- Email: [email protected]
Dynamic behavior of individual cells in developing organotypic brain slices revealed by the photoconvertable protein Kaede
Experimental Neurology 200 (2006) 430 – 437 www.elsevier.com/locate/yexnr
Dynamic behavior of individual cells in developing organotypic brain slices revealed by the photoconvertable protein Kaede T. Mutoh a,⁎, T. Miyata a , S. Kashiwagi a , A. Miyawaki b , M. Ogawa a a b
Laboratory for Cell Culture Development, Advanced Technology Development Center, Brain Science Institute, Riken Saitama, Japan Laboratory for Cell Function and Dynamics, Advanced Technology Development Center, Brain Science Institute, Riken Saitama, Japan Received 7 July 2005; revised 26 February 2006; accepted 1 March 2006 Available online 6 June 2006
Abstract In recent years, advances in optical imaging methods have facilitated the visualization of events in the developing cortex. In particular, the introduction of DNA encoding fluorescent protein into cells of the embryonic brain allows the visualization of progenitor cells; slice preparations of the cortex then allow the monitoring of the behavior of transfected cells in the context of the living cerebral wall by time-lapse microscopy. Such approaches have provided substantial information about the patterns of neuronal migration. However, as these techniques label large numbers of cells in the ventricular zone (VZ), it is difficult to follow individual cell shape changes or cell behaviors within the VZ, where neuron production and initial migration take place. Here, we report a unique method using the photoconvertable fluorescent protein Kaede, which emits green fluorescence and shifts to emitting red fluorescence upon radiation with UV. Using this method, we were able to follow the behavior of a particular pair of daughter cells among neighboring Kaede-positive cells in the SVZ of mouse brain slices. The spindle shape progenitor divided into two multipolar-shaped daughter cells. The cell–cell borders of daughter cells were clearly visualized, and easily describe the position and distance between two or more cells. The photoconvertable property of Kaede offers a powerful cell marking tool to identify the precise morphology and migratory behaviors of individual cells within living cortical slices. © 2006 Elsevier Inc. All rights reserved. Keywords: Fluorescent protein; Kaede; Photoconversion; Organotypic slice culture; Electroporation; 3D-time-lapse imaging; Neocortical histogenesis; Neuronal progenitor; Cell migration
Introduction The temporally and spatially restricted events of neuron production and migration during early cortical development are important steps in the establishment of the laminated structure of the cerebral cortex. Abnormalities in these events result in malformations of the cortex and in turn lead to brain function deficits. The mechanisms underlying neuronal migration and differentiation may consist of a combination of intrinsic cues and environmental factors. Thus, to gain insights into the potential mechanisms that drive normal morphogenetic events, it is crucial to visualize the behavior of individual cells and their progeny in the brain. Abbreviations: E, embryonic day; IZ, intermediate zone; MZ, marginal zone; SVZ, subventricular zone; VZ, ventricular zone. ⁎ Corresponding author. Fax: +81 48 467 5496. E-mail address: [email protected] (T. Mutoh). 0014-4886/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2006.03.022
Fluorescent vital dyes (e.g., DiI) have a long history as cellular markers, in part because they are very easy to apply to cells; after co-incubation, cells will readily take up such dyes. However, it is difficult to apply these dyes to a single cell, and they are therefore most often used to track the fate of particular cell groups. In recent years, fluorescent proteins have also been used as cellular markers in the field of neuroscience. Fluorescent proteins are easier to handle, less toxic, and not as subject to decreased fluorescence as are chemical dyes as a result of internalization, cell division, or cell growth. Fluorescent protein remains in a cell and its progeny as long as the vector remains present, and the appropriate transcriptional elements drive the expression of fluorescent protein. Thus far, a large number of color and functional variations of fluorescent proteins have been developed (for a review, see Matz et al., 2002; Miyawaki, 2004). In order to introduce a gene into living tissue, electroporation, replication-incompetent viral vectors, lipofection, and gene gun approaches are available. In and exo-
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
utero electroporation is one powerful tool that has been used to introduce such marker proteins into the fetal brain (Tabata and Nakajima, 2001). It is easier and less time-consuming to apply this method than it is to construct viral vectors and generate virions; moreover, electroporation is more efficient at introducing genes than are lipofection and gene gun techniques. However, when fluorescent protein is introduced into tissues by electroporation, it still remains impossible to distinguish between individual cells in the tissues, in particular those near the electroporation site, as the fluorescent protein cDNA is introduced into many cells simultaneously. Therefore, there remains the need to develop alternative methods to mark and monitor specific cells and their progeny within intact, living brain tissue. In order to establish a method of precisely identifying individual cells in cortical slices, we used the photoconvertable fluorescent protein, Kaede. Kaede (the Japanese word for “maple tree”) is a recently isolated protein from a stony coral, Trachyphyllia geoffroyi, which contains a green chromophore that is photoconverted to a red chromophore (Ando et al., 2002). This chromophore contains the tripeptide His62–Tyr83–Gly64. UV radiation induces an unconventional cleavage within the Kaede protein between the amide nitrogen and the alpha carbon at His62 (Mizuno et al., 2003). We were able to photoconvert single cells derived from a Kaede-introduced neuronal progenitor and follow their movements individually in mouse embryonic brain slices. This method enabled us to clearly distinguish a reddened cell of interest from other green fluorescent-positive cells in 3-dimensional tissue samples during extended time-lapse imaging studies. Materials and methods Mice Pregnant ICR mice (SLC Japan, Inc.) were housed in a controlled environment under a regulated 12-h light/dark cycle. The day when a vaginal plug was detected was counted as embryonic day 0.5 (E0.5). DNA preparation Kaede full-length cDNA in the pCS2 expression vector (Ando et al., 2002) was amplified and purified with the QIAGEN (Hilden, Germany) plasmid maxi kit. Electroporation The surgical procedures performed on the pregnant mice and the embryo manipulations exo-utero were carried out basically as previously described (Tabata and Nakajima, 2001). The present experiments conformed to the guidelines for animal experiments set by the Japan Neuroscience Society. On E12.5, pregnant mice were deeply anesthetized with sodium-pentobarbitone at 50 μg/g body weight. The Kaede expression vector was dissolved in PBS (5 μg/ml), and Fast Green (final concentration: 0.01%) was added to the plasmid solution. After the embryos
431
