Evidence for nuclear internalisation of biocompatible [60]fullerene¹⁾

Background

Many types of nanoparticles (NPs) have been shown to internalise within mammalian cells (1), but only a few have been observed to internalise within the cell nucleus-most likely due to the tightly-regulated nuclear membrane (2). Internalisation of NPs into the nucleus is desirable for several reasons, including their use as 1. transfection agents (3), 2. drug delivery platforms for drugs that act on DNA (4), and 3. hyperthermia-inducing agents for cancer therapy using non-invasive stimulation by radiofrequency irradiation (5), magnetic-field cycling (6), or photonic activation (7). For example, derivatised NPs, including protein-functionalised quantum dots (8) and peptide-functionalised gold NPs (9), have been shown to internalise into the nucleus. For underivatised NPs, single-walled carbon nanotubes (SWNTs), have been observed by direct transmission electron microscopy (TEM) imaging to also localise in the nucleus of human macrophage cells with dose-dependent cytotoxicity (10). Fullerene C₆₀ is another classic carbon-based NP, however it was not been shown to enter the cell nucleus until recently. In particular, a water soluble derivative of C₆₀ fluorescently labelled with a small molecule fluorophore was shown to enter cell nuclei through nuclear pore complexes in liver cancer cells (11). Here, we validate the nuclear internalisation ability of the C₆₀ derivative in several other cell types, further supporting the unique intracellular biodistribution property of this specific fullerene compound.

Results and Discussion

We present further evidence for nuclear internalisation of the fluorescently-labelled, water-soluble [60]fullerene derivative, 3′-ethoxycarbonyl-3′-({2-[6-(2-{3-[2-tert-butyl-7-(ethyl-{3-sulfopropyl}-amino)-chromen-4-ylidene]-prop-1E-enyl}-3,3-dimethyl-5-sulfonato-3H-indolium-1-yl)-hexanoylamino]-ethoxy}carbonyl)-3′′,3′′,3′′′,3′′′,3′′′′,3′′′′,3′′′′′,3′′′′′′,3′′′′′′′,3′′′′′′′-deca-(2-hydroxy-1-{hydroxymethyl} ethylcarbamoyl)-3′H,3′′H, 3′′′H,3′′′′H,3′′′′′H,3′′′′′′′H-hexacyclopropa[1,9:16,17:21,40:30,31: 44,45:52,60](C₆₀-I_h)[5,6]fullerene (abbreviated as C₆₀-serPF, structure shown in Figure 1). This observation is consistent with the fact that a similar unlabelled compound, C₆₀-ser (12), can be used as a transfection agent (13), as can other C₆₀ derivatives, both in vitro (14) and in vivo (15). The C₆₀-serPF conjugate is non-cytotoxic at the doses examined here (≤0.1 mg mL^–1) for HeLa (cervical cancer), SNU449 (liver cancer), SKOV3 (ovarian cancer), and HEK293T (immortalised fibroblast) cells. These cell types were chosen for study because (i) they are widely used (e.g., HeLa and HEK293T), allowing for easy standardised testing in the future with other fullerene derivatives; (ii) SNU449 is another liver cancer cell line, which may further support previous findings in liver cancer cells; and (iii) results in SKOV3 ovarian cancer cells may reveal the possibility of using the fullerene derivatives in other cancer applications.

Figure 1

The structure of C₆₀-serPF.

Our interest in the study of subcellular localisation of fullerenes led to the search for a fullerene derivative that can be used to establish the distribution of fullerenes in mammalian cells. Our previous work with C₆₀-ser (12, 16) established the compound to be highly water-soluble and biocompatible; therefore, we adapted the same fullerene scaffold for derivatisation with an appropriate fluorophore for intracellular trafficking studies. We settled on the PF633 fluorophore (PF), as shown in Figure 1, because (i) its absorption and emission peaks are at wavelengths longer than 600 nm, which avoids the C₆₀-ser absorption range in the visible spectrum, (ii) it is one of the more efficient fluorescent dyes with a strong absorption band and a high quantum yield for the emission band, (iii) it is more resistant to singlet oxygen production [from C₆₀-ser under visible light (17)] than other fluorescent labels that were tested, such as Fluorescent Red 610, Alexa Fluor 610, and Atto 635 (data not shown). The detailed synthesis and characterisation of C₆₀-serPF is published elsewhere (11). The absorption and emission spectra for C₆₀-serPF are shown in Figure S2B. Results of MTT viability assays (Figure S1) show no significant cytotoxicity (p-values >0.05) for the four cell lines examined due to exposure to the C₆₀-serPF concentration and longest incubation time used in this study. This observed biocompatibility for C₆₀-serPF is in sharp contrast to the behaviour of a less well-characterised colloidal fullerene material called “nC₆₀” which has been reported to be cytotoxic (18). Other well-characterised, water-soluble [60]fullerene derivatives such as the dicarboxyfullerenes, which concentrate in the mitochondria (19), or in the endosome- or lysosome-like vesicles (20) (but not in the nucleus), have been reported to be biocompatible like C₆₀-serPF.

