Visualizing morphogenesis in transgenic zebrafish embryos using BODIPY TR methyl ester dye as a vital counterstain for GFP



Green fluorescent protein (GFP) technology is rapidly advancing the study of morphogenesis, by allowing researchers to specifically focus on a subset of labeled cells within the living embryo. However, when imaging GFP-labeled cells using confocal microscopy, it is often essential to simultaneously visualize all of the cells in the embryo using dual-channel fluorescence to provide an embryological context for the cells expressing GFP. Although various counterstains are available, part of their fluorescence overlaps with the GFP emission spectra, making it difficult to clearly identify the cells expressing GFP. In this study, we report that a new fluorophore, BODIPY TR methyl ester dye, serves as a versatile vital counterstain for visualizing the cellular dynamics of morphogenesis within living GFP transgenic zebrafish embryos. The fluorescence of this photostable synthetic dye is spectrally separate from GFP fluorescence, allowing dual-channel, three-dimensional (3D) and four-dimensional (4D) confocal image data sets of living specimens to be easily acquired. These image data sets can be rendered subsequently into uniquely informative 3D and 4D visualizations using computer-assisted visualization software. We discuss a variety of immediate and potential applications of BODIPY TR methyl ester dye as a vital visualization counterstain for GFP in transgenic zebrafish embryos. Developmental Dynamics 232:359–368, 2005. © 2004 Wiley-Liss, Inc.


The study of embryogenesis has been advanced dramatically by the ability to express green fluorescent protein (GFP) or GFP–fusion proteins in specific groups of cells within living embryonic tissues (Dynes and Ngai,1998; Motoike et al.,2000; Goldman et al.,2001). By providing intense contrast enhancement, GFP expression allows cytological processes underlying patterning and morphogenesis to be visualized in vivo with unprecedented detail (Lawson and Weinstein,2002; Field et al.,2003a, b; Isogai et al.,2003). For a wide range of basic research and biotechnological applications, there are rapidly increasing needs to obtain detailed three-dimensional (3D) and four-dimensional (4D) visualizations of GFP-labeled embryonic cells within their native tissue environments.

Among vertebrates, the zebrafish embryo represents one of the most ideal experimental systems to visualize the cellular dynamics of organogenesis in vivo. Owing to their optical transparency and exceedingly rapid rate of development, detailed morphogenetic behaviors of GFP-labeled cells can be readily visualized within both wild-type and mutant zebrafish embryos, as well as in “mosaic embryos” in which wild-type or mutant cells have been transplanted (Traver et al.,2003). Once visualized, the morphogenetic cell behaviors that underlie the formation of embryonic tissues and organs can be analyzed in terms of known patterns of gene expression in the zebrafish embryo (Cooper and Kimmel,1998). For this reason, morphogenetic cell behaviors have become increasingly used as bioassays for examining gene function during zebrafish embryogenesis (Solnica-Krezel and Cooper, 2003), pathogenesis (Davis et al.,2002), and tissue regeneration (Huang et al.,2003). Thus, the need to directly visualize and interpret morphogenetic cell behaviors in vivo has stimulated many laboratories to generate transgenic lines of zebrafish, which express either GFP or GFP–fusion proteins in specific cells, tissues, or organ rudiments (Amsterdam et al.,1996; Long et al.,1997; Huang et al.,2001; Koester and Fraser,2001; Pauls et al.,2001; Gong et al.,2002; Lawson and Weinstein,2002; Davidson et al.,2003).

Although GFP expression can provide exquisite contrast enhancement of cellular and tissue morphology, it is often very useful to obtain a detailed “background” image of tissue structures in tandem with the GFP image, to provide a histological and embryological context for the fluorescently labeled cells located within the specimen. This image can be provided through the use of fluorescent counterstains that label all the cells in a living embryo (Dynes and Ngai,1998). To image GFP and a counterstain simultaneously in a transgenic embryo, it is important to have a counterstain whose fluorescence is spectrally separated from the green fluorescence of GFP.

