Real-Time Tracking of Yb³⁺, Tm³⁺ Doped NaYF₄ Nanoparticles in Living Cancer Cells

Introduction

The study of molecular transport, underlying biological processes, is one of the actively developing fields of science. The modern technical equipment for visualization and tracking of single molecules labeled with fluorescent probes allows real-time detection, identification, and tracking of individual molecules in biological systems [1]. The specificity of tracking in biological systems causes improved criteria for fluorescent dyes. In particular, it is necessary to ensure the consistency of the photophysical properties of the dye throughout the experiment. Thus, an important parameter is photostability (or resistance to photobleaching), i.e. the ability to maintain the photoluminescence (PL) intensity during prolonged excitation.

Organic dyes and genetically encoded fluorescent proteins are prone to rapid photobleaching and are not suitable for long-term studies. In addition, single fluorophores emit a relatively low number of photons, and their PL spectrum (~400–700 nm) falls within the region of high absorption of biological tissue, making it difficult to detect them. Inorganic luminescent nanoparticles are an alternative to fluorescent probes and proteins. Among them, semiconductor quantum dots have got widespread use for single molecules tracking. The successful use of quantum dots in tracking systems [2] is caused by a number of their advantages such as the exceptional brightness due to a high quantum yield (up to 90%) [3, 4], high photostability and resistance to chemical degradation. These properties make it possible to use quantum dots in long-term in vitroimaging [5, 6]. On the other hand, the presence of a fraction of non-radiating dark particles [7], as well as photoblinking [8] and cytotoxicity [9] limit their usage for tracking.

Upconversion nanoparticles (UCNPs) have recently emerged as a new class of inorganic photoluminescent nanoparticles that can successfully replace quantum dots [10]. UCNPs effectively convert the energy of absorbed infrared photons (~975 nm) into visible and near-infrared radiation. Unique photophysical properties of UCNPs, such as programmable spectra and narrow PL peaks, significant depth of optical detection in vivo [11], resistance to photobleaching [12, 13] and low cytotoxicity [14, 15], have provided their wide use in various types of biomedical research [16, 17].

Variety of inorganic matrices can serve as a base for UCNPs with a large selection of doping elements [16, 18]. However, the fluoride matrix co-doped with ytterbium and erbium or thulium ions (NaYF₄:Yb,Er/Tm) has got the most common use due to the high efficiency of upconversion (energy conversion). Remarkable photostability of UCNPs and the noninvasiveness of near-infrared excitation provide long-term visualization of processes of interest. The possibility was shown to visualize intracellular migration of NaYF₄ nanoparticles co-doped with Yb³⁺ and Er³⁺ for 6 h of continuous image acquisition [19]. Also, the use of similar UCNPs for rapid and background-free 3D visualization of dynamic processes in living cells was demonstrated [20]. Excitation of UCNPs in the infrared range provides an absence of cell autofluorescence, which almost completely eliminates the background signal out of focus and, therefore, greatly simplifies the determination of the 3D coordinates of UCNPs localization.

The aim of the study was to demonstrate the possibility of real-time tracking of NaYF₄:Yb,Tm upconversion nanoparticles coated with polyethylenimine in living cancer cells using wide-field microscopy technique.

Materials and Methods

Nanoparticles. Hexagonal NaYF₄ nanocrystals containing 8 mol% Tm³⁺, co-doped with 20 mol% Yb³⁺ [21] were used in the work. To achieve colloidal stability of NaYF₄:Yb,Tm UCNPs in aqueous solutions they were surface coated with polyethylenimine (UCNP-PEI) [14].

The transmission electron microscopy images were obtained using CM10 microscope (Philips, The Netherlands) with an accelerating voltage of 100 kV and a resolution of 5.0 Ǻ for a point and 3.4 Ǻ for a line.

Cells. The cell line SK-BR-3 (human breast adenocarcinoma, ATCC number HTB-30) was cultured in the McCoy’s 5A growth medium with 10% (v/v) fetal calf serum (HyClone, USA) and 2 mM L-glutamine in 5% CO₂ at 37°C. For passaging the cells were carefully detached using Versen solution (PanEco, Russia).

Tracking of UCNP-PEI in SK-BR-3 cells. SK-BR-3 cells were seeded in 96-well plates (BD Falcon, USA) at the density of 1·10⁴ cells per well and cultured for 24 h at 37°C in 5% CO₂. The growth medium was then replaced with PBS containing 5 μg/ml UCNP-PEI, and the cells were incubated for 2 h. After that, the UCNP-PEI solution was replaced with fresh growth medium, and the cells were incubated overnight.