were exposed, approximately 1–2 μl of plasmid solution was injected into the lateral ventricle of the telencephalon with a glass micropipette. The embryos were nipped by the tips of a tweezers-type electrode with a diameter of 5 mm (CUY650-P3; Tokiwa Science, Fukuoka, Japan). Then, five electronic pulses (20 V, 50 ms, at intervals of 950 ms) were administered to each embryo with an electroporator (CUY21E; Tokiwa Science). The embryos were carefully placed back into the abdominal cavity; care was taken to avoid damaging the placenta in order to allow the embryos to continue developing. Adequate amounts of Kaede protein for labeling the cells by electroporation were sufficiently expressed only after 12 h. Organotypic slice cultures Brain slices were prepared basically according to the method of Miyata and co-workers (2002). The embryos were removed on the next day (E13.5) of electroporation. The cerebral hemispheres were isolated in ice-cold oxygenated Dulbecco's modified Eagle's medium (DMEM), and the hemispheres were manually sliced at a thickness of 200–300 μm with handmade micro-knives. The incorporation of Kaede protein into the slices was confirmed using a fluorescence microscope under conditions of blue excitation. The slices were then transferred to a culture medium, i.e., DMEM/F12 medium lacking phenolred (Invitrogen, San Diego, CA), which was supplemented with insulin (25 μg/ml; Sigma, St. Louis, MO), transferrin (100 μg/ ml; Sigma), progesterone (20 nM/ml; Sigma), sodium selenate (30 nM/ml; Sigma), putrescine (60 μM/ml; Sigma), epidermal growth factor (EGF; 10 ng/ml; PrePro TechEC, Rocky Hill, NJ), basic fibroblast growth factor (bFGF; 10 ng/ml; Prepro TeckEC), horse serum (5%; Invitrogen), and fetal calf serum (5%; Invitrogen). Selected slices were transferred with 50 to 100 μl of culture medium onto handmade 35-mm glass-bottom culture dishes. A round cover glass (Matsunami; 22 mm in diameter) was attached to the bottom of a perforated 35-mm dish (Iwaki, Chiba, Japan) with silicon cement (KE42; Shinetzu, Tokyo, Japan). Brain slices were embedded with collagen I gel reagents according to the manufacturer's instructions (Type I Collagen; Nitta Gelatin, Tokyo, Japan). After the gel had solidified, approximately 0.6 ml of pre-warmed culture medium was poured onto the gel, and the medium was spread over the entire dish surface, including the rim, leading to a surface tension-induced reduction in the amount of medium at the center of the gel containing the slices (Miyata et al., 2002). Then, the dish was placed on a microscope stage equipped with a microincubator (Olympus, Tokyo, Japan). The atmospheric conditions were maintained by a flow of premixed N2 gas including 5% CO2 and 40% O2. Imaging and photoconverting Slices were imaged with a confocal microscope controlled by a Fluoview FV300 or FV500 scanning unit (Olympus, Tokyo, Japan), a diode pumped solid-state laser Sapphire (488–20 nm; Coherent Radiation, Palo Alto, CA), a green He/Ne laser, and a 405-nm blue-violet laser diode (DL-LS5005; Sanyo, Osaka,
432
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
Japan). Green and red fluorescence signals were captured simultaneously by the 488- and 543-nm laser lines. The microscope set-up consisted of an inverted microscope (IX70; Olympus) with a standard 100 W Hg2 lamp, a 20× objective lens (UplanApo, numerical aperture [N. A.] 0.70; Olympus), and multipoint time-lapse stage (Sigma Koki, Tokyo, Japan) controlled by Fluoview software. Fig. 1C shows an image obtained by using a 40× lens (N.A. 1.30). Interference filters (excitation and emission filters) contained in the wheels were automated by using Lambda 10-2 hardware (Sutter Instruments, Novato, CA). In this study, the power density of the Sapphire 488–20 nm laser used for the imaging study was 8–12 μW, and that used for the He/Ne laser was 0.5–1.2 μW at the focal plane. In the present study, we were able to maintain viable E13.5 mouse brain slices, and were able to trace the labeled cells for up to 2 days. To photoconvert the cells in a particular slice, a UV beam was generated from a 50-μm diameter pinhole that was attached by an L-shape lamp house unit (Olympus) and a BP330–385 band-pass excitation filter unit (U-MWU2; Olympus). The diameter of the UV beam at the focal plane was approximately 2.5 μm. The position of the beam was adjusted in advance to the center of the optical field using the grid in the eyepiece, and the culture dish was moved until the cell of interest intersected with this point. The shutter was then opened for 10–30 s. The power density of the UV beam used for photoconversion was 1.48 μW at the focal plane. The 405-nm blue-violet laser diode (DLLS5005; Sanyo, Osaka, Japan) was available for photoconversion, and the violet laser beam was adjusted to the point of interest using the clipping function of Fluoview software. To understand the kinetics of the red and green beam emission ratio, we measured the signal intensity of the area of the photoconverted cell body in each channel by using Scion Image software, and we then plotted the obtained values. To investigate the difference in the photoconversion efficiency between the different Kaede-expressing cells, we captured an image of the cell every 10 s after of each radiation. The fluorescence intensity was calculated for each image by using the Scion Image software, and an approximating curve was drawn in Excel. We then calculated the time required for a 50% reduction in the green fluorescence of Kaede. The distance between the cell and the cover glass was determined using the Z axis (focus) information obtained by focusing on the cell. Mitotic and apoptotic cell count The power density of the excitation light was measured with a power meter (TQ8210; Advantest, Munich, Germany). Each
433
Fig. 2. UV radiation at the same strength as that used for photoconversion did not alter the rate of cell mitosis or the rate of apoptosis in cultured brain slices. Entire brain slices were embedded in collagen gel on glass-bottom dishes and were radiated with UV light for 0, 10, and 30 s. Each slice was incubated for 24 h or 48 h. Neither the anti-phospho-Histon H3- (A) nor the TUNEL-positive (B) cell rate was significantly altered by UV radiation. However, significant increase of the TUNEL-positive cell was observed after 5 min exposure to UV radiation (*P = 0.0127).
brain slice was embedded in collagen gel on a glass-bottom dish, and the slices were radiated for 10 or 30 s with UV light from an objective lens lacking the pinhole unit; then, the slices were incubated for 24 or 48 h, respectively. These slices were fixed with 4% paraformaldehyde in PBS for 10 min and were
Fig. 1. A Kaede-introduced cell could be differentiated from the neighboring cells by using photoconversion. In the Kaede-introduced cells in a slice of an E12.5 brain, the red cell (arrowhead) was radiated for 30 s; yellow cell (double arrowhead), 10 s; and yellowish green cell (triple arrowhead), 5 s. Scale bar = 50 μm (A). A higher resolution image of the Kaede-introduced and photoconverted migrating neuron is shown in panel B. This is a single-plane image (optical resolution of the Z-axis: 0.717 μm/slice), and the asterisk indicates the process of another cell. Photoconverted Kaede was accumulated in the tip of the process. We were unable to find aberrant aggregations in the cell. Scale bar = 2 μm (B). Comparison between the fluorescence images of the GFP-introduced slice. It was impossible to distinguish the outline of each cell in the positive cell-rich area. Scale bar = 50 μm (C). The UV radiation time required for a 50% reduction in the green fluorescence of each cell was calculated. Cell depth is negligibly relevant to its photoconversion efficiency (D). The photoconverted cell showed normal radial migration and cell division (E). Photoconversion was performed in the VZ (t = 0.5 h), and the cells were cultured for 72 h. For photoconversion, cell no. 1 and cell no. 2 were UV radiated for 30 s and 10 s, respectively. Cell no. 1 migrated to the MZ; cell no. 2 divided into cell no. 2a and 2b and then migrated to the IZ (t = 73 h). During long-term culture, red fluorescence was slightly appeared in some non-photoconverted cells. This may be caused by nonspecific degradation of Green Kaede. Scale bar = 50 μm (E).
