Rhodamine-based fluorescent probe for dynamic STED imaging of mitochondria

Stimulated emission depletion (STED) microscopy holds tremendous potential and practical implications in the field of biomedicine. However, the weak anti-bleaching performance remains a major challenge limiting the application of STED fluorescent probes. Meanwhile, the main excitation wavelengths of most reported STED fluorescent probes were below 500 nm or above 600 nm, and few of them were between 500-600 nm. Herein, we developed a new tetraphenyl ethylene-functionalized rhodamine dye (TPERh) for mitochondrial dynamic cristae imaging that was rhodamine-based with an excitation wavelength of 560 nm. The TPERh probe exhibits excellent anti-bleaching properties and low saturating stimulated radiation power in mitochondrial STED super-resolution imaging. Given these outstanding properties, the TPERh probe was used to measure mitochondrial deformation, which has positive implications for the study of mitochondria-related diseases.


Introduction
Mitochondria, the energy factories for the smallest unit of life, are very important organelles in cells, with a smooth outer membrane and a highly curly inner membrane [1][2][3].The convoluted inner membrane is divided into many folds called cristae, which expand the surface area of the inner mitochondrial membrane and enhance its ability to produce ATP [4,5].Generally, mitochondria are approximately 1 µm wide and 4-10 µm long, while the distance between ridges in the inner mitochondrial membrane is ∼100 nm.Therefore, conventional optical microscopy cannot observe the cristae of mitochondria due to the optical diffraction limit.This is of great significance for the study of mitochondrial dynamics because recent studies have shown that dysregulation of mitochondrial dynamic can lead to mitochondrial dysfunction, affect cellular function, and be associated with many human diseases [6][7][8][9].
Various super-resolution microscopy techniques are currently used to study the submicroscopic structure of mitochondria in cells.Stochastic Optical Reconstruction Microscopy (STORM) was used to reveal the process of cristae junction formation, and the cristae junctions mainly exhibited a linear and punctate organization in COS7 and U 2 OS cells, respectively [10].Although STORM can achieve a resolution of ∼22 nm [11], it is still difficult to realize dynamic imaging of mitochondrial cristae in real time, as time is needed to reconstruct images.Structured illumination microscopy (SIM) enables dynamic imaging of mitochondrial cristae, but artifacts occur during algorithmic reconstruction, and the fidelity and quantitative properties are difficult to guarantee [12][13][14].In contrast, the high spatial resolution and temporal resolution of STED make it the most feasible option for investigating mitochondria [15][16][17][18][19].
STED has the ability to go beyond the Abbe diffraction limit and visualize structural features within mitochondria.However, obtaining live cells with STED for a long period of time remains a challenge [20,21].First, most fluorophores have limited photostability, and their fluorescence intensity deactivates within seconds.Fluorophore stability is a concern for STED microscopy since samples are corradiated with an exceptionally powerful depletion laser and a higher resolution may be achieved [22][23][24].Second, the existing dyes for labeling mitochondria have practical disadvantages, as they are not photostable enough to endure long-term STED imaging at high resolution, and the excitation wavelengths of these dyes are mostly above 600 nm, while dyes between 500-600 nm are still scarce (Supporting Information Table S1).
To address the challenges in mitochondria dynamic STED imaging, in this work, we developed a new dye called TPERh that is compatible with live cells.The primary advantage of this dye is that it has good anti-bleach properties and can be excited at 560 nm and 660 nm for depletion.With TPERh dyes, our mitochondrial cristae dynamic imaging reached 84 nm spatial resolution.In addition, the cristae structure can be observed over time, which indicates that TPERh can be used for dynamic STED imaging.Because it can make mitochondria have a dynamic imaging of super resolution and can be stimulated at 660 nm, we think that TPERh will be widely used in STED multicolor dynamic imaging.

Materials and equipment
The experimental drugs and reagents were purchased.The reagents and medicines were used without being dried and purified.The reactions were carried out in an oven-dried glassware with magnetic stirring.The fluorescent properties of TPERh were measured.Ab-sorption and fluorescence spectra were measured with a UV/Vis absorption spectrometer (GBC Cintra 2020, Australia) and a fluorescence spectrometer (Horiba iHR320, American), respectively.NMR spectra were recorded on a Bruker spectrometer at 1 H NMR (400 MHz) and 13 C NMR (100 MHz).Chemical shifts (d values) were reported in ppm downfield from internal Me 4 Si ( 1 H and 13 C NMR).High-resolution mass spectra (HRMS) were acquired on an Agilent 6510 Q-TOF LC/MS instrument (Agilent Technologies, Palo Alto, CA) equipped with an electrospray ionization (ESI) source.(Horiba iHR320, American) was also used to measure the absolute efficiency of TPERh and its stability in ethanol.