In this study, we present a new vital-staining method using a synthetic red fluorophore known as BODIPY TR methyl ester dye, which serves as an excellent counterstain for GFP-expressing tissues in transgenic zebrafish. BODIPY TR methyl ester dye readily permeates cell membranes and localizes in endomembranous organelles. However, the dye molecule does not localize strongly in plasma membranes. These localization properties make the dye an ideal vital stain, which can be used to reveal (1) the location and shapes of cell nuclei, (2) the shapes of cells within embryonic tissues, as well as (3) the boundaries of organ-forming tissues within the whole embryo. For time-lapse imaging, the staining properties of BODIPY TR methyl ester also allows arrangements of GFP-labeled cells to be identified in living tissues, thus providing essential information to elucidate the cellular behavior involved in the transformation organ primordia into organ rudiments.

Unlike other BODIPY vital dyes, we have found BODIPY TR methyl ester dye to be fixable using paraformaldehyde. The ability to fix BODIPY TR methyl ester dye, serves to broaden this vital dye's potential range of GFP-counterstain applications to a wide variety of biological systems. We also discuss how dual-channel imaging with BODIPY TR methyl ester dye can be used to obtain 3D and 4D data sets, and strategies for how this information can be used to follow the cell behaviors that underlie germ layer formation and organogenesis.


Historical Background

BODIPY is a tradename that refers to a diverse class of highly photostable, boron-containing diazaindacene fluorophores (Haugland,1996). Different versions of these fluorophores span the entire visible spectrum (Haugland,1996). Owing to their neutral charge, BODIPY fluorophores can be conjugated to biomolecules without seriously perturbing the biochemical properties of the conjugate molecule (Lipsky and Pagano,1985; Pagano et al.,1989; Haugland,1996). Thus, BODIPY dye conjugates have been used widely for several vital stains and live cell tracking purposes.

Several unconjugated BODIPY fluorophores have also been used as vital fluorescent stains for zebrafish embryos (Cooper and D'Amico,1996). These fluorophores easily partition across cell membranes, localize in intracellular yolk platelets, and provide intense contrast enhancement of cytoplasm relative to nucleoplasm and interstitial space. These dyes can be “bulk-applied” to intact embryos through bath application, allowing all cells in the embryo to be rapidly stained (Cooper et al.,1998). Below we describe the spectral, permeation, and intracellular localization properties of a novel fixable vital stain, BODIPY TR methyl ester dye, which make this dye a robust and versatile counterstain to GFP in living zebrafish embryos.

Spectral Properties of BODIPY TR Methyl Ester Dye

Vital imaging of GFP-transgenic embryos requires fluorescent excitation in the blue region of the visible spectrum (Fig. 1A). To visualize cellular and tissue structures in GFP-transgenic zebrafish embryo, a second fluorophore is needed that does not spectrally overlap with GFP. We found that a new fluorophore, BODIPY TR methyl ester dye, serves this purpose.

Figure 1.

Spectral characteristics of the red fluorescent vital stain BODIPY TR methyl ester dye. A: Excitation and emission (i.e., the longer wavelength curve) spectra of enhanced green fluorescent protein (EGFP; green) and BODIPY TR methyl ester dye (red). In living zebrafish embryos, simultaneous dual-channel imaging of these two fluorophores is greatly facilitated by the large chromatic separation in their emission spectra. The excitation maxima of EGFP and BODIPY TR methyl ester dye lie close, respectively, to the 488-nm and 568-nm lines of Ar-Kr lasers, which are commonly used as the light source for scanning laser confocal microscopes. B: Excitation and emission spectra of BODIPY TR methyl ester dye, solubilized in either dimethyl sulfoxide (DMSO, blue) or dioleoylphosphatidylcholine liposomes (DOPC, red). Only minor spectral differences are produced by solvating the dye in the two different media.