434
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
cryoprotected in a 20% sucrose solution. Serial cryostat sections (15-μm thick) were obtained. All sections were stained with anti-phospho-Histon H3 (Upstate Biotechnology Co., Waltham,
MA) and an ApopTag Apoptosis Detection Kit according to the manufacturer's instructions (Intergen Co., Manhattanville Road Purchase, NY). All nuclei in each section were stained with
Fig. 3. Time-lapse images of Kaede-introduced dividing SVZ neuronal progenitor cells and their daughter cells (A). The progenitor cell (asterisk) divided following the mitotic phase (t = 3 h). The arrowheads indicate the inherited radial fiber of each daughter cell (t = 9 h). Images were acquired every 0.5 h, and each frame shows a stack of 30 optical sections at 2-μm intervals. Scale bar = 20 μm. More details and magnified images are shown in panel C. The kinetics of the fluorescence intensity of the photoconverted cell body in panel A was calculated for the red and green channels. The intensity of red fluorescence remained constant, while that of green fluorescence increased gradually (B). After the loss of contact between the daughter cell bodies (a sequence which started at 6.5 h; blue arrowheads), the short processes of these cells moved rapidly. These processes associated with each other temporarily (sequences that started at 6.5 h and 26.5 h, respectively; white arrowheads). We projected the viewing surface through the 3D image, and investigate their detail positional relation (D; white arrowheads). Finally, the daughter cells underwent complete separation, and they eventually migrated apart (t = 36 h; A).
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
435
Fig. 3 (continued).
DAPI. Cell density was calculated from DAPI-positive nuclei number in outlined pallium area by using Scion Image software. The numbers of immuno-positive cells and DAPI-positive nuclei of the pallium in each slice were counted by using Scion Image software. We then calculated the ratio of the pH3- or TUNEL-positive cells to the total cell count. More than six slices per each condition were examined, and the standard deviations were calculated. Statistical evaluation of the differences between samples was performed using Student's paired t test and Statview J-4.5 software. Differences were considered to be significant at P < 0.05. To verify the amount of photoconversion that occurred under the present assay conditions, the kinetics involved in the decrease of the green fluorescence of Kaede was measured in the same manner as single cell radiation (the entire slice was measured). The approximating curve was drawn, and photoconversion rate was calculated in the case of the less pinhole and pinhole 30 s radiation.
Results and discussion Differentiation of individual cells from other neighboring cells by photoconversion All electroporations were performed exo-utero on E12.5 embryos. On the following day (E13.5), the animals were sacrificed, and the cortical slices were prepared. Kaedeexpressing cells in the slices were observed within the VZ and were found to have subsequently migrated out to the mantle zone of the cerebral wall. Kaede protein was widely distributed across the entire cell body, and its processes were clearly observed using a confocal microscope, which is the same as EGFP. However, it was impossible to distinguish individual cells in the cell-rich areas in the slices since all the cells displayed the same green emission (Fig. 1C). To distinguish single cells from the neighboring fluorescent protein-positive cells in the VZ, we radiated the single cells within each slice
436
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
with a UV beam (see Materials and methods). The Kaede proteins in the UV-radiated cells exhibited a red fluorescence within 30 s, whereas the fluorescence of the Kaede protein in the surrounding cells retained its original green color (Fig. 1A). The ratio of the green fluorescence to the red fluorescence in the UVradiated cells could be controlled by altering the duration of radiation (Fig. 1A, arrowheads). In the cells in these brain slices, the photoconverted Kaede protein diffused throughout the cytosol in less than 1 min (Fig. 1B). We could then distinguish the specific photoconverted cells from other neighboring cells. However, we could not selectively photoconvert the collinear cells located on the same axis of the UV beam since these cells photoconverted simultaneously. One of the possible solutions for this is to utilize the nonlinear properties of two-photon excitation. Some cell-to-cell variability was also observed in the photoconversion efficiency. The average time required for a 50% reduction in the green fluorescence was 29.5 ± 17.8 s (n = 10); the strength of the UV beam used (1.48 μW) was the same in all cases. However, we were unable to correlate the photoconversion efficiency and the distance between the cell and the cover glass (Fig. 1D). Therefore, we concluded that tissue depth had a minor effect on the photoconversion efficiency of each cell since the E12–15 embryonic brain tissue is highly transparent. The photoconversion efficiency may be affected by the amount of Kaede protein in each cell. It has previously been reported that the green-to-red photoconversion of Kaede is irreversible (Ando et al., 2002; Mizuno et al., 2003). Reddened chromophore was clearly detected, even in the processes of the cells, and we were able to observe changes in cell shape throughout the culture period. However, we did not observe a significant reduction in the intensity of the red emissions during the culture period. During this time, the green Kaede protein was being continually produced in the cells; therefore, the photoconverted cells turned yellow (Fig. 3B). In some cases, we could observe the second progenitor division of the photoconverted VZ progenitor. Further, the granddaughter cell could be traced until the end of the culture period by following the red Kaede protein marker (Fig. 1E and Supplemental Fig. 1B). However, occasionally it is slightly difficult to visualize the thin processes of the granddaughter cells by the red fluorescence due to the dilution of the red Kaede protein. These results indicate that the emission of introduced Kaede is sufficient to describe the shape of ventricular progenitor cells in brain slices. Moreover, Kaede was found to be useful for clearly visualizing the outline of the single reddened cell body, including its processes from another single Kaede-transfected cell. The photoconvertable property of Kaede is particularly effective at drawing sharp contrasts between a cell of interest and other neighboring cells. Recently, Tutsui et al. (2005) reported a novel photoconvertable protein, namely, KikGR. This protein is found in the coral Favia favus, and its fluorescence is severalfold brighter than that of Kaede. The photoconversion of KikGR using a UV beam is approximately threefold more rapid than that of Kaede (Tutsui et al., 2005). Further, we could express KikGR in cortical slices by using our
present method (data not shown). The use of KikGR may enhance the efficiency of our system since it provides brighter images and induces less oxidative stress in cells than Kaede. UV radiation at the same strength as required for Kaede photoconversion is not cytotoxic Tamamaki et al. (1999) reported that UV radiation had little effect on neuronal migration, even when the light was strong enough to generate thymine dimers. To determine the extent of UV radiation required for Kaede photoconversion itself, but that would not influence the migration or mitosis of cells in cultured slices, we radiated entire tissue with UV light (BP330-385: 2.7W/cm2) and then cultured these slices for 24 or 48 h. All slices thickened normally, and we did not detect any differences in the histological structures of the UV-radiated and control slices by the end of the culture period (data not shown). The slices were fixed and then cryosectioned for immunostaining with anti-phospho-Histon H3 (pH3) antibody in order to detect the mitotic cell. The apoptotic cells were stained by the TUNEL method. The number of pH3- or TUNEL-positive cells in the sections was calculated. No significant differences were found in the number of pH3-positive cells among the slices subjected to UV radiation for 0 (control), 10, or 30 s. Further, no statistical difference was found in the number of TUNEL-positive cells among the slices subjected to UV radiation for 0 (control), 10, or 30 s (Fig. 2). Significant UV-induced damage appeared under longer durations of radiation, i.e., TUNEL-positive cell percentages rose by 1.49 times in comparison to control samples after 5 min exposure to UV radiation (n = 5, P = 0.0127) (Fig. 2B). These findings indicate that the intensity of UV radiation required for Kaede photoconversion did not exert a significant effect on either the mitotic rate or the apoptotic rate in the cells from cultured brain slices during the 2day culture period. For the same period of radiation, a substantially higher amount of photoconversion occurred when the pinhole-less objective was used than when the objective with the pinhole was used for UV radiation. For example, 74.87% ± 23.68% (n = 4) of green Kaede was photoconverted to its red form in the sample radiated for 30 s using the pinhole-less objective. On the other hand, 46.75% ± 20.60% (n = 10) of green Kaede was photoconverted to its red form in the sample radiated for 30 s using the objective with the pinhole (see Materials and methods). This was because the diameter of the UV beam at the focus point was smaller than the diameter of the entire cell body, and therefore, only a fraction of the total Kaede protein in the cell was photoconverted. We avoided radiating the nuclear region of the cell in the case of radiation through the pinhole. In this case, radiation through the objective with the pinhole may be considerably milder than radiation through the pinhole-less objective. In fact, we observed that the photoconverted cell was able to divide, and the daughter cells migrated normally from the VZ to the MZ during observation (Fig. 1E, Supplemental Fig. 1B). These results indicate that photoconversion affects the probability of mitosis and the migration performance of a cell to a small extent.