The intracellular distribution of C₆₀-serPF in HeLa cervical cancer cells was investigated with confocal microscopy. Cells were incubated with C₆₀-serPF for 4 or 24 h, fixed, stained with DAPI, and imaged with confocal microscopy. Cellular uptake of C₆₀-serPF was observed in the majority of the cells at both timepoints (Figure S2). Since individual C₆₀-serPF molecules are only ∼3 nm in diameter, the images in Figure 2 must be of fullerene aggregates. Remarkably, 3-dimensional z-stacks obtained from confocal imaging indicate nuclear internalisation of C₆₀-serPF aggregates, as seen in the three orthogonal cross-sections (Figure 2A) and the 3D rendering model reconstructed from a z-section scan (Figure 2B, Movies S1 and S2). Interestingly, some fullerene aggregates appear associated with nucleoli found inside the nucleus (Figure 2A). Studies with the other cell lines showed similar findings (Figure S3). Further careful examination comparing fullerene intracellular distribution among various cell types may reveal important cell-type dependencies. Quantification of co-localisation indicated that approximately 27% and 21% (difference is not statistically significant), of the intracellular fullerene aggregates are within the nucleus after 4 and 24 h, respectively.

Figure 2

C₆₀-serPF aggregates imaged in 3-dimensions demonstrate intranuclear localisation. (A) Orthogonal cross-sections were reconstructed from a set of z-stack scans of C₆₀-serPF aggregates in a HeLa cell. Coloured arrows in the x-z (green and blue arrows) and y-z (yellow arrow) orthogonal planes correspond to intranuclear C₆₀-serPF aggregates (red) visible in the x-y plane. Nucleus is in blue. Scale bar is 5 μm. (B) 3D rendering of the z-stacks shows C₆₀-serPF aggregates (red) inside and outside the nuclear profile (blue). The nucleus is cut in two orthogonal planes (top and bottom panels) to display C₆₀-serPF aggregates within the nuclear volume. Scale bar is 5 μm.

C₆₀-serPF appears to enter cell nuclei through nuclear pore complexes (11). Since C₆₀-serPF is observed in the majority of cells after 4 h of incubation, breakdown of the nuclear envelope during mitosis is likely not a requirement since only a small percentage of cells would have undergone this phase of the cell cycle during such a short period of time. Future studies need to investigate the intracellular location of fullerenes that are not in the nucleus. It is presently unclear if the cytoplasmic fullerene aggregates are associated with other organelles or are membrane-free. Lastly, fluorescently-labelled C₆₀-serPF may have a different intracellular distribution compared to unlabelled C₆₀-ser, so care must be taken when extrapolating the data presented here to the behaviour of unlabelled C₆₀-ser or other types of functionalised [60]fullerenes. Fullerenes of different structure might exhibit variable nuclear localisation behaviour. Comparison of different fullerene derivatives may yield useful structure-function relationships that can be used to design other bioactive fullerene-based materials.

Conclusion

We have compelling evidence for the nuclear internalisation and accumulation of a water-soluble, fluorescently-labelled [60]fullerene. Further studies are necessary to uncover the biological mechanisms responsible for the observed intracellular distributions. The ability of C₆₀-serPF to enter the nuclear volume may have important implications in their use for biomedical applications, such as non-viral gene therapy and cancer chemotherapy, where delivery of active agents to the cell nucleus is desirable.