day follow up. The statistical significance levels of *p < 0.05 were calculated via two-way ANOVA. *, #, and & in (C) indicate statistically significant differences for the pairs of PBS vs. OX/FA/CB[7]–Fe3O4 NPs, OX vs. OX/FA/CB[7]–Fe3O4 NPs, and FA/CB[7]–Fe3O4 NPs vs. OX/FA/CB[7]–Fe3O4 NPs, respectively.

image file: d0tb00306a-f7.tif
Fig. 7 H&E stained tissue sections of major organs and tumors from different treatment groups (A) OX/FA/CB[7]–Fe3O4 NPs, (B) FA/CB[7]–Fe3O4 NPs, (C) Free OX, and (D) PBS control, harvested on Day 14 after treatment.

Conclusion

In summary, we developed a multifunctional theranostic platform based on CB[7]-covalently functionalized Fe3O4 NPs, herein referred to as CB[7]–Fe3O4 NPs, which allowed facile non-covalent surface modification and drug loading via strong host–guest recognition for MRI-image guided, targeted chemotherapy of cancer. As a proof-of-concept, amantadine conjugated-folic acid (FA-Ada), and chemotherapeutic agent oxaliplatin (OX) were respectively employed as a targeting molecule and a model drug anchored on the surface CB[7]–Fe3O4 NPs to enable targeted drug delivery and specific drug release. Furthermore, the core Fe3O4 NPs allowed MRI imaging that was employed to track their accumulation in vivo. More importantly, OX loaded FA/CB[7]–Fe3O4 NPs exhibited much improved therapeutic efficacy in vitro and in vivo, as well as superior safety profile against healthy cells and tissues/organs. This study may provide important new insights in designing versatile theranostic platforms that may allow facile, modular, and personalized surface functionalization.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (21871301) and the Research Committee at the University of Macau (Grant No. MYRG2017-00010-ICMS and MYRG2019-00059-ICMS).

References

  1. D. Ling, N. Lee and T. Hyeon, Acc. Chem. Res., 2015, 48, 1276–1285 CrossRef CAS PubMed.
  2. Y. Jun, J. Seo and J. Cheon, Acc. Chem. Res., 2008, 41, 179–189 CrossRef CAS PubMed.
  3. S. Wang, J. Luo, Z. Zhang, D. Dong and Y. Shen, Am. J. Cancer Res., 2018, 8, 1933–1946 CAS.
  4. N. Lee, D. Yoo, D. Ling, M. H. Cho, T. Hyeon and J. Cheon, Chem. Rev., 2015, 115, 10637–10689 CrossRef CAS PubMed.
  5. J. Xie, K. Chen, J. Huang, S. Lee, J. H. Wang, J. Gao, X. G. Li and X. Y. Chen, Biomaterials, 2010, 31, 3016–3022 CrossRef CAS PubMed.
  6. C. Boyer, M. R. Whittaker, V. Bulmus, J. Liu and T. P. Davis, NPG Asia Mater., 2010, 2, 23–30 CrossRef.
  7. P. Zhou, H. Zhao, Q. Wang, Z. Zhou, J. Wang, G. Deng, X. Wang, Q. Liu, H. Yang and S. Yang, Adv. Healthcare Mater., 2018, 7, 1701201 CrossRef PubMed.
  8. B. Pourbadiei, J. Nanomed. Res., 2017, 5, 114–119 Search PubMed.
  9. F. Ji, K. Zhang, J. Li, Y. Gu, J. Zhao and J. Zhang, J. Nanosci. Nanotechnol., 2018, 18, 4464–4470 CrossRef CAS PubMed.
  10. X. Peng, B. Wang, Y. Yang, Y. Zhang, Y. Liu, Y. He, C. Zhang and H. Fan, ACS Biomater. Sci. Eng., 2019, 5, 1635–1644 CrossRef CAS.
  11. Y. Cao, X. Y. Hu, Y. Li, X. Zou, S. Xiong, C. Lin, Y. Z. Shen and L. Wang, J. Am. Chem. Soc., 2014, 136, 10762–10769 CrossRef CAS PubMed.
  12. C. Sun, H. Zhang, S. Li, X. Zhang, Q. Cheng, Y. Ding, L. H. Wang and R. Wang, ACS Appl. Mater. Interfaces, 2018, 10, 25090–25098 CrossRef CAS PubMed.
  13. J. Zhou, G. Yu and F. Huang, Chem. Soc. Rev., 2017, 46, 7021–7053 RSC.
  14. L. Yue, C. Sun, Q. Cheng, Y. Ding, J. Wei and R. Wang, Chem. Commun., 2019, 55, 13506–13509 RSC.
  15. P. Xu, Q. Feng, X. Yang, S. Liu, C. Xu, L. Huang, M. Chen, F. Liang and Y. Cheng, Bioconjugate Chem., 2018, 29, 2855–2866 CrossRef CAS PubMed.
  16. F. Benyettou, I. Milosevic, Y. Lalatonne, F. Warmont, R. Assah, J.-C. Olsen, M. Jouaid, L. Motte, C. Platas-Iglesias and A. Trabolsi, J. Mater. Chem. B, 2013, 1, 5076–5082 RSC.
  17. S. Tanaka, Y. V. Kaneti, N. L. W. Septiani, S. X. Dou, Y. Bando, M. S. A. Hossain, J. Kim and Y. Yamauchi, Small Methods, 2019, 3, 1800512 CrossRef.
  18. Y. Ahn, Y. Jang, N. Selvapalam, G. Yun and K. Kim, Angew. Chem., Int. Ed., 2013, 52, 3140–3144 CrossRef CAS PubMed.
  19. D. Shetty, J. K. Khedkar, K. M. Park and K. Kim, Chem. Soc. Rev., 2015, 44, 8747–8761 RSC.
  20. A. Ray Chowdhuri, D. Bhattacharya and S. K. Sahu, Dalton Trans., 2016, 45, 2963–2973 RSC.
  21. K. K. Cheng, P. S. Chan, S. Fan, S. M. Kwan, K. L. Yeung, Y. X. Wang, A. H. Chow, E. X. Wu and L. Baum, Biomaterials, 2015, 44, 155–172 CrossRef CAS PubMed.
  22. M. A. Romero, N. Basilio, A. J. Moro, M. Domingues, J. A. Gonzalez-Delgado, J. F. Arteaga and U. Pischel, Chem. – Eur. J., 2017, 23, 13105–13111 CrossRef CAS PubMed.
  23. H. Chen, Y. Chen, H. Wu, J. F. Xu, Z. Sun and X. Zhang, Biomaterials, 2018, 178, 697–705 CrossRef CAS PubMed.
  24. Y. Chen, Z. Huang, H. Zhao, J. F. Xu, Z. Sun and X. Zhang, ACS Appl. Mater. Interfaces, 2017, 9, 8602–8608 CrossRef CAS PubMed.
  25. L. Yue, Z. Dai, X. Chen, C. Liu, Z. Hu, B. Song and X. Zheng, Nanoscale, 2018, 10, 17858–17864 RSC.

Footnote

Electronic supplementary information (ESI) available: Experimental section and supplementary results including 1H NMR of AO1CB[7], fluorescence assay, flow cytometry, and corresponding 1/T2 value in vivo. See DOI: 10.1039/d0tb00306a

This journal is © The Royal Society of Chemistry 2020