Spatial and temporal information about intracellular objects and their dynamics within a living cell are essential for dynamic analysis of such objects in cell biology. A specific intracellular object can be discriminated by photoactivatable fluorescent proteins that exhibit pronounced light-induced spectral changes. Here, we report on selective labeling and tracking of a single organelle by using two-photon conversion of a photoconvertible fluorescent protein with near-infrared femtosecond laser pulses. We performed selective labeling of a single mitochondrion in a living tobacco BY-2 cell using two-photon photoconversion of Kaede. Using this technique, we demonstrated that, in plants, the directed movement of individual mitochondria along the cytoskeletons was mediated by actin filaments, whereas microtubules were not required for the movement of mitochondria. This single-organelle labeling technique enabled us to track the dynamics of a single organelle, revealing the mechanisms involved in organelle dynamics. The technique has potential application in direct tracking of selective cellular and intracellular structures.
© 2007 Optical Society of America
Intracellular objects, including organelles and protein complexes, dynamically move within living cells and are involved in cell division, metabolism, and signal transduction. In vivo examination of their location and velocity at specific spatial and temporal intervals has the potential to allow functional analysis of intracellular events. Positioning and movement of intracellular organelles are essential for cellular homeostasis and adaptation to external stress. Individual organelles have different motilities and functions according to their intracellular environments and locations. Direct tracking of the movement and velocity of organelles within living cells can be accomplished with the introduction of a fluorescent marker that can be activated locally by light. Photoactivation and photoconversion enables selective conversion or activation of fluorescence signals after optical illumination, and are recognized as powerful tools for studying the dynamic processes of fluorescently labeled individual cells, organelles, and proteins [1,2].
Photoactivation refers to the transformation of an essentially non-fluorescent molecule into a highly fluorescent product by the absorption of light [2–8]. Photoactivatable fluorescent protein, for example, photoactivatable green fluorescent protein (PAGFP), is suitable for protein tracking due to its monomeric property [2–4]. PAGFP is initially non-fluorescent, but it emits green light after irradiation with ultraviolet (UV) light. Therefore, additional staining is required for distinguishing photoactivated regions from unphotoactivated regions.
Photoconvertible fluorescent proteins capable of pronounced light-induced spectral changes have been developed [9–14]. For example, Kaede, a cloned fluorescent protein extracted from the open brain coral Trachyphyllia geoffroyi, has an emission color that can be irreversibly changed from green to red by exposure to UV light (specifically, 350 to 400 nm in wavelength) [11–15]. These color-changing properties of photoconvertible Kaede offer several advantages. First, a larger separation of the green and red emission wavelengths enables simultaneous observation of the interaction between labeled organelles and unlabeled ones. Second, the red fluorescence emitted from the photoconverted Kaede is bright and stable without requiring rigorous anaerobic conditions. Third, image acquisition of the green and red fluorescence can be performed with excitation using blue and green light, respectively, neither of which induces photoconversion of Kaede.
Most photoactivation and photoconversion processes via one-photon excitation require light in the UV region. One-photon excitation leads to collateral photoconversion outside the focal volume, which limits the spatial resolution . In contrast, although there have been some reports describing the use of near-infrared femtosecond laser pulses to attempt two-photon excitation of photoactivatable fluorescent proteins, problems such as low contrast and low spatial resolution were encountered [5,13,15].
In this study, we developed an imaging method that combines the photoconvertible protein Kaede and two-photon excitation. We demonstrated selective labeling of a single mitochondrion by using two-photon conversion of Kaede, and tracking the dynamics a single mitochondrion. The ability to label selective organelles is a potentially promising method for analysis of intracellular structures by tracking the dynamics of a single organelle in a living cell, revealing the mechanisms involved in organelle dynamics.
