2.1.

Synthesis and Peptide Conjugation of Au NPs

Au NPs were synthesized according to the well-established citrate-reduction method.^{42, 43} Briefly, $50\mathrm{mL}$ of a 0.01% (by weight) auric acid (Sigma-Aldrich) aqueous solution was heated to boiling while stirring in a $100\text{-}\mathrm{mL}$ beaker. Next, $350\mu \mathrm{L}$ of a 1% (by weight) trisodium citrate (Sigma-Aldrich) aqueous solution was quickly added. The solution changed color within several minutes from yellow to black and then to red, indicating NP formation. The absorption spectrum of the NPs measured by a UV-Vis spectrometer (Ocean Optics) shows the surface plasmon resonance wavelength at around $526\mathrm{nm}$ . Transmission electron microscopy (TEM, JEOL 100CX II) and dynamic light scattering (DLS, ZetaSizer Nano ZS90) show the size of the NPs at about $30\mathrm{nm}$ in diameter.

Before peptide conjugation, Au NPs were modified with thiolated poly(ethylene) glycol (PEG). This was done by mixing $0.8\mathrm{nM}$ citrate capped Au NPs (calculated concentration using ${\epsilon }_{\mathrm{abs}}=3.0\times {10}^9{\mathrm{M}}^{-1}{\mathrm{cm}}^{-1}$ determined by Hurst ⁴⁴) with $2\mu \mathrm{M}$ methoxy-PEG-thiol (mPEG-SH, molecular weight 5000, Laysan Bio, Inc.) for $2\mathrm{h}$ at room temperature. The PEGylated Au NPs were centrifuged at $\mathrm{14,000}\mathrm{rpm}$ for $20\min$ and redispersed in PBS (phosphate-buffered saline) buffer to a final optical density of 2.0. The NLS (CGGGPKKKRKVGG), originating from the SV-40 large T antigen^{45, 46, 47} and ${\alpha }_{\mathrm{v}}{\beta }_3$ integrin targeting arginine-glycine-aspartate containing peptide (RGDRGDRGDRGDPGC) were bound to Au NPs via Au-S bonds at the cysteine residues. This was done by mixing $50\mu \mathrm{M}$ RGD alone (for cytoplasmic delivery) or both $50\mu \mathrm{M}$ RGD and $50\mu \mathrm{M}$ NLS (for nuclear delivery) with $1\mathrm{nM}$ PEGylated Au NPs. The mixtures were allowed to react overnight. The peptide-conjugated Au NPs were purified by centrifugation at $\mathrm{14,000}\mathrm{rpm}$ for $20\min$ and then redispersed in PBS buffer. The surface modification at each step was monitored by changes in the absorption spectra, size, and zeta potential.

2.2.

Cell Culture and Au NP treatment

Human oral squamous cell carcinoma (HSC-3) cells were cultured in Dulbecco’s modification of Eagle’s medium (DMEM, Cellgro) plus 10% fetal bovine serum (FBS, Gem Cell) at $37°\mathrm{C}$ under 5% $\mathrm{C}{\mathrm{O}}_2$ . The cells were first placed in a $35\text{-}\mathrm{mm}$ glass-bottomed tissue culture dish and allowed to grow for $2\text{days}$ . Then a given volume of Au NPs with an optical density of 2.0 was added to the medium to a final concentration of $0.05\text{to}0.4\mathrm{nM}$ . The cells were incubated with the NP-containing medium for $24\mathrm{h}$ . Then, the cells were washed with $1\times$ PBS, immersed in fresh medium, and put in the cell incubation minichamber on the microscope for real-time dark-field live cell imaging. For the particle-tracking experiments, the cell dish was put into the minichamber for imaging immediately after Au NP addition.

2.3.