The red emission maximum of BODIPY TR methyl ester dye is around 625 nm, making this fluorophore ideal for dual-channel imaging with GFP, whose emission maximum lies at 508 nm (Fig. 1A). The excitation curve separation between GFP and BODIPY TR methyl ester dye is substantial, with only minor overlap in the 475- to 525-nm range. The molar extinction coefficient of BODIPY TR methyl ester dye at 489 nm is ∼1,750 M−1 cm−1 or approximately 3% of that of EGFP (55,000 M−1 cm−1; Patterson et al.,2001). The excitation separation indicates that these two fluorophores can be separately excited with the 488 nm and 568 nm lines of an Ar-Kr air-cooled gas laser, a common light source for dual-channel epifluorescence confocal imaging.

The emission spectra of GFP and BODIPY TR methyl ester dye are also well separated (Fig. 1A). This allows concurrent dual-channel confocal imaging, without significant spectral “bleedthrough” of GFP fluorescence into the BODIPY TR methyl ester dye imaging channel. Moreover, the spectral properties of BODIPY TR methyl ester dye do not change when the dye is solvated in dimethyl sulfoxide (DMSO) or liposomes (Fig. 1B), indicating that the functionally important spectral separation of BODIPY TR methyl ester dye and GFP is unlikely to be affected by fluorophore environment.

Permeation and Cytological Localization of Vital Stain

Several unconjugated BODIPY fluorophores are lipophilic and readily partition into the lipidic yolk platelets that are dispersed within the cellular cytoplasm of teleost fish embryos (e.g., BODIPY 505/515, BODIPY FL dyes; Cooper et al.,1999). In comparison to BODIPY TR methyl ester dye, these unconjugated BODIPY dyes do not localize in nucleoplasm or interstitial space, thus providing vivid contrast enhancement to living tissue architecture. After vital staining with BODIPY TR methyl ester dye, it is possible to simultaneously image GFP-expressing cells, as well as the cellular structure of organ-forming primordia within a zebrafish embryo, using dual-channel epifluorescence confocal imaging (Figs. 2, 3). By combining these two images, the GFP expression domain can be readily recognized within the 3D structure of the embryo (Fig. 2C).

Figure 2.

BODIPY TR methyl ester dye serves as a fluorescent counterstain to green fluorescent protein (GFP) in live zebrafish embryos. A 19-somite embryo. A: Tail rudiment of a tbx6-GFP transgenic embryo counterstained with BODIPY TR methyl ester dye (lateral view). As shown in the GFP channel, tbx6 promoter drives GFP expression in the caudal mass stem cells, as well as in the postanal presomitic mesoderm (psm). D indicates dorsal and P indicates posterior for A, B, and C. B: Cellular structure in tissues not labeled with GFP, such as the enveloping layer (EVL) and neural plate (np), are readily visible using BODIPY TR methyl ester dye fluorescence. Loosely organized mesenchymal stem cells in the embryo's caudal cell mass migrate and join cells in the posterior limit of the more compacted presomitic mesoderm (psm). C: Spectral bleedthrough between the GFP and BODIPY TR methyl ester dye fluorescent images is minimal (compare images A and B). This low bleedthrough allows the two fluorescent images to be confidently pseudocolored and merged to detect the true spatial overlap of the two fluorophores (yellow). Scale bar = 100 microns in A (applies to A–C).

Figure 3.