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
Behavior of a daughter-pair generated in SVZ As described above, Kaede protein photoconversion is useful for distinguishing individual cells from adjacent fluorescencepositive cells. In particular, this approach was shown to be useful for observing the differential behavior of two daughter cells that had generated from a common parent cell. Using the introduction of Kaede into E13.5 mouse cerebral cortex samples, we observed progenitor division, as well as the individual movements of each daughter cell (Fig. 3A and Supplemental Movie 1). The germinal zone of the E13.5 cerebral cortex is subdivided into the VZ (inner) and the SVZ (outer). At this stage, paired cycling daughter cells are typically generated in the VZ. We have previously reported that some progenitors migrate into the SVZ, where they divide and produce neuron pairs. It has been observed that mitosis during E13–14 in the SVZ or lower intermediate zone (IZ) (non-surface division: NS division) produces a pair of daughter cells that are committed to forming neurons (Miyata et al., 2004; Noctor et al., 2004). In the present study, a spindle-shaped progenitor emerged from the VZ to the SVZ, became round, and then entered the mitotic (M) phase. It was observed that the apical radial process lost contact with the apical plane after cell division. Shortly after cell division, we applied the UV beam to radiate one daughter cell to induce photoconversion. Photoconverted Kaede enabled clear differentiation of the daughter cells from each other, as they were visualized in two different colors red and green. The UVradiated sister cell inherited the parental apical process, and the other sister cell inherited the radial basal process of the parent cell (Fig. 3A). Then, each daughter cell repeatedly extended and retracted new short processes in a dynamic manner, which were thinner than the initial radial processes. After losing contact with the cell body, the short motile processes of each daughter cell seemed to be rapidly and temporarily associating with each other (Fig. 3C and Supplemental Movie 3). Further detailed studies using ultrastructual analysis are required to understand the significance of this phenomenon. Then, the daughter cells escaped from each other, and migrated in different directions (Figs. 3A, B and Supplemental Movie 1). Thus, it was possible to follow the behavior of an individual pair of daughter cells in a four-dimensional manner by using the photoconvertable fluorescent protein Kaede. It is generally believed that newly generated neurons in the germinal zone migrate along radial glial fibers to the mantle zone (Pinto-Lord et al., 1982). The interaction between a migrating neuron and radial glia is crucial for the process of construction of cortical structures (Sanada et al., 2004). Intercellular positional relation is known to be important for layer construction and for the specification of the cerebral cortex. As discussed above, the use of Kaede in a slice-culture
437
system may shed light on such intercellular mechanisms by providing very detailed observations of cell movement and shape, as well as by rendering visible the interactions that take place between multiple cells of interest. Acknowledgments We are grateful to Dr. K. Nagai, Dr. H. Hama, Dr. H. Mizuno, Ms. C. Hara, and Ms. R. Ando (RIKEN, Japan) for the provision of the Kaede expression vector and for their kind help with the confocal microscopy. We would also like to thank Mr. Y. Watanabe (Olympus) for his help with and maintenance of the confocal microscope. This work was supported by MEXT, KAKENHI 16710041. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.expneurol.2006.03.022. References Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., Miyawaki, A., 2002. An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. U. S. A. 99, 12651–12656. Mizuno, H., Mal, T., Tong, K., Ando, R., Furuta, T., Ikura, M., Miyawaki, A., 2003. Photo-induced peptide cleavage in the green-to-red conversion of a fluorescent protein. Mol. Cell 4, 1051–1058. Matz, M., Lukyanov, K., Lukyanov, S., 2002. Family of the green fluorescent protein: journey to the end of the rainbow. BioEssays 24, 953–959. Miyata, T., Kawaguchi, A., Saito, K., Kuramochi, H., Ogawa, M., 2002. Visualization of cell cycling by an improvement in slice culture methods. J. Neurosci. Res. 69, 861–868. Miyata, T., Kawaguchi, A., Saito, K., Kawano, M., Mutoh, T., Ogawa, M., 2004. Asymmetric production of surface-dividing and non-surfacedividing cortical progenitor cells. Development 131, 3133–3145. Miyawaki, A., 2004. Fluorescent proteins in a new light. Nat. Biotechnol. 22, 1374–1376. Noctor, S., Martinez-Cerdeno, V., Ivic, L., Kriegstein, A., 2004. Cortical neurons arise in symmetric and asymmetric division zones and migrate through specific phases. Nat. Neurosci. 7, 136–144. Pinto-Lord, MC., Evrard, P., Caviness Jr., V.S., 1982. Obstructed neuronal migration along radial glial fibers in the neocortex of the reeler mouse: a Golgi-EM analysis. Brain Res. 256, 379–393. Sanada, K., Gupta, A., Tsai, LH., 2004. Disabled-1-regulated adhesion of migrating neurons to radial glial fiber contributes to neuronal positioning during early corticogenesis. Neuron 42, 197–211. Tabata, H., Nakajima, K., 2001. Efficient in utero gene transfer system to the developing mouse brain using electroporation: visualization of neuronal migration in the developing cortex. Neuroscience 103, 865–872. Tamamaki, N., Sugimoto, Y., Tanaka, K., Takauji, R., 1999. Cell migration from the ganglionic eminence to the neocortex investigated by labeling nuclei with UV radiation via a fiber-optic cable. Neurosci. Res. 35, 241–251. Tutsui, H., Karasawa, S., Shimizu, H., Nukina, N., Miyawaki, A., 2005. Semirational engineering of a coral fluorescent protein into an efficient highlighter. EMBO Rep. 6, 233–238.
