Imaging the morphogenesis of insect embryos with light sheet-based fluorescence microscopy has become state of the art. This protocol outlines and compares three mounting techniques appropriate for Tribolium castaneum embryos, introduces two novel custom-made transgenic lines well suited for live imaging, discusses essential quality controls and indicates current experimental limitations.
The red flour beetle Tribolium castaneum has become an important insect model organism in developmental genetics and evolutionary developmental biology. The observation of Tribolium embryos with light sheet-based fluorescence microscopy has multiple advantages over conventional widefield and confocal fluorescence microscopy. Due to the unique properties of a light sheet-based microscope, three dimensional images of living specimens can be recorded with high signal-to-noise ratios and significantly reduced photo-bleaching as well as photo-toxicity along multiple directions over periods that last several days. With more than four years of methodological development and a continuous increase of data, the time seems appropriate to establish standard operating procedures for the usage of light sheet technology in the Tribolium community as well as in the insect community at large. This protocol describes three mounting techniques suitable for different purposes, presents two novel custom-made transgenic Tribolium lines appropriate for long-term live imaging, suggests five fluorescent dyes to label intracellular structures of fixed embryos and provides information on data post-processing for the timely evaluation of the recorded data. Representative results concentrate on long-term live imaging, optical sectioning and the observation of the same embryo along multiple directions. The respective datasets are provided as a downloadable resource. Finally, the protocol discusses quality controls for live imaging assays, current limitations and the applicability of the outlined procedures to other insect species.
This protocol is primarily intended for developmental biologists who seek imaging solutions that outperform standard laboratory equipment. It promotes the continuous attempt to close the gap between the technically orientated laboratories/communities, which develop and refine microscopy methodologically, and the life science laboratories/communities, which require ‘plug-and-play’ solutions to technical challenges. Furthermore, it supports an axiomatic approach that moves the biological questions into the center of attention.
The red flour beetle Tribolium castaneum, which belongs to the large family of darkling beetles (Tenebrionidae), has a long history within the agricultural and life sciences and is the second best studied model insect model organism after the fruit fly Drosophila melanogaster. During the last four decades, it became a powerful and popular insect model organism in developmental genetics, in evolutionary developmental biology and, during the last twenty years, in embryonic morphogenesis for a variety of reasons:
Drosophila and Tribolium both belong to the Holometabola, but diverged approximately 300 million years ago1,2,3,4. While the embryonic development of Drosophila is commonly considered as highly derived, Tribolium shows a more ancestral mode of development that is found in a considerably larger proportion of insect species5,6,7,8,9. Firstly, Tribolium exhibits non-involuted head development, i.e. its mouthparts and antennae emerge already during embryogenesis10,11,12,13,14,15. Secondly, Tribolium follows the principles of short-germ development, i.e. abdominal segments are added sequentially from a posterior growth zone during germband elongation16,17,18,19. Thirdly, Tribolium develops and later degrades two extra-embryonic membranes i.e. the amnion, which covers the embryo only ventrally, and the serosa, which envelops the embryo completely20,21,22. Both membranes play a crucial morphogenetic23 as well as protective role against microorganisms24,25 and desiccation26. Fourthly, the embryonically developing legs are fully functional during the larval life stage and serve as the primordia for the adult legs during pupal metamorphosis27,28,29,30,31.
Due to their small size and modest demands, cultivation of Tribolium in the laboratory is fairly straightforward. Cultures of wild-type (WT) strains or transgenic lines typically consist of around 100-300 adults and can be kept within one-liter glass bottles (footprint 80 cm2) filled three to four centimeters high (about 50 g) with growth medium that consists of full grain wheat flour supplemented with inactive dry yeast. A water supply is not necessary. This allows even small laboratories to keep dozens of beetle cultures within small- or medium-sized commercially available insect incubators. Later developmental stages of Tribolium (larvae after approximately the fourth instar, pupae and adults) are easily separated from the growth medium by sieving. Synchronized embryos are obtained by incubating adults for short periods on egg-laying medium. For rapid development, beetle cultures are kept at 32 °C (about four weeks per generation), while stock keeping is typically performed at 22-25 °C (about ten weeks per generation).