Localization patterns of BODIPY TR methyl ester dye in zebrafish embryonic tissues. A: The lines indicate where B and C represent cross-sectional views from the animal pole of a 30% epiboly stage embryo. B: Deep cell blastomeres and enveloping layer (EVL) epithelial cells. Cell nuclei and interstitial space are not stained with BODIPY TR methyl ester dye. C: Higher magnification of the EVL superficial epithelium. BODIPY TR methyl ester dye localizes to numerous intracellular membranous organelles. Dark boundaries at the epithelium's tight junctions reveal that BODIPY TR methyl ester dye does not localize strongly to plasma membranes. D: Boxes indicate where E and F represent lateral views of a 20-somite embryo. E: Eye rudiment in a 20-somite embryo; right side of the embryo. BODIPY TR methyl ester dye lightly stains the interstitial fluid. A, anterior. F: Cellular structure of organ rudiments revealed by BODIPY TR methyl ester dye staining. nc, notochord; nr, neural rod; fp, floor plate. The dye stains intracellular organelles preferentially, before localizing in yolk platelets. This staining is seen in the yolk extension, where yolk sac layer (YSL) nuclei remain unstained, whereas YSL cytoplasm becomes brightly stained. Underlying yolk platelets in the yolk extension remain unlabeled. P, posterior. Scale bars = 50 microns in B,C,E,F.

Although BODIPY TR methyl ester dye is both membrane-permeant and lipophilic, this vital dye localizes in numerous intracellular organelles (Fig. 3). BODIPY TR methyl ester dye does not localize in either plasma membranes or nucleoplasm (Fig. 3C,E,F). Similar to other BODIPY vital dyes (Cooper et al.,1999), we have not observed any teratogenic effects of BODIPY TR methyl ester dye on embryonic development (data not shown).

BODIPY TR Methyl Ester Dye Is Fixable

Several lipophilic BODIPY vital dyes (e.g., BODIPY 505/515 dye) are not fixable in zebrafish embryos. The inability to fix most unconjugated BODIPY fluorophores often limits the type of experiment that can be conducted with these probe molecules. Fortunately, BODIPY TR methyl ester dye can be easily fixed with 4% paraformaldehyde. Although, fixation mildly increases background fluorescence, cellular cytoplasm can still be easily distinguished from nucleoplasm and interstitial space in fixed tissue (Fig. 4). In addition, the cellular structure of organ primordia and organ rudiments are readily discerned in fixed embryos.

Figure 4.

Zebrafish embryos vitally stained with BODIPY TR methyl ester dye are fixable with 4% paraformaldehyde. A, anterior; P, posterior. A,B: A vitally stained 36-hr embryo after fixation. BODIPY TR methyl ester dye fluorescence does not extend into the green wavelengths (A); the fluorescence remains in red wavelengths (B). The boxes in B show the regions seen in C and D. C: Close up of embryo from B, showing cellular structure in the notochord and myotomes are visible after fixation. A, anterior; P, posterior. D: Close up of embryo from B, showing tissue structures in the eye rudiment after fixation. E,F: Tail rudiment in a tbx6–green fluorescent protein (GFP) transgenic embryo vitally stained with BODIPY TR methyl ester dye after fixation. GFP and BODIPY TR methyl ester dye fluorescence remain separate. A 20-somite embryo. psm, presomitic mesoderm. Scale bars = 100 microns in A (applies to A,B), in C, in D, in E (applies to E,F).

Even after paraformaldehyde fixation, the emission fluorescence of GFP and the emission fluorescence of BODIPY TR methyl ester dye remain well separated chromatically (Fig. 4E,F). For both live and paraformaldehyde-fixed zebrafish embryos, the lack of spectral bleedthrough between the GFP and BODIPY TR methyl ester dye allows for a multitude of postimaging visualization strategies, using computer-assisted image processing.

Visualization Strategies Using the GFP Vital Counterstain

Dual-channel confocal images are commonly presented as side-by-side visualizations. Figure 5 shows a membrane targeted GFP (mGFP) transgenic zebrafish embryo counterstained with BODIPY TR methyl ester dye. The fluorescent counterstain helps reveal the position of cellular nuclei and other cellular morphologies, which help confirm cellular phenotypes within nascent somites and presomitic mesoderm.

Figure 5.

