Multiphoton laser scanning microscopy (MPLSM) enables the production of long timelapse recordings from live fluorescent specimens. 1047- and 900-nm excitation were used to image both a vital fluorescent membrane probe, FM 4-64, and a modified green fluorescent protein (GFP) in live Caenorhabditis elegans embryos. Automated four-dimensional (4D) data collection yielded individual recordings comprising thousands of images, each allowing analysis of all of the cell divisions, contacts, migrations, and fusions that occur during a span of several hours of embryogenesis.
© Optical Society of America
Fluorescence imaging by MPLSM provides optical sectioning with a low degree of photobleaching and phototoxicity to the sample [1, 2]. It is therefore well suited to the long-term repeated acquisition of stacks of optical sections, the basis of 4D recording from live specimens. The model organism C. elegans lends itself to microscopy in general because of its transparency and small size, and the reproducibility and speed of its development. The advanced molecular genetics of this system has enabled widespread use of fluorescently tagged cloned proteins for in vivo microscopic studies of individual molecules in whole animals . C. elegans embryos are particularly qualified for 4D analysis, as the whole specimen (approximately 60 × 30 μm) fits within a high magnification microscope field, and the lineage of virtually all of the > 500 cells that make up the hatching animal is completed within six hours of fertilization [4–7]. Furthermore the initial differentiation and morphogenesis of a number of tissues commence while the embryo is still amenable to timelapse recording, before the beginning of relatively rapid muscular movement.
We have begun to explore the potential of MPLSM in our studies of C. elegans development. In this report we describe imaging of live embryos labeled with either vital dyes or endogenously expressed fluorescent proteins, and we present animations demonstrating some of the interesting features of embryogenesis which are revealed by these techniques.
2. Materials and methods
Imaging was performed on either of two Nikon Diaphot Quantum inverted microscopes, one outfitted with a modified MRC 600 (Bio-Rad) control system and laser scanhead and a pulsed 1047-nm Nd:YLF excitation laser (Microlase DPM1000) , the other with a modified MRC 1024 (Bio-Rad) control system, a custom-designed scanhead, and a pulsed, tunable Ti:Sapphire (Spectra-Physics Tsunami) excitation laser (unpublished). Both microscopes were configured for direct detection of emitted fluorescent light passing through either an 850-nm short-pass (Nd:YLF excitation) or 650-nm short-pass (Ti:Sapphire) dichroic mirror. Each system was controlled by commercial software accompanying the scanning/imaging hardware. 4D acquisition was directed by either built-in (MRC 1024) or custom macro-based (MRC 600) program features. All imaging was done using a 60X, 1.4 NA Nikon oil immersion lens.
In a typical experiment a box size of 768 × 512 pixels was used for slow-scan image collection, with 30–60 optical sections, spaced 1.0 or 0.5 μm apart, collected per timepoint. The photomultiplier tube was set to full gain, and excitation power was set empirically to the minimum required to give a discernible image. In practice, this was usually ~4 mW average power (~160 W average peak power) for the Nd:YLF laser and ~3 mW average power (~350 W average peak power) for the Ti:Sapphire laser tuned to 900 nm. Although excessive excitation could result in rapid photobleaching, we were able through judicious use of excitation power to achieve good 4D recordings of 3–6 hours of development comprising up to 2800 images.
2.2 Sample preparation
Embryos were adhered to poly-L-lysine-coated glass coverslips, mounted in physiological buffer, and the coverslip edges sealed with silicone oil. In dye-labeling experiments, the impermeable eggshell was perforated using a coumarin-440 dye ablation laser (Photonics Instruments), allowing the aqueous-soluble vital probe FM 4–64 (Molecular Probes, used at 10 μm/mL in the surrounding medium) to enter the embryo. Specimens were imaged at an ambient room temperature between 18°C and 22°C.
2.3 Image Data Analysis
For standard 4D animation of a specimen, a series of separate stacks of optical sections, one from each timepoint, was converted to a single 4D QuickTime movie using the program 4D Turnaround . For stereo-4D reconstruction, timepoint stacks were volume-rendered and edited in three dimensions before conversion to 4D movies . Converted 4D data were viewed using the program 4D Viewer , and selected tracks were post-processed with Adobe Premiere.
3.1 Cell membranes labeled with FM 4–64
When imaged using the 1047-nm laser, embryos labeled with FM 4–64 displayed brightly fluorescent cell membranes while the cellular contents generally remained dark. Fig. 1 depicts a timelapse sequence from a single focal plane of mid-cleavage stage embryo. Cytokinetic furrowing of the membranes of dividing cells is clearly seen. Also apparent are the migration and ingression of external cells into the embryo interior during the process of gastrulation.
The “hollow-bubble” distribution of label in most cells is probably due to exclusion of the aqueous-soluble dye by the intact cell membrane. In Fig. 2, (additional data from the same embryo shown in Fig 1) a programmed cell death is followed in timelapse from the cell’s birth and migration between focal planes through its apparent engulfment by a neighboring cell. The pronounced brightening of the dying cell is most likely due to loss of cell membrane integrity.