Real-Time Dark-Field Light Scattering Imaging

Figure 1 shows the instrumental setup of the real-time dark-field light scattering imaging system. This system is a combination of a commercial inverted Olympus IX70 microscope, an angled white light illumination system, and a homebuilt environmental cell culture minichamber. In the conventional dark-field microscope, a high-NA condenser is used to deliver a narrow beam of white light onto the top of the sample. In our study, the dark-field illumination system was angled with respect to the sample plane, similar to the setup described by Yguerabide ³⁰ This side-illumination arrangement enabled for the location of a homebuilt environmental cell culture minichamber on the microscope stage. The side illumination system was composed of a Fiber-Lite MI-150 white light illuminator (Dolan-Jenner Industries), an optical fiber delivering the light closer to the sample, and a condenser lens focusing the light onto the sample. Using a sample of dried Au NPs on a microscope cover glass, the illumination angle and the focus were optimized so that there was no uneven illumination and the transmission light did not enter the objective ( $40\times$ iris, U Plan FL, NA 0.6, WD2.7). At the sample position, the light beam was about $2\mathrm{mm}$ in diameter.

Fig. 1

Imaging setup and particle targeting strategy. (a) Schematic setup of real-time light scattering imaging system for live cell studies. A dark-field light scattering imaging microscope with a side-illumination arrangement is combined with a homebuilt environmental cell incubation chamber for long-time study of live cells. (b) Absorption spectra of Au NPs before and after surface modification. Surface modification with mPEG-SH molecules induces an $\sim 3\text{-}\mathrm{nm}$ red shift of the localized surface plasmon resonance wavelength of the Au NPs. Additional binding of RGD/NLS or RGD leads to a further $2\text{-}\text{to}3\text{-}\mathrm{nm}$ red shift of the plasmon wavelength maximum of the nanoparticles. (c) Typical images of HSC-3 cancer cells after incubation with $0.1\mathrm{nM}$ PEG conjugated (left), RGD/NLS/PEG conjugated (middle) and RGD/PEG conjugated (right) Au NPs for overnight in cell culture medium at $37°\mathrm{C}$ . While the PEG conjugated Au NPs do not bind to the cells, RGD/NLS/PEG conjugated ones were accumulated in cell nuclear region and the RGD/PEG conjugated ones were accumulated in cytoplasm. Scale bar: $10\mu \mathrm{m}$ . (Color online only.)

The environmental chamber, which was made of a double wall of organic glass and a copper bottom with a hole in the middle for light transmission, was connected with a heated/refrigerated circulator (VWR) and with a gas tank containing 5% $\mathrm{C}{\mathrm{O}}_2$ in air. A temperature of $57°\mathrm{C}$ of the water in the circulator tank gave a temperature of $37°\mathrm{C}$ of the medium in the cell culture dish in the chamber. The opened water channel in the wall of the chamber served as a humidity reservoir as well to prevent culture from desiccation (humidity: $>95\mathrm{\%}$ ). The $\mathrm{C}{\mathrm{O}}_2$ flow rate was monitored by a solid acrylic flow meter (Davis Instruments) and adjusted to around 2 scfh (standard cubic feet per hour) to maintain cell culture medium pH. This environmental cell culture system enabled live cell studies in real time over $30\mathrm{h}$ . The images were taken using a Nikon D200 digital camera. A blank microscope cover glass was used to perform a manual white balance of the camera.

3.1.