2.1 Optical setup
Figure 1 shows a schematic diagram of the setup used for tracking and labeling a single organelle within a living cell. We used a confocal microscope to simultaneously image unphotoconverted and photoconverted fluorescence. The laser scanning microscope was adapted from an Olympus FV300 scanning unit combined with an Olympus IX71 inverted microscope. The CW beams from a He-Ne laser (wavelength 543 nm, Melles Griot, 05-LGR-171) and an Ar+ laser (wavelength 488 nm, Melles Griot, IMA101-010-BOS) were reflected by dichroic mirrors DM1 (Olympus Corp., FV3-RDM650-2) and DM2 (Olympus Corp., FVX-DM BG(488/533)) and then focused into the BY-2 cell through an oil-immersion objective lens (OB; Olympus Corp., PlanApo60×O2, numerical aperture (NA) 1.4). Green fluorescence was excited by the Ar+ laser, and the back-propagated fluorescence was collected using the same objective lens and detected with a photomultiplier tube (PMT1; Hamamatsu Photonics, R928P), after passing through a bandpass filter BP1 (transmission wavelength: 510 to 540 nm, Olympus Corp., FVX-BA510-540). Red fluorescence after photoconversion of Kaede was excited by the He-Ne laser (wavelength 543 nm), and the back-propagated fluorescence was detected with a photomultiplier tube (PMT2), after passing through a dichroic mirror DM3 (Olympus Corp., FV-SDM570) and a bandpass filter BP2 (transmission wavelength: 560 to 600 nm, Olympus Corp., FV3-BA560-600). Green and red fluorescence images were alternately captured to avoid cross-talk between the two PMTs. Two-dimensional confocal cross-sectional images were obtained by scanning the focused laser beams in the XY plane with a pair of high-speed galvanometer mirrors (GM; Cambridge, 6210) inside the laser-scanning microscope. Scanning in the depth direction (Z direction) was achieved by moving the objective lens with a stepping motor to obtain three-dimensional (3D) confocal fluorescence images. To obtain the stacked 3D images, we obtained 13 confocal cross-sectional images by translating the objective lens by 2.4 μm in the depth (Z) direction in steps of 0.5 μm. Transmission images were obtained using the Ar+ laser with lower emission intensity. Imaging and marking by laser irradiation were performed with Fluoview software.
2.2 Photoconversion of Kaede
Photoconversion of Kaede was performed using near-infrared femtosecond laser pulses from a mode-locked Ti:sapphire laser oscillator with a wavelength of 750 nm and a repetition rate of 76 MHz (Coherent, Mira). The laser pulses were passed through a Faraday isolator to block reflections from the optical components. The laser pulses were then passed through a series of SF10 prisms (P1, P2) to compensate for the dispersion of the optical components in the light path and the microscope. The laser pulses were then directed onto the pair of GMs via DM1 and DM2 and were focused into the BY-2 cell through the oil-immersion objective (OB). During femtosecond irradiation, DM2 is replaced by a dichroic mirror (Olympus, FV3-DDM488/650). The laser energy was controlled with neutral density filters (ND). The pulse duration measured at the specimen, after passing through the microscope, was 135 fs. The number of pulses supplied was selected by an electro-magnetic shutter (Sigma Koki, Σ-65L).
2.3 Materials (cell preparation).
We constructed the expression vector, which includes the N-terminal presequence of the Arabidopsis mitochondrial ATPase subunit and the Kaede coding region (Amalgaam, Tokyo) under the control of the CaMV35S promoter.
Kaede-expressing tobacco BY-2 cells were prepared by a previously described method , and the cells were maintained as previously described by Nagata et al. . The cells were cultured in modified Linsmaier and Skoog medium in a rotary shaker at 25 °C in the dark. Two-day-old BY-2 cells after subculturing were used for our analyses. Inhibition of actin filaments and microtubules was performed by incubation with 1 μM bistheonellide A and 3 μM propyzamide for 30 min and 1hr, respectively. All experiments were carried out at 20°C.