Membrane-targeted green fluorescent protein (mGFP) transgenic fish counterstained with BODIPY TR methyl ester dye. Anterior is at the top and dorsal (D) is to the left in all panels. P, posterior. A 20-somite embryo. A: Tract of presomitic mesoderm (psm) and blood cells. The most recently formed somite (S0) and the first presumptive somite (S-1) are visible in the mGFP transgenic zebrafish. B: Red fluorescence reveals positions of cell nuclei, confirming that the presomitic mesoderm is covered by a monolayer of epithelial cells. C: Nascent somites (i.e., S0 and S-1) in an mGFP transgenic zebrafish counterstained with BODIPY TR methyl ester dye, showing the epithelial border cells and internal mesenchymal cells (GFP fluorescence, excited with 488-nm laser light). D: The same embryo as in C, illuminated with 568-nm laser light. The BODIPY TR methyl ester dye fluorescence reveals complimentary information about cellular structure within nascent somites. Scale bars = 100 microns in A (applies to A,B), in C (applies to C,D).

In cases where two fluorescent images both contain highly detailed structural information throughout the field of view, a side-by-side comparison is best suited for distinguishing similarities and differences in tissue architecture. Merged images, on the other hand, from dual-channel confocal imaging, provide one of the best means of determining colocalization of fluorescently labeled objects (Fig. 2). In the case of GFP-labeled embryos counterstained with BODIPY TR methyl ester dye, merged images can be used to reveal microdomains of cells that either express, or fail to express, GFP-reporter genes. In the case of a tbx6-GFP transgenic zebrafish embryo, microdomains of cells that fail to express GFP in the posterior presomitic mesoderm can be clearly identified in merged images (Fig. 6A–C). When spatially contiguous cells in a tissue simultaneously express GFP, the cellular morphology of these cells can be easily distinguished using BODIPY TR methyl ester dye as a vital counterstain (Fig. 6D–F).

Figure 6.

Cellular domains can be visualized and distinguished in a BODIPY TR methyl ester dye counterstained tbx6–green fluorescent protein (GFP) transgenic embryo. Posterior (P) is to the right and dorsal (D) is at the top in all panels. A–C: Several cellular microdomains, which do not express tbx6-GFP in trunk somites, are easily identified in a pseudocolored merged image (C) of A (GFP) and B (BODIPY TR methyl ester dye). A 15-somite embryo, lateral view. D,E: BODIPY TR methyl ester dye counterstaining (E) helps reveal that bipolar myocytes are differentiating in the dorsal aspect of posterior somites. D: GFP fluorescence does not reveal enough cellular detail to make this conclusion. F: Merged image of D and E. A 19-somite embryo, lateral view. Scale bars = 50 microns in A (applies to A–C), in D (applies to D–F).

A merged-image time-lapse recording of a tbx6-GFP transgenic embryo reveals the onset of tbx6 expression in migrating presumptive blood cells, which subsequently aggregate into a blood island (Supplementary Movie S1, which can be viewed at http:// This time-lapse recording illustrates that the expression of a GFP-reporter gene is not affected by vital staining with BODIPY TR methyl ester dye.

Single image plane time-lapse recordings are one of many types of visualization strategies to analyze the morphogenesis of embryonic tissues. Other imaging and visualization strategies require an analysis of how BODIPY TR methyl ester dye withstands repeated photoillumination. Multiple scans of a specimen are needed for a time-lapse series (Supplementary Movie 1), a 3D reconstruction, or a 4D morphogenetic visualization. For live specimens, there is a limit on how many times a specimen can be scanned before photobleaching reduces fluorescence below usable levels. We have found that approximately 100 scans can be made of a zebrafish embryo stained with BODIPY TR methyl ester dye before photobleaching degrades image quality. Highly informative 2D time-lapse recordings can easily be obtained with fewer than 100 image scans of a specimen (Supplementary Movie S1). We have found that BODIPY TR methyl ester dye has low phototoxicity similar to other BODIPY vital dyes (Cooper and Kimmel,1998; Cooper et al.,1999), owing to its low photobleaching rate.