Cytotoxicity of TPREh and TPERh1
SOV3 cells were cultured in medium containing TPERh and TPERh1 dyes (0, 0.1, 1, 2, 4 µM).Cell viability was determined by MTT assay at 30 min, 60 min, 120 min and 240 min.The experiment was repeated 5 times, and the data values represent the mean ± standard deviation.

Cell culture and fluorescence imaging
Live SOV3 cells were stained in DMEM containing the fluorescent probe TPREh or TPERh1 (0.2 µL, 1 mM) and 10% fetal bovine serum for 15 min in a CO 2 incubator.Cells were then washed 3 times with PBS to remove free probes and incubated with fresh medium for 30 min.Then, the mitochondrial tracker (MitoTrack-er@Deep Red FM 644/665, 0.5 µL, concentration at 1 mM) or lysosome tracker (FluoLyso Deep Red 649/665, 0.5 µL, concentra-tion at 1 mM) was added and incubated for 20 min in the incubator.After the waste solution was removed, the cells were washed three times with PBS and incubated in an incubator for 30 min after add-ing fresh medium for confocal colocalization imaging.Additionally, the stained cells were incubated with high glucose medium (3∼5 times higher than normal DMEM) for 45 min.Then, mitochondrial fragmentation was observed on a Nikon confocal microscope (80×/1.40OIL, 1 µW, Ex: 560 nm, Em: 580-600 nm) for 30 min.Afterwards, the high-glucose medium was replaced with normal medium (DMEM), and the mitochondrial dynamics were further observed on a confocal microscope.

STED Imaging and colocalization
The Leica TCS SP8 STED system was used for STED imaging with excitation and depletion wavelengths of 560 and 660 nm, respectively.A HyD detector and STED WHITE objective (80×/1.40oil) were used.Unless otherwise stated, STED images were acquired with excitation at 560 nm (1 µW), a detector detection range of 570-600 nm, and a depletion laser at 660 nm (CW-STED, 23 mW).Images were processed using Lecia X software, and the full width at half maximum (FWHM) was used to define the resolution in STED super-resolution imaging.
After cells were stained, mitochondrial colocalization dye (Mito-Tracker@Deep Red FM 644/665, 0.1 µL, concentration at 1 mM) or lysosomal dye (FluoLyso Deep Red 647/668, 0.5 µL, concentration at 1 mM) was added and incubated for 20 min in the incubator.After the waste solution was removed, the cells were washed three times with PBS and incubated in an incubator for 30 min after adding fresh medium for confocal imaging.

Photophysical property of TPERh and TPERh1
The photophysical properties of the rhodamine-based fluorescent probe TPERh were initially investigated in different solvents.The TPERh probe exhibited the maximum molar absorption coefficient in the range of 555-565 nm (Fig. 2(A)) and strong fluorescence emission ranging from 575 nm to 595 nm in common organic solvents (Fig. 2(B)), which indicates that the probe is a typical xanthene dye.While TPERh1 has two absorption peaks, one in the range of 509 nm to 517 nm, and another in the range of 553 nm to 549 nm.Excited with a 515 nm laser, the emission wavelength of the dye ranges from 583 nm to 593 nm (Figure S7).In comparison, due to the hydrophobic tetraphene, the probe generated obvious H aggregation in aqueous solution, which resulted in an extremely broad absorbance (ε = 2.82 × 10 4 L mol −1 cm −1 ) and weak fluorescence at approximately 580 nm in TPERh.Meanwhile, the high absolute fluorescence quantum efficiency (ϕf = 42.3%) in DMSO ensures that significant fluorescence signals could be observed during live cell imaging (Figure S8).Additionally, the lifetime of TPERh and TPERh1 in different organic solvents was measured (Table 1, Table S2, Figure S9, Figure S10), and we can see that the lifetime of TPERh is shorter in water and DMSO than in any other solvent.We attribute the variation in lifetime to the polarity of the solvent and the solubility of TPERh and TPERh1 in different solvents.The fluorescence lifetime of TPERh is longer than that of TPERh1, which may be related to the symmetry of the material structure.