2.4 Intracellular photoconversion experiments
We selected a cell and recorded fluorescence of the mitochondria with one-photon excitation. Femtosecond laser pulses with a mean power of 4 to 8 mW were tightly focused at a target mitochondrion in the cell. After photoconversion of the mitochondrion, we acquired time-lapse images of the green and red fluorescence using a one-photon fluorescence microscope (see Fig. 1). Note that the wavelength used for one-photon excitation does not induce any photoconversion.
3.1 Single Organelle Labeling in a Fixed Cell
First, we describe photoconversion of a single mitochondrion by two-photon excitation. We used fixed BY-2 cells whose mitochondria were labeled with Kaede. Near-infrared femtosecond laser pulses from a mode-locked Ti:sapphire laser were focused into a target mitochondrion to alter the fluorescence from green to red. (See Fig. 1)
Figure 2 shows stacked 3D confocal fluorescence images and transmitted images before and after femtosecond laser irradiation. Green fluorescence images were obtained through a bandpass filter (transmission wavelength: 510 to 540 nm, left column of Fig. 2), and red fluorescence images were obtained through another bandpass filter (transmission wavelength: 560 to 600 nm, centre column of Fig. 2). To identify the cellular boundaries, transmission images were simultaneously recorded for each frame and were overlaid with the activated mitochondrion (right column of Fig. 2). Laser pulses with an energy of 0.053 nJ/pulse (4 mW) were focused at the mitochondrion indicated by the arrow, and the shutter was opened for an exposure time of 1 s, which corresponded to 76×106 pulses. After photoconversion by the femtosecond laser irradiation, the red fluorescent signal increased, resulting in a seven-fold increase of red fluorescence before and after photoconversion. The contrast between the unphotoconverted green fluorescence and the photoconverted red fluorescence was 2:1. The results demonstrate that some of the fluorescence from a single mitochondrion was converted from green to red. The spatial resolution in cross-section and the depth resolution of photoconversion were approximately 1 μm. We estimate that a volume of the mitochondrion of approximately 1 femtoliter (= 1.0 μm×π×(0.5 μm)2) was photoconverted, indicating that a volume of less than one femtoliter of a single organelle, or part of some organelles, can be photoconverted around the focal point. The results demonstrate that spatially selective labeling of a single organelle can be performed in 3D space.
3.2 Single-Organelle Tracking in a Living BY-2 cell
To examine the movement of a single mitochondrion, we demonstrated, as an example, tracking of the movement of a single mitochondrion among several hundreds of mitochondria in a living BY-2 cell. Figure 3 shows a time-lapse series of confocal images of mitochondria labeled with Kaede in living BY-2 cells, obtained at 30-min intervals. Laser pulses with an energy of 0.11 nJ (8 mW) were focused at the mitochondrion indicated by the arrow, and the shutter was opened for an exposure time of 2 s. The irreversible photoconversion of Kaede, giving stable red fluorescence, allowed tracking of the movement of the target mitochondrion for as long as 5 hours. During observation, no photobleaching was observed. It should be noted that after femtosecond laser irradiation, both the red and green signals, as well as their ratio, were stable, suggesting that this technique allows an organelle to be tracked for a long time without loss of image contrast. This indicates that labeling and tracking of an organelle using a femtosecond laser can be performed without compromising the viability.
From time-lapse stacked 3D confocal images taken at 30 s intervals after labeling, we observed the dynamics of an individual mitochondrion. Figures 4(a) and (b) show the position and trajectory of the labeled mitochondrion in 3D space and projections of 3D tracks on the XY, XZ, and YZ planes, respectively. The velocity of the photoconverted mitochondrion was calculated to be 0.18 μm s-1. The results demonstrated the ability to track the dynamics of a single mitochondrion and to reveal detailed spatial information in a living cell, such as the position and velocity. Site-specific organelle labeling enabled us to track the dynamics of a single organelle at different sites in a living cell. The activities of individual mitochondria vary depending on their location in the cell. Mitochondria that move along cytoskeleton strands (average velocity, 0.50 μm s-1, n=10) move faster than those around the nucleus (average velocity, 0.22 μm s-1) and those around the cortical region (average velocity, 0.10 μm s-1). Our method can therefore analyze the dynamics of a single organelle in a specific region within a cell.