Dynamic behavior of individual cells in developing organotypic brain slices revealed by the photoconvertable protein Kaede T. Mutoh a,⁎, T. Miyata a , S. Kashiwagi a , A. Miyawaki b , M. Ogawa a a b
Laboratory for Cell Culture Development, Advanced Technology Development Center, Brain Science Institute, Riken Saitama, Japan Laboratory for Cell Function and Dynamics, Advanced Technology Development Center, Brain Science Institute, Riken Saitama, Japan Received 7 July 2005; revised 26 February 2006; accepted 1 March 2006 Available online 6 June 2006
Abstract In recent years, advances in optical imaging methods have facilitated the visualization of events in the developing cortex. In particular, the introduction of DNA encoding fluorescent protein into cells of the embryonic brain allows the visualization of progenitor cells; slice preparations of the cortex then allow the monitoring of the behavior of transfected cells in the context of the living cerebral wall by time-lapse microscopy. Such approaches have provided substantial information about the patterns of neuronal migration. However, as these techniques label large numbers of cells in the ventricular zone (VZ), it is difficult to follow individual cell shape changes or cell behaviors within the VZ, where neuron production and initial migration take place. Here, we report a unique method using the photoconvertable fluorescent protein Kaede, which emits green fluorescence and shifts to emitting red fluorescence upon radiation with UV. Using this method, we were able to follow the behavior of a particular pair of daughter cells among neighboring Kaede-positive cells in the SVZ of mouse brain slices. The spindle shape progenitor divided into two multipolar-shaped daughter cells. The cell–cell borders of daughter cells were clearly visualized, and easily describe the position and distance between two or more cells. The photoconvertable property of Kaede offers a powerful cell marking tool to identify the precise morphology and migratory behaviors of individual cells within living cortical slices. © 2006 Elsevier Inc. All rights reserved. Keywords: Fluorescent protein; Kaede; Photoconversion; Organotypic slice culture; Electroporation; 3D-time-lapse imaging; Neocortical histogenesis; Neuronal progenitor; Cell migration
Introduction The temporally and spatially restricted events of neuron production and migration during early cortical development are important steps in the establishment of the laminated structure of the cerebral cortex. Abnormalities in these events result in malformations of the cortex and in turn lead to brain function deficits. The mechanisms underlying neuronal migration and differentiation may consist of a combination of intrinsic cues and environmental factors. Thus, to gain insights into the potential mechanisms that drive normal morphogenetic events, it is crucial to visualize the behavior of individual cells and their progeny in the brain. Abbreviations: E, embryonic day; IZ, intermediate zone; MZ, marginal zone; SVZ, subventricular zone; VZ, ventricular zone. ⁎ Corresponding author. Fax: +81 48 467 5496. E-mail address: [email protected] (T. Mutoh). 0014-4886/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2006.03.022
Fluorescent vital dyes (e.g., DiI) have a long history as cellular markers, in part because they are very easy to apply to cells; after co-incubation, cells will readily take up such dyes. However, it is difficult to apply these dyes to a single cell, and they are therefore most often used to track the fate of particular cell groups. In recent years, fluorescent proteins have also been used as cellular markers in the field of neuroscience. Fluorescent proteins are easier to handle, less toxic, and not as subject to decreased fluorescence as are chemical dyes as a result of internalization, cell division, or cell growth. Fluorescent protein remains in a cell and its progeny as long as the vector remains present, and the appropriate transcriptional elements drive the expression of fluorescent protein. Thus far, a large number of color and functional variations of fluorescent proteins have been developed (for a review, see Matz et al., 2002; Miyawaki, 2004). In order to introduce a gene into living tissue, electroporation, replication-incompetent viral vectors, lipofection, and gene gun approaches are available. In and exo-
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
utero electroporation is one powerful tool that has been used to introduce such marker proteins into the fetal brain (Tabata and Nakajima, 2001). It is easier and less time-consuming to apply this method than it is to construct viral vectors and generate virions; moreover, electroporation is more efficient at introducing genes than are lipofection and gene gun techniques. However, when fluorescent protein is introduced into tissues by electroporation, it still remains impossible to distinguish between individual cells in the tissues, in particular those near the electroporation site, as the fluorescent protein cDNA is introduced into many cells simultaneously. Therefore, there remains the need to develop alternative methods to mark and monitor specific cells and their progeny within intact, living brain tissue. In order to establish a method of precisely identifying individual cells in cortical slices, we used the photoconvertable fluorescent protein, Kaede. Kaede (the Japanese word for “maple tree”) is a recently isolated protein from a stony coral, Trachyphyllia geoffroyi, which contains a green chromophore that is photoconverted to a red chromophore (Ando et al., 2002). This chromophore contains the tripeptide His62–Tyr83–Gly64. UV radiation induces an unconventional cleavage within the Kaede protein between the amide nitrogen and the alpha carbon at His62 (Mizuno et al., 2003). We were able to photoconvert single cells derived from a Kaede-introduced neuronal progenitor and follow their movements individually in mouse embryonic brain slices. This method enabled us to clearly distinguish a reddened cell of interest from other green fluorescent-positive cells in 3-dimensional tissue samples during extended time-lapse imaging studies. Materials and methods Mice Pregnant ICR mice (SLC Japan, Inc.) were housed in a controlled environment under a regulated 12-h light/dark cycle. The day when a vaginal plug was detected was counted as embryonic day 0.5 (E0.5). DNA preparation Kaede full-length cDNA in the pCS2 expression vector (Ando et al., 2002) was amplified and purified with the QIAGEN (Hilden, Germany) plasmid maxi kit. Electroporation The surgical procedures performed on the pregnant mice and the embryo manipulations exo-utero were carried out basically as previously described (Tabata and Nakajima, 2001). The present experiments conformed to the guidelines for animal experiments set by the Japan Neuroscience Society. On E12.5, pregnant mice were deeply anesthetized with sodium-pentobarbitone at 50 μg/g body weight. The Kaede expression vector was dissolved in PBS (5 μg/ml), and Fast Green (final concentration: 0.01%) was added to the plasmid solution. After the embryos
431