Within the last decade, many standard techniques have been gradually adapted and optimized for Tribolium, as summarized in the Emerging Model Organisms books32. Of great importance are advanced genetic methods such as embryonic33, larval34,35 or parental36,37 RNA interference-based gene knockdown, germline transformation with either the piggyBac38,39 or the Minos40 transposase system and CRISPR/Cas9-based genome engineering41. Furthermore, the Tribolium genome has been sequenced about a decade ago42, and is now in the third round of genome assembly release43, which allows efficient and genome-wide identification and systematic analysis of genes44 or other genetic elements45,46. Additionally, the genomes of four other coleopteran species are available for comparative genetic approaches47,48,49,50. In association with the sequenced genome, two large-scale genetic analyses have been performed, i.e. an insertional mutagenesis screen51 and a systematic RNA interference-based gene knockdown screen52,53.
Fluorescence live imaging with widefield, confocal or light sheet-based microscopy (LSFM) allows to observe the embryonic morphology of Tribolium as a function of time (i.e. the morphogenesis) in a multi-dimensional context (Table 1). In widefield and confocal fluorescence microscopy, the excitation and emission light is guided through the same objective lens. In both approaches, the entire specimen is illuminated for every recorded two-dimensional plane. Hence, the specimens are subjected to very high energy levels. In LSFM, only the fluorophores in the focal plane are excited due to a decoupling of illumination and detection by using two perpendicularly arranged objective lenses (Figure 1). LSFM comes in two canonic implementations – the single plane illumination microscope (SPIM) and the digital scanned laser light sheet-based fluorescence microscope (DSLM, Figure 2) – and offers several crucial advantages over traditional approaches: (i) intrinsic optical sectioning capability, (ii) good axial resolution, (iii) strongly reduced level of photo-bleaching, (iv) very low photo-toxicity, (v) high signal-to-noise ratio, (vi) relatively high acquisition speed, (vii) imaging along multiple directions and (viii) deeper tissue penetration due to the usage of low numerical aperture illumination objective lenses54,55,56.
LSFM has already been successfully applied in Tribolium to document nearly the entire embryonic morphogenesis57 and to analyze the principles of extra-embryonic membrane rupture at the beginning of dorsal closure23. To raise the attractiveness of LSFM in the Tribolium community and for insect science in general, it is of great importance to establish standard operating procedures and to improve the methods, protocols and the pool of resources to a level where the microscope becomes an ease-of-use standard tool in developmental biology laboratories, and the biological questions stay in the center of attention.
This protocol begins with the basics of Tribolium cultivation, i.e. maintenance, reproduction and embryo collection. Next, two experimental strategies are illustrated: (i) live imaging of custom-made transgenic lines and (ii) imaging of fixed embryos that were stained with fluorescent dyes (Table 2). Subsequently, three mounting techniques with slightly different purposes are explained in detail (Figure 3 and Table 3): (i) the agarose column, (ii) the agarose hemisphere and (iii) the novel cobweb holder. The protocol then explains the data acquisition procedure with LSFM. Imaging modalities and key considerations are outlined. Finally, embryo retrieval is explained and suggestions for basic data processing are provided. In the representative results, live imaging data from two novel custom-made and the Glia-blue58 transgenic lines are shown and the respective imaging datasets are provided as a downloadable resource. Additionally, image data of fixed embryos that were stained with a variety of fluorescence dyes are presented. The discussion focuses on quality control, current limitations of the live imaging approach and the adaptation of the protocol to other species.
The protocol is written for light sheet-based fluorescence microscopes that are equipped with a sample chamber and a rotatable clamp mechanism for standardized sample holders54,59,60, which are typically cylinder-shaped elements made of metal, plastic or glass with a diameter in the millimeter range. The protocol is also suitable for both canonic implementations, i.e. SPIM and DSLM, as well as for setups with two or more illumination and detection arms61,62,63. The representative results show data in two spectral channels, green (illumination with a 488 nm laser, detection through a 525/50 bandpass filter) and red (illumination with a 561 nm laser, detection through a 607/70 bandpass filter), but the protocol can be expanded to three or four spectral channels.