3.3 Interaction between Mitochondria and Cytoskeletons and Mitochondrial Transport
We studied the interaction between mitochondria and cytoskeletons and mitochondrial transport. In higher plants, the movement of an organelle predominantly depends on actin filaments, whereas, in yeast and animals, it depends on microtubules . In order to analyze the relationship between cytoskeletons and the dynamics of mitochondria, we selectively labeled and tracked mitochondria that move along cytoskeleton strands. Then, we quantified the velocities of the labeled mitochondria, whose cells were treated with the actin-polymerization inhibitor bistheonellide A or the tubulin-polymerization inhibitor propyzamide . The average velocities of the individual mitochondria were calculated from tracing the positional difference of 100 successive positions at intervals of 1.1 s (n=10). The tracking data were used to calculate the average velocities of the individual mitochondria, as well as the instantaneous velocities. Figures 5(a) and (b) show typical examples for the tracking positions and instantaneous velocities of the mitochondria, respectively. The variations in instantaneous velocity in untreated control cells and the cells treated with propyzamide were reflected in changes between rapid moving and wiggling events. Our results agree with published work showing that mitochondria repeatedly stop and go . The high velocity probably reflects movement along cytoplasmic strands, and the extremely low velocity suggests that mitochondria occasionally pause at one point. The average velocities of the individual mitochondria treated with bistheonellide A and propyzamide, and that of the control cells were 0.08±0.04 μm s-1, 0.44±0.22 μm s-1, and 0.51±0.24 μm s-1, respectively (Fig. 6). The velocity of the individual mitochondria treated with bistheonellide A drastically reduced, whereas it was not significantly affected by propyzamide. This suggests that, in plant cells, the directional movement of mitochondria strongly depends upon actin filaments, whereas microtubules are not required for the movement of mitochondria, indicating that the movement is probably regulated by myosin motors. This finding is supported by the analysis of individual mitochondria in plant leaves .
Photoconversion using two-photon excitation allows selective photoconversion at any location in the cell within a photoconversion volume of a few femtoliters down to less than one femtoliter. Previous studies on photoactivation or photoconversion using near-infrared femtosecond laser pulses suffered from the problem of low spatial resolution. It should be noted that our experiments demonstrated selective labeling and tracking of a single organelle by using two-photon conversion of a photoconvertible fluorescent protein with near-infrared femtosecond laser pulses. In two-photon conversion of Kaede, it is important to use the optimal laser parameters. Optimal conditions to perform successful photoconversion of a mitochondrion labeled with Kaede were a laser energy of 0.08 to 0.12 nJ and an exposure time of 1 to 2 s. Kaede forms tetramer in vivo, and the oligomerization properties of Kaede prevent rapid diffusion and exchange within living cells and contributes to stable retention of the photoconvertible protein in mitochondria. Therefore, monitoring the photoconverted fluorescence of Kaede enabled us to analyze the dynamics of a single living mitochondrion over a long period of time (five hours). In contrast, the properties of monomeric photoconvertible proteins including PA-GFP and Dendra2, which have been used for labeling and tracking of intracellular objects such as proteins, make them suitable for tracking rapid movements .