A single confocal z-series (also known as a “focus-through”), containing approximately 20–40 image planes, can be rendered easily into a single QuickTime movie, which can be used to navigate through the 3D tissue structure of an embryo (Fig. 7A–C, Supplementary Movie S2). By arranging two channels of fluorescence, as well as a merged pseudocolor image, in a side-by-side presentation, a very informative histological visualization can be generated (Fig. 7D–I, Supplementary Movie S2).

Figure 7.

Two visualization strategies for viewing tissue structures in vitally stained green fluorescent protein (GFP) -transgenic embryos. A 20-somite stage embryo. A–C: Three images from a three-dimensional virtual reality (VR) visualization (see Supplementary Movie S3), showing GFP in tracts of presomitic mesoderm and posterior somites in the tail rudiment of a tbx6–GFP transgenic zebrafish embryo. The embryo has been counterstained with BODIPY TR methyl ester dye. The different panels show the same tail rudiment in different orientations. D–I: Panels from a left to right dual-channel focus-through QuickTime visualization (Supplementary Movie S2). D–F show one focal plane cutting through one region of presomitic mesoderm. G–I show a focal plane through the midline of the embryo. D,G: GFP fluorescence. E,H: BODIPY TR methyl ester dye fluorescence. F,I: merged image. Posterior (P) is at the top, dorsal (D) is to the right. Scale bars = 50 microns in A (applies to A–C), in D (applies to D–I).

A 3D data set of approximately 50–100 image planes can also be converted into a virtual reality (VR) 3D-rotation visualization using commercially available rendering software (see Fig. 7A–C and Supplementary Movie S3). The great advantage to these images is that the GFP expression domain, within the context of the nonexpressing regions, can be viewed from any angle (Supplementary Movie S3). The 4D confocal data stacks can be arranged conveniently into multilevel time-lapse visualization (not shown), which can be viewed and analyzed using navigation freeware (e.g., 4D Turnaround) or commercial software (Cooper et al.,1999).


Although GFP expression can provide detailed views of microscopic cellular processes in living tissues, additional visual information is often needed to answer questions concerning the morphogenetic and social behavior of embryonic cells within the context of the whole embryo. This information is not located entirely at the microscopic scale of cell processes. Rather, it lies in a higher level of spatial order, known as the mesoscopic scale (Felderhof,2003; Lund and Jonsson,2003). Mesoscopic information deals with the population behavior of interacting objects. In the case of embryonic cells, mesoscopic behaviors include the morphogenetic movements of embryonic cell populations during morphogenesis and organogenesis.

Germ layers of gastrula and organ-forming primordia are dynamic metastructures (i.e., transient structures), which can be analyzed in terms of the expression of structural and behavioral phenotypes. To perceive the social dynamics of GFP-labeled cells in situ, one must consider optimal ways of enhancing contrast in neighboring cells as well. One way is to collect a brightfield, differential interference contrast (DIC) or phase contrast image concurrently with the fluorescent image. The nonfluorescent image can be presented side-by-side with the fluorescent image, or the two images can be superimposed. In either case, the nonfluorescent image provides mesoscopic information essential for the visual perception of histological structure and orientation. However, it is not possible to merge 3D DIC or phase images with 3D fluorescence data to produce informative 3D visualizations, because edge boundaries in DIC and phase contrast images do not superimpose well in 3D reconstructions.

BODIPY TR methyl ester dye provides excellent contrast enhancement to the tissues in which GFP is concurrently expressed. Mesoscopic (i.e., intermediate spatial scale) visual information about tissue structures and boundaries provides a richer image context for the cells whose contrast is enhanced by GFP. GFP-transgenic embryos counterstained with BODIPY TR methyl ester dye are also ideally suited for dual-channel fluorescence time-lapse imaging. Other researchers have used red BODIPY fluorophores (e.g., BODIPY 548/578 dye and BODIPY 564/570 dye) as counterstains for GFP (Dynes and Ngai,1998; Picker et al.,2002). BODIPY TR methyl ester dye is superior to these dyes because of its spectral separation from GFP. The ability to rapidly stain all cells in an entire transgenic zebrafish embryo (labeled with GFP and/or other biofluorophores) with a fluorescent, nonteratogenic counterstain dramatically improves the dimensions of experimental investigation.