Some specialized cell types showed atypical labeling which included more than just the outer plasma membrane of the cell. The precursors of the intestine displayed a unique pattern of fluorescence, characterized by small spots of intracellular labeling (Fig. 3). Interestingly, this distribution of dye may indicate the onset of endocytotic activity in these cells, even before the final cell division of this lineage, ~2 hours before the gut takes shape, and ~9–10 hours before the animal hatches and ingests its first meal.
In another example, the outgrowing axons of neurons within the nerve ring - the processing center of the nematode nervous system - became distinct as they formed the bundles which define and extend away from the central ring (see stereo-4d reconstruction in Fig. 4).
In late-stage embryos comprising >500 cells, the membranes of individual cells were distinguishable only either in standard 4D format (i.e. as optical sections) or after isolation by 3D editing of a stereo-4D reconstruction . The precursors of the external epithelium (hypodermis) could be viewed as a contiguous cellular sheet only in stereo-4D format, and only after elimination of signal from other cells deep within the embryo (see Fig. 5). Through use of these “cored” stereo-4D reconstructions, we have recorded and made quantitative measurements of the fusion of precursor cells during the formation of a large multinucleated hypodermal syncytium .
3.2 Epithelial cell junctions labeled with GFP
Another useful marker in studies of cell movements and fusions has been a GFP-tagged version of the protein MH27, which is localized to the intercellular adherens junctions between epithelial cells [11, 10]. Having studied this marker successfully by conventional confocal microscopy, we explored the potential of imaging transgenic embryos with MPLSM. By tuning the Ti:Sapphire laser to several wavelengths between 800 and 900 nm, we settled upon 900 nm as an practical optimum for imaging of this variant of GFP. Continuous 4D recording with this excitation was possible over a time period extending from the onset of MH27-GFP expression, through epithelial cell migrations and fusions leading to full enclosure of the embryo, to the onset of muscular movement (Fig. 6). Unlike confocal recordings, no conspicuous photobleaching was seen in these data sets, and development proceeded normally over the entire imaged time span.
Recently colleagues have extended this approach to the systematic study of morphogenetic mutations, imaging fields of up to 20 transgenic embryos simultaneously. This scheme allows spatially and temporally detailed phenotypic characterization of several recessive mutant individuals alongside normal siblings, all within a single 4D recording (P. Heid, W. Raich, J. Hardin, personal communication).
MPLSM has begun to afford embryologists with opportunities to study the behavior of single molecules over significant stretches of time within the context of whole living animals. The initial success we have had with two fluorescent markers in C. elegans embryos makes us optimistic about extension of this system to the study of other organelles, molecules and signals in development through the use of alternative fluorescent probes. Understanding of physiological processes can be greatly enhanced by retrospective analysis of animated 4D recordings.
We thank D. Wokosin for construction, maintenance, and tuning of the microscopes and K. Elicieri for computing assistance. This work was supported by NIH grants GM52454 to J.G.W. and RR00570 to the Integrated Microscopy Resource, and by NIH fellowship GM18200-02 to W.A.M.
References and Links
2. V. E. Centonze, D. L. Wokosin, and J. G. White, “Improved deep optical sectioning capabilities rendered by 2-photon excitation imaging,” Mol. Biol. of Cell 6113a (1995).
6. A. Fire, “A four-dimensional digital image archiving system for cell lineage tracing and retrospective embryology,” Comput. Appl. Biosci. 10443–447 (1994). [PubMed]
7. C. Thomas, P. DeVries, J. Hardin, and J. White, “Four-dimensional imaging: computer visualization of 3D movements in living specimens,” Science 273603–607 (1996). http://www.bocklabs.wisc.edu/imr/facility/4D/4d.htm [CrossRef]
8. D. L. Wokosin, V. E. Centonze, J. G. White, S. N. Hird, S. Sepsenwol, G. P. A. Malcolm, G. T. Maker, and A. I. Ferguson, “Multiple-photon excitation imaging with an all-solid-state laser,” Proc. of Optical Diagnostics of Living Cells and Biofluids, SPIE 267838–49 (1996).
9. W. A. Mohler and J. G. White, “Stereo-4-D reconstruction and animation from living fluorescent specimens,” BioTechniques 241006–1012 (1998).http://www.bocklabs.wisc.edu/imr/stereo4d/stereo4d.html [PubMed]
10. W. A. Mohler, J. S. Simske, E. M. Williams-Masson, J. D. Hardin, and J. G. White, “Dynamics and ultrastructure of developmental cell fusions in the Caenorhabditis elegans hypodermis,” Curr. Biol. 81087–1090 (1998). View supplementary movie files at: http://current-biology.com/supmat/cub/bb8s53s1.movhttp://current-biology.com/supmat/cub/bb8s53s2.mov [CrossRef]
11. R. H. Waterston, “Muscle” in The Nematode Caenorhabditis elegans, W. B. Wood, ed. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1988).