The results presented here demonstrate that labeling by two-photon conversion is a powerful tool for dynamic analysis of a single organelle. Introduction of a photoconvertible fluorescent label enabled precise photolabeling in a sub-femtoliter focal volume and tracking of the organelle, or an intracellular protein complex of interest, and gave information on its velocity and position. This type of analysis will give new insights into the dynamic mechanisms involved in various intracellular phenomena, including vesicular transport, axonal transport, redistribution of intracellular structures, and chromatin dynamics. Moreover, monomeric photoconversion by two-photon excitation will allow subcellular region-specific cascades in signal transduction to be monitored in real time. Selective labeling of an intracellular object using two-photon conversion will open the door to various new biological and medical applications. This will allow selective analysis of organelles in abnormal cells, including degenerative diseased cells or cancer cells. Moreover, intracellular imaging of a single organelle in combination with optical manipulation, such as laser nanosurgery [21–26] and trapping , will help in understanding organelle-organelle communication.
We have demonstrated tracking the movement of a mitochondrion by monitoring the photoconverted fluorescence of Kaede. By photoconversion of the fluorescence properties at one specific point using tightly focused femtosecond laser pulses, the movement of a specific organelle in a living cell could be tracked.
The authors would like to thank T. Higashi and S. Kataoka from Osaka University for useful discussions. This work was supported by Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Culture, Sports, Science and Technology of Japan, awarded to one of the authors (S.M.; grant no. 18687005).
References and links
3. J. N. Post, K. A. Lidke, B. Rieger, and D. J. Arndt-Jovin, “One- and two-photon photoactivation of a paGFP-fusion protein in live Drosophila embryos,” FEBS Lett. 579, 325–330 (2005). [CrossRef] [PubMed]
4. M. Karbowski, D. Arnoult, H. Chen, D. C. Chan, C. L. Smith, and R. J. Youle, “Quantitation of mitochondrial dynamics by photolabeling of individual organelles shows that mitochondrial fusion is blocked during the Bax activation phase of apoptosis,” J. Cell Biol. 164, 493–499 (2004). [CrossRef] [PubMed]
5. M. Schneider, S. Barozzi, I. Testa, M. Faretta, and A. Diaspro, “Two-photon activation and excitation properties of PA-GFP in the 720–920 nm region,” Biophys J. 89, 1346–1352 (2005). [CrossRef] [PubMed]
6. D. M. Chudakov, V. V. Belousov, A. G. Zaraisky, V. V. Novoselov, D. B. Staroverov, D. B. Zorov, S. Lukyanov, and K. A. Lukyanov, “Kindling fluorescent proteins for precise in vivo photolabeling,” Nature Biotechnol. 21, 191–194 (2003). [CrossRef]
7. D. M. Chudakov, V. V. Verkhusha, D. B. Staroverov, E. A. Souslova, S. Lukyanov, and K. A. Lukyanov, “Photoswitchable cyan fluorescent protein for protein tracking,” Nature Biotechnol. 22, 1435–1439 (2004). [CrossRef]
8. Y. Chen, P. J. Macdonald, J. P. Skinner, G. H. Patterson, and J. D. Müller, “Probing nucleocytoplasmic transport by two-photon activation of PA-GFP,” Microscopic Res. Tech. 69, 220–226 (2006). [CrossRef]
9. J. Wiedenmann, S. Ivanchenko, F. Oswald, F. Schmitt, C. Rocker, A. Salih, K. D. Spindler, and G. U. Nienhaus, “EosFP, a fluorescent marker protein with UV-inducible green-to-red fluorescence conversion,” Proc Natl. Acad. Sci. USA 101, 15905–15910 (2004). [CrossRef] [PubMed]