Monitoring Nuclear Uptake of Peptide-Conjugated Au NPs

Nuclear targeting of NPs is a topic of great importance not only for understanding fundamental gene functions but also for developing gene delivery technology. Nuclear delivery is largely dependent on the use of nuclear localization signals, amino acid sequences that viruses generally use for nuclear entry via interaction with the nuclear pore complex.^{48, 49} The nuclear localization of Au NPs modified with an NLS peptide from the SV-large T antigen was reported by Xie using TEM and surface-enhanced Raman spectroscopy⁵⁰ and by Ryan using inductively coupled plasma optical emission spectroscopy (ICP-OES) and nuclear fractionation techniques.⁵¹ In our previous study, we observed the nuclear localization of NLS-conjugated Au nanorods using a dark-field microscope and confirmed it with surface-enhanced Raman spectroscopy.⁴⁷ In this work, we modified NLS-conjugated Au NPs by adding a linear RGD peptide to facilitate nuclear targeting of cancer cells, taking advantage of the specific binding of RGD peptides to ${\alpha }_{\mathrm{v}}{\beta }_3$ integin receptors overexpressed on many types of cancer cells and tumor vasculature.^{52, 53, 54} To overcome NP aggregation by the peptide molecules, we stabilized the NPs with a minimal amount of PEG in advance. Subsequent addition of both RGD and NLS produces Au NPs capped with a mixed monolayer of RGD, NLS, and PEG molecules. Surface modification with mPEG-SH molecules induces about a $3\text{-}\mathrm{nm}$ red shift of the NP-localized surface plasmon resonance wavelength [Fig. 1] and a $10\text{-}\mathrm{nm}$ increase on the hydrodynamic diameter. The zeta potential dramatically dropped from $-40\text{to}-10\mathrm{mV}$ after PEGylation (Table 1 ). Additional binding of RGD/NLS leads to a further $2\text{to}3\text{-}\mathrm{nm}$ plasmon wavelength shift and a $5\text{-}\mathrm{nm}$ hydrodynamic size change due to the conformational change of PEG molecules after adding peptides. Comparatively, for cytoplasmic NP delivery, the NPs were conjugated with PEG and RGD only. Surface characterization with an absorption spectrum and DLS shows similar changes for dual functionalization. The difference in zeta potential is due to the positive charge contributed by the NLS peptides. Figure 1 shows that PEG-conjugated Au NPs do not bind to the HSC-3 cancer cells. The RGD/PEG-conjugated Au NPs were found in the cytoplasm only. The addition of the NLS peptide leads to a predominantly nuclear accumulation of NPs. Our recent study⁵⁵ shows that the NLS/RGD/PEG-conjugated Au NPs, at a high concentration, cause DNA damage in cancer cells, but the RGD/PEG-conjugated Au NPs do not. This confirms the success of nuclear targeting by the dual functional Au NPs.

Table 1

Physicochemical properties of plain and surface modified Au NPs with PEG, NLS/RGD/PEG, and RGD/PEG molecules.

Using the real-time light scattering imaging system, we could continuously observe, over hours, how the NPs enter into cells and how they accumulate in the nuclear region. The transport of NLS/RGD/PEG-conjugated Au NPs into single cells within $15\mathrm{h}$ is shown in Fig. 2 . Consecutive dark-field light scattering images show the full process of the nuclear localization of NPs in the presence of $0.05\mathrm{nM}$ NLS/RGD/PEG-conjugated Au NPs [Fig. 2], which shows that NPs are found in nuclear region within $3\mathrm{h}$ . Massive nuclear accumulation takes about $13\mathrm{h}$ , which is consistent with results reported by the literature using a commercial incubator.⁵⁰ This confirms that our homebuilt minichamber used for imaging live cells does provide a growth environment very similar to that offered by the commercial incubator.

Fig. 2

Real-time tracking of the nuclear uptake of RGD- and NLS-conjugated Au NPs in HSC-3 cancer cells. (a) Time-dependent nuclear uptake of RGD/NLS/PEG-conjugated Au NPs (also see supplementary information in Video 1). (b) Changes of color of Au NPs in the three representative stages of nuclear localization, the color of the Au NPs changes from cyan at low density to yellow at higher density after they transport from the cytoplasm (left) to a smaller volume nucleus (right). For both (a) and (b), the cells were immersed in cell culture medium at $37°\mathrm{C}$ overnight in the presence of $0.05\text{-}\mathrm{nM}$ RGD/NLS/PEG-conjugated Au NPs. Scale bar: $10\mu \mathrm{m}$ . (Color online only.)