In this study, we have counterstained GFP transgenic zebrafish with BODIPY TR methyl ester dye to visualize the formation of the somites from the somitic stem cells, as well as the cytomechanics of somitogenesis. In addition, BODIPY TR methyl ester dye does not interfere with the imaging of GFP-reporter genes (see Supplementary Movie S1). Presumptive blood cells begin expressing GFP through the tbx6 promoter hours after the cells have been vitally stained with BODIPY TR methyl ester dye. The ability to (1) visualize somitic stem cells, (2) observe the activation of gene expression, and then (3) track the behavior of individual cells demonstrates how BODIPY TR methyl ester dye can be used in tandem with GFP to study cellular dynamics underlying tissue morphogenesis in zebrafish.

Morphogenesis of embryonic tissues can be readily analyzed in 3D time-lapse renderings (also known as 4D visualizations). 4D visualizations of embryological development, like 3D renderings, usually require tight spacing between image planes, to reduce interpolation errors during computer rendering calculations (Majlof and Forsgren,1993; Thomas et al.,1996). With single-photon confocal microscopy, one is limited to approximately 100 image scans, before BODIPY TR methyl ester dye fluorescence is substantially photobleached. By using multiphoton confocal microscopy, however, it may be possible to obtain many more individual scans of a zebrafish embryo vitally stained with BODIPY TR methyl ester dye. Larger 4D confocal image data sets of living embryos permit longer and more detailed 4D visualizations to be generated.

Multiphoton confocal microscopy allows much more latitude in repeated photoillumination of living specimens (e.g., 3D and 4D data sets), because photoexcitation and photobleaching is vastly reduced in out-of-focus image planes (Dickinson et al.,2003). A fluorescent counterstain to neighboring cells is a case where a spectrally separate counterstain could be of great utility in multiphoton confocal imaging applications. The localization properties of BODIPY TR methyl ester dye in intracellular organelles, as well as its ability to be fixed by paraformaldehyde, suggest that this dye may have wide applications outside of zebrafish embryos.

New confocal imaging detectors, such as the META system, may profit from the photostability of a vital GFP counterstain (Dickinson et al.,2003). In the META detector system, images are sequentially acquired using different wavelengths (termed Lambda stacks) of excitation or emission. By using a linear de-mixing algorithm, excitation or emission fingerprinting of multiple spectrally overlapping fluorophores is possible. Fluorescence from these spectrally overlapping fluorophores can be separated subsequently into individual imaging channels. The stability of the BODIPY TR methyl ester dye in DMSO and liposomes suggests that the fluorescence of BODIPY TR methyl ester dye will be similar within living cells. This stability in its fluorescence spectra could be a technical advantage for use with the META detector system, because the imaging detection scheme relies upon fluorophore excitation or emission fingerprinting.

The potential uses of BODIPY TR methyl ester dye as a vital counterstain to GFP are likely to increase as biomedical and basic research uses of zebrafish expand (Gong et al.,2003). In this regard, BODIPY TR methyl ester dye will likely serve as a valuable counterstain, to facilitate the investigation of transgenic embryos, in which multiple fluorescent fusion proteins (e.g., GFP, DS Red, and HC1; Miyawaki et al.,1997; Finley et al.,2001; Verkhusha et al.,2001; Hadjantonakis et al.,2003) have been engineered into the genome (Wan et al.,2002).



The systematic name for the BODIPY TR methyl ester dye is 4-(4,4-difluoro-5-(2-thienyl)-4-bora-3a,4a-diaza-s-indacene-3-yl)phenoxyacetic acid, methyl ester. We refer to the dye throughout this study in the abbreviated form: BODIPY TR methyl ester dye. BODIPY TR methyl ester dye was obtained from Molecular Probes (Eugene, OR) and will be marketed by Molecular Probes under the product name CellTrace BODIPY TR methyl ester (catalog no. C34556) in July 2004.