10. N. G. Gurskaya, V. V. Verkhusha, A. S. Shcheglov, D. B. Staroverov, T. V. Chepurnykh, A. F. Fradkov, S. Lukyanov, and K. A Lukyanov, “Engineering of a monomeric green-to-red photoactivatable fluorescent protein induced by blue light,” Nature Biotechnol. 24, 461–465 (2006). [CrossRef]
11. R. Ando, H. Hama, M. Yamamoto-Hino, H. Mizuno, and A. Miyawaki, “An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein,” Proc. Natl. Acad. Sci. USA , 99, 12651–12656 (2002). [CrossRef] [PubMed]
12. H. Mizuno, T. Kumar Mai, K. I. Tong, R. Ando, T. Furuta, M. Ikura, and A. Miyawaki, “Photo-induced peptide cleavage in the green-to-red conversion of a fluorescent protein,” Mol. Cell 12, 795–1058 (2003). [CrossRef]
13. H. Tsutsui, S. Karasawa, H. Shimizu, N. Nukina, and A. Miyawaki, “Semi-rational engineering of a coral fluorescent protein into an efficient highlighter,” EMBO Rep. 6, 233–238 (2005). [CrossRef] [PubMed]
14. S. Arimura, J. Yamamoto, G. P. Ada, M. Nakazono, and N. Tsutsumi, “Frequent fusion and fission of plant mitochondria with unequal nucleoid distribution,” Proc. Natl. Acad. Sci. USA 101, 7805–7808 (2004). [CrossRef] [PubMed]
15. S. Ivanchenko, C. Röcker, F. Oswald, J. Wiedenmann, and U. Nienhaus, “Targeted green-red photoconversion of EosFP, a fluorescent marker protein,” J. Biol. Phys. 31, 249–259 (2005). [CrossRef]
16. T. Nagata, Y. Nemoto, and S. Hasezawa, “Tobacco BY-2 cell line as the ‘HeLa’ cell in the cell biology of higher plants,” Int. Rev. Cyt. 132, 1–30 (1992). [CrossRef]
18. A. Yoneda, M. Akatsuka, H. Hoshino, F. Kumagai, and S. Hasezawa, “Decision of spindle poles and division plane by double preprophase bands in a BY-2 cell line expressing GFP-tubulin,” Plant Cell Physiol. 46, 531–538 (2005). [CrossRef] [PubMed]
19. A. Nebenführ, L. A. Gallagher, T. G. Dunahay, J. A. Frohlick, A. M. Mazurkiewicz, J. B. Meehl, and L. A. Staehelin, “Stop-and-go movements of plant Golgi stacks are mediated by the acto-myosin system,” Plant Physiol. 121, 1127–1141 (1999). [CrossRef] [PubMed]
20. K. Van Gastel, R. H. Köhler, and J.-P. Vervelen, “Plant mitochondria move on F-actin, but their positioning in the cortical cytoplasm depends on both F-actin and microtubules” J. Exp. Bot. 53, 659–667 (2002). [CrossRef]
22. W. Watanabe, N. Arakawa, S. Matsunaga, T. Higashi, K. Fukui, K. Isobe, and K. Itoh, “Femtosecond laser disruption of subcellular organelles in a living cell,” Opt. Express 12, 4203–4213 (2004), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-18-4203. [CrossRef] [PubMed]
23. A. Heisterkamp, I. Z. Maxwell, E. Mazur, J. M. Underwood, J. A. Nickerson, S. Kumar, and D. E. Ingber, “Pulse energy dependence of subcellular dissection by femtosecond laser pulses,” Opt. Express 13, 3690–3696 (2005), http://www.opticsexpress.org/abstract.cfm?URIOPEX-13-10-3690. [CrossRef] [PubMed]
25. T. Shimada, W. Watanabe, S. Matsunaga, T. Higashi, H. Ishii, K. Fukui, K. Isobe, and K. Itoh, “Intracellular disruption of mitochondria in living HeLa cells with a 76-MHz femtosecond laser oscillator,” Opt. Express 13, 9869–9880 (2005), http://www.opticsinfobase.org/abstract.cfm?URI=oe-13-24-9869. [CrossRef] [PubMed]
26. A. Vogel, J. Noack, G. Hüttman, and G. Paltauf, “Mechanisms of femtosecond laser nanosurgery of cells and tissues,” Appl. Phys. B 81, 1015–1047 (2005). [CrossRef]
27. K. König, “Laser tweezers and multiphoton microscopes in life sciences,” Histochem. Cell Biol. 114, 79–92 (2000). [PubMed]