The observed nuclear localization indicates that the NPs do escape from the endosome by a yet unknown mechanism and are able to gain access to the nucleus. From Fig. 2, we can see that the Au NPs are first distributed in the cytoplasm ( $1\text{-}\mathrm{h}$ image). Then Au NPs penetrate the nuclear membrane and accumulate in the nucleus over time. Finally, the NPs were found to be localized predominately inside the nuclear region after $7\mathrm{h}$ (Video 1 ). The nuclear internalization is possibly via the following steps. First, the NPs must enter into the cytoplasm, most probably by receptor-mediated endocytosis because the RGD on the NPs binds to the integrin receptors on the cancer cell surface. Then the NPs escape from the endosome and transport to the nuclear region. Finally, the NPs transpass nuclear membrane and are accumulated in the nuclear region. There is an alternative way for the RGD/NLS/PEG-conjugated Au NPs to accumulate in the nuclear region after long-term internalization. First, the NPs enter into the cell via the endocytic pathway, probably via receptor-mediated endocytosis. Second, the NPs move toward the Golgi apparatus and endoplasmic reticulum (ER) via retrograde transport and late endosomes to the Golgi apparatus pathway. Third, once at the ER via retrograde transport, the NPs have access to the nucleus envelope and finally to the nuclear region.

Video 1

Real-time observation of nucleus internalization of NLS Au NPs into cancer cell nucleus (QuickTime, 6 MB).

It is possible that the Au NPs inside the nuclear region bind to macromolecules, such as DNA, RNA, or proteins. As a matter of fact, molecular signatures from DNA and proteins in the nucleus were reported by our group⁴⁷ and other researchers⁵⁰ using Raman spectroscopy enhanced by the nuclear targeting Au NPs. Note that although we show here a $15\text{-}\mathrm{h}$ tracking experiment, this imaging technique can be used for real-time studies up to at least $30\mathrm{h}$ . This gives us the opportunities to study the full cycle of a single live cell, as well as the division of second-generation cells as shown in the following section.

Another interesting finding is the change in the color of Au NPs during the process of nuclear internalization. As shown in Fig. 2, the image on the left shows Au NPs in the cytoplasm. The image in the middle shows that some NPs penetrate the nuclear membrane and accumulate inside the nuclear region. The image on the right shows the majority of NPs initially in the cytoplasm is now localized in the nucleus. From these results, it can be seen that the color of the Au NPs change from cyan to yellow after transport from the cytoplasm to the nucleus. The red shift of the light scattering wavelength is attributed to the NP plasmonic coupling.^{56, 57, 58} As the plasmonic fields overlap on aggregation, the color shifts from blue to red.

Using the dark-field imaging system, we demonstrated that the behavior of light scattering Au NPs in living cells as well as that of the cells themselves can be monitored in real time over a long-time duration. However, like all other technologies, our current imaging system has some disadvantages. One of these disadvantages is the uneven illumination. Traditional dark-field microscopy setups are designed to provide even illumination from all sides. In our case, the cells were illuminated from one side. This uneven illumination could induce color artifacts in the images, as indicated by the blue shadows in Fig. 2. To eliminate such effects to a large extent, dried Au NPs on a microscope cover glass was used as a standard sample to optimize the illumination angle and the focus. The focus was made as sharp as possible and the transmission light does not enter the objective. In addition to the uneven illumination, there are other limitations, such as low spatial resolution in the direction perpendicular to the image plane and inability to performing multiplex imaging. Therefore, intensive research efforts are still required in the future to address these limitations.

3.2.