were exposed, approximately 1–2 μl of plasmid solution was injected into the lateral ventricle of the telencephalon with a glass micropipette. The embryos were nipped by the tips of a tweezers-type electrode with a diameter of 5 mm (CUY650-P3; Tokiwa Science, Fukuoka, Japan). Then, five electronic pulses (20 V, 50 ms, at intervals of 950 ms) were administered to each embryo with an electroporator (CUY21E; Tokiwa Science). The embryos were carefully placed back into the abdominal cavity; care was taken to avoid damaging the placenta in order to allow the embryos to continue developing. Adequate amounts of Kaede protein for labeling the cells by electroporation were sufficiently expressed only after 12 h. Organotypic slice cultures Brain slices were prepared basically according to the method of Miyata and co-workers (2002). The embryos were removed on the next day (E13.5) of electroporation. The cerebral hemispheres were isolated in ice-cold oxygenated Dulbecco's modified Eagle's medium (DMEM), and the hemispheres were manually sliced at a thickness of 200–300 μm with handmade micro-knives. The incorporation of Kaede protein into the slices was confirmed using a fluorescence microscope under conditions of blue excitation. The slices were then transferred to a culture medium, i.e., DMEM/F12 medium lacking phenolred (Invitrogen, San Diego, CA), which was supplemented with insulin (25 μg/ml; Sigma, St. Louis, MO), transferrin (100 μg/ ml; Sigma), progesterone (20 nM/ml; Sigma), sodium selenate (30 nM/ml; Sigma), putrescine (60 μM/ml; Sigma), epidermal growth factor (EGF; 10 ng/ml; PrePro TechEC, Rocky Hill, NJ), basic fibroblast growth factor (bFGF; 10 ng/ml; Prepro TeckEC), horse serum (5%; Invitrogen), and fetal calf serum (5%; Invitrogen). Selected slices were transferred with 50 to 100 μl of culture medium onto handmade 35-mm glass-bottom culture dishes. A round cover glass (Matsunami; 22 mm in diameter) was attached to the bottom of a perforated 35-mm dish (Iwaki, Chiba, Japan) with silicon cement (KE42; Shinetzu, Tokyo, Japan). Brain slices were embedded with collagen I gel reagents according to the manufacturer's instructions (Type I Collagen; Nitta Gelatin, Tokyo, Japan). After the gel had solidified, approximately 0.6 ml of pre-warmed culture medium was poured onto the gel, and the medium was spread over the entire dish surface, including the rim, leading to a surface tension-induced reduction in the amount of medium at the center of the gel containing the slices (Miyata et al., 2002). Then, the dish was placed on a microscope stage equipped with a microincubator (Olympus, Tokyo, Japan). The atmospheric conditions were maintained by a flow of premixed N2 gas including 5% CO2 and 40% O2. Imaging and photoconverting Slices were imaged with a confocal microscope controlled by a Fluoview FV300 or FV500 scanning unit (Olympus, Tokyo, Japan), a diode pumped solid-state laser Sapphire (488–20 nm; Coherent Radiation, Palo Alto, CA), a green He/Ne laser, and a 405-nm blue-violet laser diode (DL-LS5005; Sanyo, Osaka,
432
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
Japan). Green and red fluorescence signals were captured simultaneously by the 488- and 543-nm laser lines. The microscope set-up consisted of an inverted microscope (IX70; Olympus) with a standard 100 W Hg2 lamp, a 20× objective lens (UplanApo, numerical aperture [N. A.] 0.70; Olympus), and multipoint time-lapse stage (Sigma Koki, Tokyo, Japan) controlled by Fluoview software. Fig. 1C shows an image obtained by using a 40× lens (N.A. 1.30). Interference filters (excitation and emission filters) contained in the wheels were automated by using Lambda 10-2 hardware (Sutter Instruments, Novato, CA). In this study, the power density of the Sapphire 488–20 nm laser used for the imaging study was 8–12 μW, and that used for the He/Ne laser was 0.5–1.2 μW at the focal plane. In the present study, we were able to maintain viable E13.5 mouse brain slices, and were able to trace the labeled cells for up to 2 days. To photoconvert the cells in a particular slice, a UV beam was generated from a 50-μm diameter pinhole that was attached by an L-shape lamp house unit (Olympus) and a BP330–385 band-pass excitation filter unit (U-MWU2; Olympus). The diameter of the UV beam at the focal plane was approximately 2.5 μm. The position of the beam was adjusted in advance to the center of the optical field using the grid in the eyepiece, and the culture dish was moved until the cell of interest intersected with this point. The shutter was then opened for 10–30 s. The power density of the UV beam used for photoconversion was 1.48 μW at the focal plane. The 405-nm blue-violet laser diode (DLLS5005; Sanyo, Osaka, Japan) was available for photoconversion, and the violet laser beam was adjusted to the point of interest using the clipping function of Fluoview software. To understand the kinetics of the red and green beam emission ratio, we measured the signal intensity of the area of the photoconverted cell body in each channel by using Scion Image software, and we then plotted the obtained values. To investigate the difference in the photoconversion efficiency between the different Kaede-expressing cells, we captured an image of the cell every 10 s after of each radiation. The fluorescence intensity was calculated for each image by using the Scion Image software, and an approximating curve was drawn in Excel. We then calculated the time required for a 50% reduction in the green fluorescence of Kaede. The distance between the cell and the cover glass was determined using the Z axis (focus) information obtained by focusing on the cell. Mitotic and apoptotic cell count The power density of the excitation light was measured with a power meter (TQ8210; Advantest, Munich, Germany). Each
433
Fig. 2. UV radiation at the same strength as that used for photoconversion did not alter the rate of cell mitosis or the rate of apoptosis in cultured brain slices. Entire brain slices were embedded in collagen gel on glass-bottom dishes and were radiated with UV light for 0, 10, and 30 s. Each slice was incubated for 24 h or 48 h. Neither the anti-phospho-Histon H3- (A) nor the TUNEL-positive (B) cell rate was significantly altered by UV radiation. However, significant increase of the TUNEL-positive cell was observed after 5 min exposure to UV radiation (*P = 0.0127).
brain slice was embedded in collagen gel on a glass-bottom dish, and the slices were radiated for 10 or 30 s with UV light from an objective lens lacking the pinhole unit; then, the slices were incubated for 24 or 48 h, respectively. These slices were fixed with 4% paraformaldehyde in PBS for 10 min and were
Fig. 1. A Kaede-introduced cell could be differentiated from the neighboring cells by using photoconversion. In the Kaede-introduced cells in a slice of an E12.5 brain, the red cell (arrowhead) was radiated for 30 s; yellow cell (double arrowhead), 10 s; and yellowish green cell (triple arrowhead), 5 s. Scale bar = 50 μm (A). A higher resolution image of the Kaede-introduced and photoconverted migrating neuron is shown in panel B. This is a single-plane image (optical resolution of the Z-axis: 0.717 μm/slice), and the asterisk indicates the process of another cell. Photoconverted Kaede was accumulated in the tip of the process. We were unable to find aberrant aggregations in the cell. Scale bar = 2 μm (B). Comparison between the fluorescence images of the GFP-introduced slice. It was impossible to distinguish the outline of each cell in the positive cell-rich area. Scale bar = 50 μm (C). The UV radiation time required for a 50% reduction in the green fluorescence of each cell was calculated. Cell depth is negligibly relevant to its photoconversion efficiency (D). The photoconverted cell showed normal radial migration and cell division (E). Photoconversion was performed in the VZ (t = 0.5 h), and the cells were cultured for 72 h. For photoconversion, cell no. 1 and cell no. 2 were UV radiated for 30 s and 10 s, respectively. Cell no. 1 migrated to the MZ; cell no. 2 divided into cell no. 2a and 2b and then migrated to the IZ (t = 73 h). During long-term culture, red fluorescence was slightly appeared in some non-photoconverted cells. This may be caused by nonspecific degradation of Green Kaede. Scale bar = 50 μm (E).