Fluorescence and STED imaging in living cells
According to the optical physical properties, fluorescence imaging experiments were carried out to investigate the intra-cellular staining capability of the probe TPERh and TPERh1.Before live cell imaging, the cytotoxicity of TPERh and TPERh1was first assessed.The TPERh and TPERh1 probe exhibited low toxicity in SOV3 cells at a concentration of 0.1 µM after 4 h of incubation (Figure S11).However, probe cytotoxicity may also occur with increasing concentration and incubation time.we suppose the hydrophobic moieties in the dye result in increased toxicity.In addition, the concentration of STED dynamic imaging was 0.1 µM, at which cell viability was sufficient for STED imaging (Fig. 3(A)).Compared with TPERh1, the cell viability is higher, indicating less toxic property.From the structure of TPERh and TPERh1, we can see that the TPERh has more hydrophobic moieties, the hydrophobic moieties in the dye may result in increased toxicity.At the same time, we also carried out glucose simulation experiments to verify whether the toxicity of TPERh was sufficient to cause mitochondrial fragmentation (Figure S12).As you could see, mitochondria remained tubular until 20 minutes after adding high concentration of glucose, and then from 30 to 40 minutes, shorter and smaller mitochondria begin to be present in cells, indicating mitochondrial fragmentation.Even if glucose was removed for 20 minutes, mitochondrial fragments also remained.Finally, 40 minutes after glucose removal, mitochondria returned to their tubular shape.This phenomenon was consistent with Tianzheng Yu et al [26,27].Through this experiment, we believed that TPERh was less toxic at low concentration.Because if TPERh was toxic enough to cause mitochondria fragmentation, then the mitochondria should not return to their normal tubular shape after glucose was removed.Bright fluorescence was clearly observed in SOV3 cells that were incubated with TPERh (0.5 µM, 1 µM and 2 µM) for 20 min (Figure S13A-C), and the background fluorescence intensity in the cytoplasm was quite low; meanwhile, the probe displayed a concentration-dependent fluorescence intensity that was positively related to the concentration (Figure S13D).These results demonstrate that TPERh exhibits enough membrane penetration ability and mitochondrial selectivity.To determine the lowest concentration of TPERh for STED imaging, we conducted the experiment under different concentrations for STED imaging (Figs. 3).The mitochondrial cristae could be seen under different concentrations.Considering cytotoxicity of TPREh, we preferred the lowest concentration for further STED imaging.The same experiments were carried out for TPERh1.Unfortunately, no depleted effect was observed for STED imaging (Figure S14).To test the mitochondrial targeting ability of the TPERh and TPERh1 probe, colocalization experiments were performed by containing SOV3 cells with TPERh and organelle-targetable tracker to confirm the fluorogenic of the probe in mitochondria.From Fig. 4, the fluorescence signals of TPERh (green channel) were perfectly colocalized with that of Mito-Tracker (red channel) with a Pearson's correlation coefficient of 0.920, while they merged poorly with those of Lyso-Tracker (red channel, Pearson's correlation coefficient for the colocalization of lysosomes was 0.261).The high Pearson's correlation coefficient of TPERh colocalized with mitochondria clearly indicated that the TPERh probe could specifically target mitochondria in living cells.Obvious, there was an overlap between Lysosome and mitochondria, we supposed that TPERh could incorporate into both membrane as a hydrophobic probe.To testify whether tetra-phenyl ethylene fragment contributes to the targeting ability, we carried colocalization experiment for TPERh1.The Pearson's correlation coefficient was 92% (Figure S15), which indicated the tetra-phenyl ethylene fragment may contribute the probe to targeting at mitochondria.To demonstrate the superior utility of the TPREh probe, the anti-photobleaching property was also performed.After scanning 200 frames (approximately 160 s) by STED (depletion laser power is 23 mW) and confocal mode, less than 10% attenuation of the fluorescence signal in SOV3 cells stained with TPERh was observed, as shown in Fig. 5(A) and Figure S16.Meanwhile, the saturated stimulated emission power of the TPERh probe was obtained as Psat = 2.44 ± 0.04 mW by measuring the lateral resolution with different powers of STED light (660 nm, Fig. 5(B), Figure S17, Figure S18).The Isat for TPERh is 0.813 MW/cm 2 at 660 nm based on the doughnut area we measured earlier (∼ 3 × 10 −9 cm 2 ) [28].In addition, the fluorescence intensity of STED and confocal images collected at different presented slight intensity differences, which indicated that the TPERh probe can be used for dynamic mitochondria imaging (Fig. 5(C, D)).The Isat of TPERh is smaller than that of MitoEsq-635 [15]; Meanwhile, TPERh has better anti-bleaching properties because the light intensity of TPREh only decreased by 0.1 after scanning 100 frames, while that of MitoESP-635 decreased by 0.2 (Figure S19, Table S1), which indicates that TPERh has stronger anti-bleaching properties than MitoEsq635.Similarly, compared with commercial dyes MitoTracker@Deep Red FM, the STED and confocal images collected at different sights change quickly, indicating that MitoTracker have inferior anti-bleaching property (Figure S20).Also, the P sat of MitoTracker@Deep Red FM was also measured (18.81 mW), ∼ 7.7 times higher than that of TPERh (Figure S21).Besides, PK Mito and TPERh have almost the same quantum efficiencies in alcohol solutions, and both can be used for mitochondrial dynamic imaging with a resolution around 85 nm [16].The relevant information of TPERh, MitoEsq-635, MitoTracker@Deep Red FM and PK Mito were summarized Table 2 below.Meanwhile, TPERh has a smaller I sat than most dyes (Table S1).Therefore, the high brightness, permeability, photostability and mitochondrial-targeting ability of the developed probe make it ideally suited for live cell STED imaging.
As expected, the TPERh probe can vividly achieve super-resolution of mitochondrial dynamics.The raw data of confocal and STED images clearly showed the mitochondrial cristae (Fig. 6(A,  Compared with confocal images of the inner mitochondrial membrane with a resolution of 330 nm, the STED imaging results revealed the refined structure of the cristae inside the mitochondria with a resolution of 84 nm.Also, another example that resolution achieved 84 nm was given in Figure S22.Taking advantage of the high resolution of the TPERh probe in fluorescence imaging of mitochondria, the dynamic processes and subtle morphological changes within mitochondria can be monitored by STED.In Fig. 6(F), we achieved dynamic mitochondrial cristae imaging.The yellow arrow shows that the crista could be seen at 10th frame, and disappeared in 25th frame.While the yellow arrow shows crista with high intensity, and the crista intensity start to fall in 15th frame, and the crista finally disappeared in 40th frame.These phenomena indicate that TPERh can be used for studying dynamic changes in cristae.