Tracking the Full Cycle of Cancer Cells from Birth to Division

Knowledge about the complex processes of the cell cycle is indispensable in understanding cancer formation and progression, and the signaling pathways responsible for human tumorigenesis. It is undoubted that the most straightforward method to investigate the cell cycle is imaging. For mammalian cells, it is a great challenge because of the very long period of the cell cycle (about $24\mathrm{h}$ ) and we could not rely on fluorescence microscopy. Our technique of real-time light scattering imaging enables continuous, intermittence-free, and long-time imaging, providing an approach for dealing with this challenge. Figure 3 shows that consecutive dark-field light scattering images of a living cancer cell in a full cycle from birth to division. Although the G1, S, and G2 phases cannot be clearly resolved, the M phase can be seen occurring between $13$ and $15\mathrm{h}$ . The morphological changes of the cell at these phases can also be observed in real time. While the whole cycle takes as long as over $16\mathrm{h}$ , the division takes only about $2\mathrm{h}$ . Note that the NPs remain in the nuclear region after division. This demonstrates that the instrument developed here is capable of real-time cell cycle studies at the single-cell level. This is helpful for investigating the dynamics of a single cell at different phases in a cell cycle.

Fig. 3

Real-time tracking of a full cycle of a HSC-3 cancer cell with Au NPs localized in the nuclear region from birth to division. The cells were incubated with $0.1\mathrm{nM}$ RGD/NLS/PEG-conjugated Au NPs in cell culture medium at $37°\mathrm{C}$ overnight before the tracking experiment. Using the enhanced real-time light scattering imaging technique, we were able to track the full cycle of a cell from birth to division and monitor the distribution of the nanoparticles in single cells in real time. Labels on the images indicate the time lapse. Scale bar: $10\mu \mathrm{m}$ .

3.3.

Following Nuclear Dynamics during Cell Mitosis

Using the imaging technique and Au NPs as contrast agents, we followed especially the mitosis process at a higher time resolution (Fig. 4 ). The mitosis of eukaryotic cells is a complex process involving chromosome condensation, nuclear envelope and lamina disassembly, spindle formation, and movement of varied nuclear proteins.^{59, 60, 61} Although fluorescence live-cell imaging techniques facilitate the detection of chromosomes and various cellular proteins, the study of dynamical properties of centrosomes and kinetochores, and gain us insights on the molecular mechanisms of tubulin,^{14, 15, 16, 17} the disadvantages of photobleaching and phototoxicity limit the applications in long-term imaging. The light scattering imaging technique assisted by Au NP contrast agents could be a potential alternative for imaging the complicated process of mitosis in real time. As shown in Fig. 4, subphases from prophase to cytokinesis during the course of mitosis can be well resolved by the light scattering Au NP contrast agents.

Fig. 4

Real-time tracking of the mitosis process of a HSC-3 cancer cell with Au NPs localized in the nuclear region. The motion of the chromosome at different stage of mitosis is resolved by the light scattering imaging of RGD/NLS/PEG-conjugated Au NPs in the nuclear region (also see supplementary information in Video 2). The cells were incubated with $0.1\mathrm{nM}$ RGD/NLS/PEG-conjugated Au NPs in a cell culture medium at $37°\mathrm{C}$ overnight before the tracking experiment. The time lapse shown in each image is as hour:minute. Scale bar: $10\mu \mathrm{m}$ .

Binding of NLS-conjugated Au NPs to chromosomes were previously confirmed by the detection of surface enhanced Raman signals from DNA in single cells,^{47, 50} so it is not surprising that the RGD/NLS/PEG-modified Au NPs in our study might bind to the chromosomes and enable us to track the chromosome dynamics. It is well known that the configuration of the chromosome keeps changing during mitosis, and the characteristic chromosome pattern is used to identify the different stages of mitosis. From Fig. 4, it is obvious that the evolution of the pattern of Au NP distribution inside of the nucleus exactly follows that of the chromosomes, indicating the binding of Au NPs to the chromosomes. In agreement with results in Fig. 3, it is found that Au NPs, initially inside the nucleus of the parent cell, remain in the nucleus of the daughter cells. Another interesting phenomenon is that Au NPs do not seem to divide equally between the two daughter cells (see Video 2 ). The capability of resolving every stage of the mitosis process at a single-cell level based on light scattering from peptide-conjugated Au NPs implicates that this new technology has great potential to provide us with new insights in understanding mitosis, a very important cellular process.

Video 2

Following the nuclear dynamics using Au NPs during cell mitosis (QuickTime, 5 MB).