Spectroscopic Measurements

Absorption spectra were measured on a Perkin-Elmer Lambda 35 spectrophotometer. Fluorescence emission spectra were measured on a Hitachi F4500 spectrofluorometer. Dioleoylphosphatidylcholine (DOPC; Avanti Polar Lipids, Alabaster, AL) liposomes containing BODIPY TR methyl ester dye (1:100 mol:mol DOPC) were prepared by the ethanol injection technique described by Kremer et al. (1977).

Vital Staining Procedures

The vital staining procedures for BODIPY TR methyl ester dye are the same used for other unconjugated BODIPY dyes (for more details, see Cooper et al.1999). Briefly, anhydrous DMSO is used to solubilize BODIPY TR methyl ester dye into a 5 mM stock solution. The stock DMSO solution is diluted 1:50 into Embryo Rearing Medium (ERM) (Westerfield,1995) buffered with 5 mM HEPES (pH 7.2; Cooper et al.,1999), making a final labeling solution of 100 μM BODIPY TR methyl ester dye with 2% DMSO. Aliquots of the stock BODIPY TR methyl ester dye solution can be conveniently placed in microfuge tubes and stored at −20°C until needed.

Embryos are stained in 100 μM BODIPY TR methyl ester dye for 1 hr, then passed through three successive washes in HEPES-buffered ERM. The embryos are then mounted in an open-faced chamber for imaging (Cooper et al.,1999). For time-lapse recordings, vitally stained embryos were deyolked and secured to a coverslip by using a modified plasma clot technique (Langenberg et al.,2003).

Embryos used for fixation experiments were fixed in 4% paraformaldehyde in HEPES-buffered ERM at 4°C for 1 hr, before being washed in ERM and mounted for confocal imaging.

Imaging Optics and Procedures

All images in this study were collected by using a Bio-Rad MRC-600 confocal microscope, equipped with a 640 × 480 pixel frame buffer and an air-cooled Ar-Kr laser. Air-cooled Ar-Kr lasers are commonly used in laser confocal microscopes, owing to their stability and relatively low-cost. An effective way of coexciting GFP and BODIPY TR methyl ester dye is to illuminate the zebrafish embryo with two wavelengths of the major Ar-Kr laser lines (i.e., 488 nm and 568 nm) simultaneously. The fluorescent light can then be split by a dichroic mirror into two beam paths leading to separate photodetectors. Light from GFP is collected using 490- to 550-nm bandpass filter, whereas BODIPY-TR methyl ester dye is collected using a 575-nm long-pass filter.

Production of Transgenic Zebrafish

Two lines of GFP-transgenic zebrafish were used in this study. The tbx6-GFP transgenic zebrafish line Tg(tbx6:gfp) is described in Szeto and Kimelman (manuscript submitted for publication). The mGFP transgenic line Tg(β-actin:mgfp) is a membrane-targeted GFP that uniformly labels cell outlines and was constructed by first fusing a GFP construct with the last 20 amino acids of c-Ha-Ras (Jiang and Hunter,1998), a sequence that provides farnesylation and palmitoylation signals for targeting the GFP to the plasma membrane. This construct was placed under the control of the medaka β-actin promoter (Chou et al.,2001; details of the construction available from J.T. and L. S-K.). The transgenic animals were generated by injecting plasmid DNA into one-cell stage embryos and a founder line was generated (J. Topczewski and L. Solnica-Krezel, unpublished results).

Visualization Procedures

QuickTime movies and VR visualizations were generated from confocal data sets by using Volocity 2.0 software (Improvision, Lexington MA).


M.S.C. was funded by the the NSF, and L.S-K. and D.K. were funded by the NIH. D.S. was supported by a PHS postdoctoral fellowship.