Conclusion
In this paper, a new organic dye for mitochondrial STED imaging in live cells was found.
Compared with commercial mitochondrial tracers, the dye had a good mitochondrial localization effect.Using this probe for STED super-resolution imaging, it has low saturation depletion power and good resistance to photobleaching and can achieve a spatial resolution of up to 84 nm.Additionally, this probe is very suitable for STED super-resolution imaging to detect the dynamic changes in cristae and provides a useful tool for the study of mitochondrial structure and function.
In addition, because it can achieve dynamic imaging of mitochondria with superior resolution and can be depleted at 660 nm, we think that TPERh will be widely used in STED multicolor dynamic imaging.Thus, it is a powerful tool and has great significance and broad application prospects in the study of mitochondria-related diseases.

Fig. 4 .
Fig. 4. Mitochondrial colocalization.(A, B) Mitochondrial colocalization using MI tracker and Lyso-tracker.The last plot on right was Pearson's correlation coefficient calculated by merged photos.A higher Pearson coefficient indicates more overlapping area.

Fig. 5 .
Fig. 5. STED anti-bleaching test of TPERh.(A) Fluorescence intensity changes of 200 frames of confocal and STED images.(B) Resolution of STED images obtained at increased depletion power, and the data was fitted with equation FWHM = λ/ (︃ 2NA (︂ 1 + P P sat

Fig. 6 .
Fig. 6.Confocal and STED images of SOV3 labeled with TPERh.(A) Confocal images; (B) STED image of mitochondria when the depletion power is 50 mW.(C, D) Deconvoluted Image of Figure A and Figure B. (E) FWHM of yellow line in figure C and D. (F) mitochondrial dynamics at different frames.The red arrow and yellow arrow showed mitochondria fission.