Tracking of UCNP-PEI in SK-BR-3 cells was carried out using the home-built setup created on the basis of the Axio Observer A1 microscope (Carl Zeiss, Germany). The scheme of the setup is shown in Figure 1.

---

Figure 1. The scheme of the experimental setup for upconversion nanoparticles tracking in living cells LED is the light-emitting diode, EMCCD is the electron multiplying charge coupled device

---

The exciting radiation is generated by a laser diode LDD-10 (Semiconductor devices, Russia) at 975 nm. Then, the laser diode emission is assembled into a multimode optical fiber (200 μm diameter) connected to a collimator F280SMA-980 (Thorlabs, USA), which forms a Gaussian beam with a diameter of 2 mm. Further, a uniform illumination of the sample is created using a lens focusing on the back focal plane of an immersion oil objective C Plan Apochromat 63× NA1.4 Oil (Carl Zeiss, Germany). On the way to the sample, the excitation radiation is spectrally filtered by a laser clean-up filter LD01-975/10 (Semrock, USA), after that the radiation is reflected from the dichroic mirror ZT1064rdc-sp (Chroma, USA) and focused on the sample using the objective. The PL from the sample passes through the objective, dichroic mirror, and two successively set edge short-pass filters FF01-842/SP (Semrock, USA), which cut off laser radiation and transmit the spectral range of the PL registration (420–842 nm). The image is captured through the side port of the microscope to the iXon3 897 EMCCD camera (Andor, UK).

The excitation of the samples was carried out in a continuous mode at 975 nm with a radiation intensity of 950 W/cm² and an image acquisition time of up to 30 min. Registration of epi-luminescence images was performed in a mode of continuous image acquisition (frame transfer kinetics) with a frequency of ~5 Hz and an exposure of each frame of 200 ms. Using a red LED, the sample was illuminated every 5 frames to obtain brightfield images in which the location and morphology of the cells were determined. Synchronization of LED start was carried out by means of the pulse generator AKIP-3413/1 (Prist, Russia), which received a control signal from the iXon3 897 camera (Andor, UK). All images were acquired at room temperature.

Analysis of trajectories of UCNP-PEI. The PL signal of nanoparticles with a size below the diffraction limit is well described by the normal distribution. Therefore, to determine the coordinates of the UCNP-PEI location in cells, we fitted the diffraction spots corresponding to UCNP-PEI by the two-dimensional Gaussian function (G) [22]:

where x₀, y₀ is the center of the diffraction spot, 2√2ln2·s_{x, y} is its full width at half maximum along the axis x and y, respectively and I₀ is the peak intensity.

To determine the transport speed and the type of motion of the marked objects, the mean square displacement (MSD) was calculated for each time point of the UCNP-PEI trajectory:

where t is the acquisition time, N is the total number of frames, and n is the frame number in sequence.

The plot MSD(nt) has characteristic features for each type of particle motion. The linear dependence of MSD on time with a slope of 4D, where D is diffusion coefficient, corresponds to an isotropic random motion. If there is an additional directional motion with a velocity ν, MSD(nt) is expressed as:

In the case when the random motion of a particle is bounded by the region L, the graph MSD(nt) is described by the following expression:

where L² is the confined area in which the diffusion is restricted.

Results and Discussion

According to the transmission electron microscopy data on the size distribution, NaYF₄:Yb,Tm UCNPs were monodisperse with an average nanocrystal size of 28.1±2.2 nm (Figure 2). This type of Tm³⁺-doped UCNPs is characterized by PL emission with several main peaks in the visible (455 nm, 514 nm, and 660 nm) and near-infrared (744 nm, 782 nm, and 802 nm) spectral regions when excited at 980 nm [21].

---

Figure 2. The transmission electron microscopy image of NaYF4:Yb,Tm nanoparticles On the insert, the size distribution of the nanoparticles

---

Incubation of UCNP-PEI with SK-BR-3 cells for 2 h was accompanied by receptor-independent endocytosis. The internalized nanoparticles were retained in the cell for a long time (at least 12 h). We visualized the behavior of particles in the cell after this time period. Figure 3 shows images of UCNP-PEI in SK-BR-3 cell extracted from a video file recorded in real-time at a rate of 5 frames per second. Taking into account the size of UCNPs, which is about ~30 nm, the most likely mechanism of the nanoparticles uptake is pinocytosis, leading to encapsulation of particles in vesicular organelles such as endosomes and their subsequent intracellular transport [23, 24]. The positive surface charge of the PEI-coated UCNPs, presumably, leads to the electrostatic interaction with negatively charged membrane phospholipids, such as phosphatidylserine and phosphatidylglycerol, and facilitates the first steps of the internalization.