3.4.

Observing Intracellular Distribution of Au NPs during Cell Division

To futher understand the behavior of Au NPs during cell mitosis, we tracked the distribution of Au NPs before and after cell division. By selectively conjugating Au NPs with specific molecules (NLS/RGD or RGD alone), we were able to deliver them either into the nucleus or into the cytoplasm. It is obvious that the final distribution of the Au NPs in daughter cells depends on their initial location in the parent cell (Fig. 5 ). Both NLS- and RGD-conjugated Au NPs, initially localized in the nucleus of the mother cell, are found in the nucleus of the daughter cells after division (see Video 3 ). This is consistent with the images shown in the Figs. 3 and 4. RGD-conjugated Au NPs initially localized in the cytoplasm of the mother cell, are found only in the cytoplasm of the daughter cells (see Video 4 ). It is well known that in mitosis, the nuclear envelope disintegrates during prophase and reforms in telophase. This implies that during the period of metaphase and anaphase, the environment of the nucleus and the cytoplasm completely mix with one another. However, Au NPs remain either in the nuclear or cytoplasmic regions after cell division. As explained previously, the preservation of Au NPs in the nuclear region is possibly due to the binding of the NPs to the chromosomes. The preservation in the cytoplasmic region is probably due to NP encapsulation in endosomes, which keep them inside the cytoplasm duiring mitosis.

Fig. 5

Motion and distributon of Au NPs during cell mitosis. When the particles are localized in the nuclear region of mother cells, they remain in the nuclear region of daughter cells. When they are localized in the cytopalsm of mother cells, they remain in the cytoplasm of daughter cells (also see Video 3 and 4). The cells were incubated with $0.1\mathrm{nM}$ RGD/NLS/PEG (top) and RGD/PEG (bottom) conjugated Au NPs in cell culture medium at $37°\mathrm{C}$ overnight before the tracking experiment. Scale bar: $10\mu \mathrm{m}$ . The time is shown as hour:minute.

Video 3

Movement and distributon of Au NPs initially inside the nucleus during cell mitosis (QuickTime, 1.79 MB).

Video 4

Movement and distributon of Au NPs initially inside the cytoplasm during cell mitosis (QuickTime, 2.09 MB).

3.5.

Visualizing the Effects of Au NPs on Cell Viability

Another application of this imaging technique is the observation of the effects of Au NPs on the viability of daughter cells. In our experiments, we found that Au NPs did not induce any effect on the cell growth or division, nor on the growth of daughter cells until the dose of RGD/NLS/PEG-conjugated Au NPs added to the culture medium was increased above $0.4\mathrm{nM}$ . In the specific case shown in Fig. 6 , we found that Au NPs were not equally divided between the two daughter cells, and the daughter cell with more Au NPs in the nucleus did not survive (see Video 5 ). Therefore, although Au NPs at a concentration of $0.4\mathrm{nM}$ do not have obvious effects on the parent cell (which remained alive until division), they could cause the death of the daughter cell in the later stages of cell division. This indicates the potential of developing Au NPs as anticancer drugs. Currently, we do not know much about the mechanism, but we have confirmed that the death of the daughter cell is a consequence of the DNA damage induced by the presence of the high density of Au NPs inside the cell nucleus.⁵⁵ Further investigation of the interaction between Au NPs and chromosomes will help us to develop Au NPs as drug for cancer therapy.

Fig. 6

Genotoxicity of high-dose Au NPs. In this case, the concentration of RGD/NLS/PEG-conjugated Au NPs is $0.4\mathrm{nM}$ . It is obvious that Au NPs are not equally divided among the two daughter cells and the daughter cell with more Au NPs in nucleus does not survive, as indicated by arrow (also see Video 5). The time is shown as hour:minute and the scale bar is $10\mu \mathrm{m}$ .

Video 5

Effects of bioconjugated Au NPs on the viability of daughter cells under a high dose (QuickTime, 1.5 MB).