434
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
cryoprotected in a 20% sucrose solution. Serial cryostat sections (15-μm thick) were obtained. All sections were stained with anti-phospho-Histon H3 (Upstate Biotechnology Co., Waltham,
MA) and an ApopTag Apoptosis Detection Kit according to the manufacturer's instructions (Intergen Co., Manhattanville Road Purchase, NY). All nuclei in each section were stained with
Fig. 3. Time-lapse images of Kaede-introduced dividing SVZ neuronal progenitor cells and their daughter cells (A). The progenitor cell (asterisk) divided following the mitotic phase (t = 3 h). The arrowheads indicate the inherited radial fiber of each daughter cell (t = 9 h). Images were acquired every 0.5 h, and each frame shows a stack of 30 optical sections at 2-μm intervals. Scale bar = 20 μm. More details and magnified images are shown in panel C. The kinetics of the fluorescence intensity of the photoconverted cell body in panel A was calculated for the red and green channels. The intensity of red fluorescence remained constant, while that of green fluorescence increased gradually (B). After the loss of contact between the daughter cell bodies (a sequence which started at 6.5 h; blue arrowheads), the short processes of these cells moved rapidly. These processes associated with each other temporarily (sequences that started at 6.5 h and 26.5 h, respectively; white arrowheads). We projected the viewing surface through the 3D image, and investigate their detail positional relation (D; white arrowheads). Finally, the daughter cells underwent complete separation, and they eventually migrated apart (t = 36 h; A).
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
435
Fig. 3 (continued).
DAPI. Cell density was calculated from DAPI-positive nuclei number in outlined pallium area by using Scion Image software. The numbers of immuno-positive cells and DAPI-positive nuclei of the pallium in each slice were counted by using Scion Image software. We then calculated the ratio of the pH3- or TUNEL-positive cells to the total cell count. More than six slices per each condition were examined, and the standard deviations were calculated. Statistical evaluation of the differences between samples was performed using Student's paired t test and Statview J-4.5 software. Differences were considered to be significant at P < 0.05. To verify the amount of photoconversion that occurred under the present assay conditions, the kinetics involved in the decrease of the green fluorescence of Kaede was measured in the same manner as single cell radiation (the entire slice was measured). The approximating curve was drawn, and photoconversion rate was calculated in the case of the less pinhole and pinhole 30 s radiation.
Results and discussion Differentiation of individual cells from other neighboring cells by photoconversion All electroporations were performed exo-utero on E12.5 embryos. On the following day (E13.5), the animals were sacrificed, and the cortical slices were prepared. Kaedeexpressing cells in the slices were observed within the VZ and were found to have subsequently migrated out to the mantle zone of the cerebral wall. Kaede protein was widely distributed across the entire cell body, and its processes were clearly observed using a confocal microscope, which is the same as EGFP. However, it was impossible to distinguish individual cells in the cell-rich areas in the slices since all the cells displayed the same green emission (Fig. 1C). To distinguish single cells from the neighboring fluorescent protein-positive cells in the VZ, we radiated the single cells within each slice
436
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
with a UV beam (see Materials and methods). The Kaede proteins in the UV-radiated cells exhibited a red fluorescence within 30 s, whereas the fluorescence of the Kaede protein in the surrounding cells retained its original green color (Fig. 1A). The ratio of the green fluorescence to the red fluorescence in the UVradiated cells could be controlled by altering the duration of radiation (Fig. 1A, arrowheads). In the cells in these brain slices, the photoconverted Kaede protein diffused throughout the cytosol in less than 1 min (Fig. 1B). We could then distinguish the specific photoconverted cells from other neighboring cells. However, we could not selectively photoconvert the collinear cells located on the same axis of the UV beam since these cells photoconverted simultaneously. One of the possible solutions for this is to utilize the nonlinear properties of two-photon excitation. Some cell-to-cell variability was also observed in the photoconversion efficiency. The average time required for a 50% reduction in the green fluorescence was 29.5 ± 17.8 s (n = 10); the strength of the UV beam used (1.48 μW) was the same in all cases. However, we were unable to correlate the photoconversion efficiency and the distance between the cell and the cover glass (Fig. 1D). Therefore, we concluded that tissue depth had a minor effect on the photoconversion efficiency of each cell since the E12–15 embryonic brain tissue is highly transparent. The photoconversion efficiency may be affected by the amount of Kaede protein in each cell. It has previously been reported that the green-to-red photoconversion of Kaede is irreversible (Ando et al., 2002; Mizuno et al., 2003). Reddened chromophore was clearly detected, even in the processes of the cells, and we were able to observe changes in cell shape throughout the culture period. However, we did not observe a significant reduction in the intensity of the red emissions during the culture period. During this time, the green Kaede protein was being continually produced in the cells; therefore, the photoconverted cells turned yellow (Fig. 3B). In some cases, we could observe the second progenitor division of the photoconverted VZ progenitor. Further, the granddaughter cell could be traced until the end of the culture period by following the red Kaede protein marker (Fig. 1E and Supplemental Fig. 1B). However, occasionally it is slightly difficult to visualize the thin processes of the granddaughter cells by the red fluorescence due to the dilution of the red Kaede protein. These results indicate that the emission of introduced Kaede is sufficient to describe the shape of ventricular progenitor cells in brain slices. Moreover, Kaede was found to be useful for clearly visualizing the outline of the single reddened cell body, including its processes from another single Kaede-transfected cell. The photoconvertable property of Kaede is particularly effective at drawing sharp contrasts between a cell of interest and other neighboring cells. Recently, Tutsui et al. (2005) reported a novel photoconvertable protein, namely, KikGR. This protein is found in the coral Favia favus, and its fluorescence is severalfold brighter than that of Kaede. The photoconversion of KikGR using a UV beam is approximately threefold more rapid than that of Kaede (Tutsui et al., 2005). Further, we could express KikGR in cortical slices by using our
present method (data not shown). The use of KikGR may enhance the efficiency of our system since it provides brighter images and induces less oxidative stress in cells than Kaede. UV radiation at the same strength as required for Kaede photoconversion is not cytotoxic Tamamaki et al. (1999) reported that UV radiation had little effect on neuronal migration, even when the light was strong enough to generate thymine dimers. To determine the extent of UV radiation required for Kaede photoconversion itself, but that would not influence the migration or mitosis of cells in cultured slices, we radiated entire tissue with UV light (BP330-385: 2.7W/cm2) and then cultured these slices for 24 or 48 h. All slices thickened normally, and we did not detect any differences in the histological structures of the UV-radiated and control slices by the end of the culture period (data not shown). The slices were fixed and then cryosectioned for immunostaining with anti-phospho-Histon H3 (pH3) antibody in order to detect the mitotic cell. The apoptotic cells were stained by the TUNEL method. The number of pH3- or TUNEL-positive cells in the sections was calculated. No significant differences were found in the number of pH3-positive cells among the slices subjected to UV radiation for 0 (control), 10, or 30 s. Further, no statistical difference was found in the number of TUNEL-positive cells among the slices subjected to UV radiation for 0 (control), 10, or 30 s (Fig. 2). Significant UV-induced damage appeared under longer durations of radiation, i.e., TUNEL-positive cell percentages rose by 1.49 times in comparison to control samples after 5 min exposure to UV radiation (n = 5, P = 0.0127) (Fig. 2B). These findings indicate that the intensity of UV radiation required for Kaede photoconversion did not exert a significant effect on either the mitotic rate or the apoptotic rate in the cells from cultured brain slices during the 2day culture period. For the same period of radiation, a substantially higher amount of photoconversion occurred when the pinhole-less objective was used than when the objective with the pinhole was used for UV radiation. For example, 74.87% ± 23.68% (n = 4) of green Kaede was photoconverted to its red form in the sample radiated for 30 s using the pinhole-less objective. On the other hand, 46.75% ± 20.60% (n = 10) of green Kaede was photoconverted to its red form in the sample radiated for 30 s using the objective with the pinhole (see Materials and methods). This was because the diameter of the UV beam at the focus point was smaller than the diameter of the entire cell body, and therefore, only a fraction of the total Kaede protein in the cell was photoconverted. We avoided radiating the nuclear region of the cell in the case of radiation through the pinhole. In this case, radiation through the objective with the pinhole may be considerably milder than radiation through the pinhole-less objective. In fact, we observed that the photoconverted cell was able to divide, and the daughter cells migrated normally from the VZ to the MZ during observation (Fig. 1E, Supplemental Fig. 1B). These results indicate that photoconversion affects the probability of mitosis and the migration performance of a cell to a small extent.