---

Figure 3. Dynamics of the movement of UCNP-PEI-containing vesicles in a living SK-BR-3 cell: video file shots at time points 0, 9.6, 19.2, 28.8, 37.3 s (see video) Upper row: brightfield images; middle row: photoluminescence images of UCNP-PEI, recorded in the range of 420–842 nm; lower row: overlay images

---

The proposed scheme for recording luminescence causes a selective signal acquisition only from UCNP-PEI (Figure 3, middle row). When exciting UCNPs by a continuous wave laser diode at 975 nm, the autofluorescence signal is undetectable [20]. The blurred PL signal is caused by UCNPs located outside the focus at the time of exposure of the frame.

While visually analyzing the video files obtained, we noted that some particles undergo flick and directional movements. These movements usually last for several seconds and resume after short pauses, which is typical for the intracellular transport of nanoparticles encapsulated in vesicles [19, 25]. In particular, such behavior is demonstrated by the particle marked under number 1 in Figure 3. At the same time, some of the UCNP-PEI particles perform random spatial fluctuations with relatively small amplitudes (see the particle marked under number 2 in Figure 3).

To determine the location of particles 1 and 2, we have fitted the PL spots corresponding to these particles by the Gaussian function (1) on each frame of the time interval from 0 to 37.3 s. Figure 4 shows an example of determining the UCNP-PEI location.

---

Figure 4. Determination of the location of the vesicle containing the UCNP-PEI particles by fitting by the Gaussian function. The scale bar is 1 μm

---

The particle trajectories calculated in this way are shown in Figure 5. Particle 2 moves in a limited region with an area of less than 4 μm², while particle 1 is moved to more than 14 μm from its original location. The most likely mechanism for the particle 1 moving is active intracellular transport through microtubules or actin filaments [26].

---

Figure 5. Trajectories of the UCNP-PEI particles 1 and 2 in SK-BR-3 cell On the right and left inserts, the red color indicates the pathway traversed in 37.3 s by the particles 1 and 2, respectively. The scale bar on the inserts is 1 μm

---

Figure 6 shows the graphs of cumulative and relative displacements, depending on the time of observation of the UCNP-PEI particles indicated in Figure 3. The velocity of motion of the two types of particles noted above calculated on the basis of cumulative displacement differs several times (about 2.5). In order of magnitude, the particle velocities determined in our work (0.4±1.2 μm/s) agree with the data of [19], in which the tracking of UCNPs in a cell was performed.

---

Figure 6. Cumulative and relative displacements of particles 1 and 2, depending on the time of the observation

---

Another indicator shown in Figure 6 is the relative displacement (blue curves). In the case of direction motion, the displacement reached a significant value (Figure 6 (a)), whereas for random motion the displacement did not exceed 2 μm (Figure 6 (b)).

An important aspect that must be taken into account when carrying out long-term observations of living cells is the preservation of their viability throughout the experiment. In our study, the cells remained viable for 3 h. This opens possibilities for long-term monitoring of biological processes of interest in a cancer cell.

Conclusion

In this paper, we implemented a scheme for tracking of NaYF₄:Yb,Tm nanoparticles coated with polyethylenimine in living cells using wide-field microscopy. The observed nonspecific tracking of UCNP-PEI in human breast adenocarcinoma cells SK-BR-3 showed that UCNPs are an advanced agent for dynamic studies of intracellular processes.

The key advantages of the proposed tracking scheme with the use of UCNP-PEI and wide-field microscopy are the almost complete absence of autofluorescence, which ensures a high selectivity of PL registration, and fast rate of image acquisition, which makes it possible to study fast intracellular processes.

Further improvement of the photophysical properties of UCNPs and the use of surface conjugation schemes will expand the use of UCNPs as a PL marker for monitoring intracellular processes.

Acknowledgments. The work was financially supported by the Ministry of Education and Science of Russia (contract No.14.Z50.31.0022).

Conflicts of interest. The authors declare that there are no conflicts of interest regarding the publication of this paper.ф