T. Mutoh et al. / Experimental Neurology 200 (2006) 430–437
Behavior of a daughter-pair generated in SVZ As described above, Kaede protein photoconversion is useful for distinguishing individual cells from adjacent fluorescencepositive cells. In particular, this approach was shown to be useful for observing the differential behavior of two daughter cells that had generated from a common parent cell. Using the introduction of Kaede into E13.5 mouse cerebral cortex samples, we observed progenitor division, as well as the individual movements of each daughter cell (Fig. 3A and Supplemental Movie 1). The germinal zone of the E13.5 cerebral cortex is subdivided into the VZ (inner) and the SVZ (outer). At this stage, paired cycling daughter cells are typically generated in the VZ. We have previously reported that some progenitors migrate into the SVZ, where they divide and produce neuron pairs. It has been observed that mitosis during E13–14 in the SVZ or lower intermediate zone (IZ) (non-surface division: NS division) produces a pair of daughter cells that are committed to forming neurons (Miyata et al., 2004; Noctor et al., 2004). In the present study, a spindle-shaped progenitor emerged from the VZ to the SVZ, became round, and then entered the mitotic (M) phase. It was observed that the apical radial process lost contact with the apical plane after cell division. Shortly after cell division, we applied the UV beam to radiate one daughter cell to induce photoconversion. Photoconverted Kaede enabled clear differentiation of the daughter cells from each other, as they were visualized in two different colors red and green. The UVradiated sister cell inherited the parental apical process, and the other sister cell inherited the radial basal process of the parent cell (Fig. 3A). Then, each daughter cell repeatedly extended and retracted new short processes in a dynamic manner, which were thinner than the initial radial processes. After losing contact with the cell body, the short motile processes of each daughter cell seemed to be rapidly and temporarily associating with each other (Fig. 3C and Supplemental Movie 3). Further detailed studies using ultrastructual analysis are required to understand the significance of this phenomenon. Then, the daughter cells escaped from each other, and migrated in different directions (Figs. 3A, B and Supplemental Movie 1). Thus, it was possible to follow the behavior of an individual pair of daughter cells in a four-dimensional manner by using the photoconvertable fluorescent protein Kaede. It is generally believed that newly generated neurons in the germinal zone migrate along radial glial fibers to the mantle zone (Pinto-Lord et al., 1982). The interaction between a migrating neuron and radial glia is crucial for the process of construction of cortical structures (Sanada et al., 2004). Intercellular positional relation is known to be important for layer construction and for the specification of the cerebral cortex. As discussed above, the use of Kaede in a slice-culture
437
system may shed light on such intercellular mechanisms by providing very detailed observations of cell movement and shape, as well as by rendering visible the interactions that take place between multiple cells of interest. Acknowledgments We are grateful to Dr. K. Nagai, Dr. H. Hama, Dr. H. Mizuno, Ms. C. Hara, and Ms. R. Ando (RIKEN, Japan) for the provision of the Kaede expression vector and for their kind help with the confocal microscopy. We would also like to thank Mr. Y. Watanabe (Olympus) for his help with and maintenance of the confocal microscope. This work was supported by MEXT, KAKENHI 16710041. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.expneurol.2006.03.022. References Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., Miyawaki, A., 2002. An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. U. S. A. 99, 12651–12656. Mizuno, H., Mal, T., Tong, K., Ando, R., Furuta, T., Ikura, M., Miyawaki, A., 2003. Photo-induced peptide cleavage in the green-to-red conversion of a fluorescent protein. Mol. Cell 4, 1051–1058. Matz, M., Lukyanov, K., Lukyanov, S., 2002. Family of the green fluorescent protein: journey to the end of the rainbow. BioEssays 24, 953–959. Miyata, T., Kawaguchi, A., Saito, K., Kuramochi, H., Ogawa, M., 2002. Visualization of cell cycling by an improvement in slice culture methods. J. Neurosci. Res. 69, 861–868. Miyata, T., Kawaguchi, A., Saito, K., Kawano, M., Mutoh, T., Ogawa, M., 2004. Asymmetric production of surface-dividing and non-surfacedividing cortical progenitor cells. Development 131, 3133–3145. Miyawaki, A., 2004. Fluorescent proteins in a new light. Nat. Biotechnol. 22, 1374–1376. Noctor, S., Martinez-Cerdeno, V., Ivic, L., Kriegstein, A., 2004. Cortical neurons arise in symmetric and asymmetric division zones and migrate through specific phases. Nat. Neurosci. 7, 136–144. Pinto-Lord, MC., Evrard, P., Caviness Jr., V.S., 1982. Obstructed neuronal migration along radial glial fibers in the neocortex of the reeler mouse: a Golgi-EM analysis. Brain Res. 256, 379–393. Sanada, K., Gupta, A., Tsai, LH., 2004. Disabled-1-regulated adhesion of migrating neurons to radial glial fiber contributes to neuronal positioning during early corticogenesis. Neuron 42, 197–211. Tabata, H., Nakajima, K., 2001. Efficient in utero gene transfer system to the developing mouse brain using electroporation: visualization of neuronal migration in the developing cortex. Neuroscience 103, 865–872. Tamamaki, N., Sugimoto, Y., Tanaka, K., Takauji, R., 1999. Cell migration from the ganglionic eminence to the neocortex investigated by labeling nuclei with UV radiation via a fiber-optic cable. Neurosci. Res. 35, 241–251. Tutsui, H., Karasawa, S., Shimizu, H., Nukina, N., Miyawaki, A., 2005. Semirational engineering of a coral fluorescent protein into an efficient highlighter. EMBO Rep. 6, 233–238.