1. Husbandry of Tribolium Cultures
NOTE: Standard conditions are defined as an incubation temperature of 25 °C and 70% relative humidity in a 12 h bright/12 h dark cycle. For more information on Tribolium husbandry, respective guidelines are available64. This protocol requires two different flour-based media, which can be prepared in kilogram quantities and stored for several months.
2. Microscope Setup and Calibration
NOTE: The Tribolium embryo has an anterior-posterior length of about 600-660 µm and a transversal diameter of about 300-330 µm. Most modern CCD cameras have a sensor chip size of about 9 mm x 7 mm, i.e. a diameter of 11.4 µm, and a pixel-pixel distance of 6.45 µm. When combined with a 10X magnification objective lens, the embryo can be imaged in toto with a pixel pitch of 0.645 µm. Most modern sCMOS cameras have a larger sensor chip size of about 16 mm x 14 mm, i.e. a diameter of 21.3 µm, and a pixel-pixel distance of 6.5 µm. When combined with a 20X magnification objective lens, the embryo can be imaged in toto with a pixel pitch of 0.325 µm.
3. Collection of Transgenic or WT Embryos
4. Embryo Dechorionation
NOTE: The chorion is the outer layer of the eggshell and consists of proteins and polysaccharides. It is strongly light absorbent but not essential for proper development and can thus be chemically removed.
5. Vitelline Membrane Permeabilization, Fixation, and Staining
NOTE: The vitelline membrane is the inner layer of the eggshell. It is impervious for fluorescent dyes and thus has to be chemically permeabilized before staining.
6. Preparation of Mounting Agarose
7. Embryo Mounting and Insertion into the Sample Chamber
8. Light Sheet-based Fluorescence Microscopy
9. Embryo Retrieval and Quality Control
10. Image Data Processing
NOTE: For image data processing, the open-source image processing software ImageJ67 or its derivate FIJI68 are recommended. Both versions as well as comprehensive documentation are found at www.imagej.net. The standard file format for LSFM data is the tagged image file format (TIFF). The intrinsic container option allows the storage of image stacks such as z-stacks or time series of z maximum projections within a single file that can be opened in ImageJ or FIJI via drag-and-drop.
This protocol describes an experimental framework for fluorescence imaging of living or fixed and stained Tribolium embryos with LSFM. Due to the low levels of photo-bleaching and photo-toxicity, a direct consequence of its optical sectioning capability, LSFM is particularly well suited for long-term live imaging.
The novel AGOC{ATub'H2B-mEmerald} #1 transgenic line expresses a histone2B-mEmerald fusion protein under control of the alpha-tubulin 1 promoter74. It shows an enhancer trap-like expression pattern51, since the fluorescence signal is mainly obtained from the serosa, the yolk sac and several neuronal cell clusters. Using this line, 66 hr of embryonic development at room temperature (23±1 °C) were recorded, covering germband retraction and dorsal closure (Supplementary Movie 1). This line can be used to visualize neuronal clusters within the head appendages (Figure 4A) and the dynamics of serosa migration during dorsal closure (Figure 4B). The novel AGOC{ATub'H2B-mEmerald} #4 line carries the same transgene as the #4 line, but at a different genomic location. It does not show an obvious enhancer trap pattern, but the fluorescence signal is spatiotemporally ubiquitously detectable as described previously for this promoter74. This line was imaged at the transition from gastrulation to germband elongation to demonstrate the optical sectioning capability at two depth levels (Figure 4C). Another feature, especially for live imaging, is the acquisition of z-stacks along multiple directions via sample rotation, which is standard for LSFM setups used in developmental biology. For example, in the Glia-blue58 line, sample rotation allows the observation of glia cell reorganization along and around the ventral nerve cord as well as the proliferation dynamics in the left head lobe, which can only be seen properly from the dorsal site, in the same embryo (Figure 4D, Supplementary Movie 2). The live imaging datasets associated with this study are provided as a downloadable resource, meta and access information are found in Supplementary Table 1.
Since the choice of transgenic lines for live imaging is still limited, small fluorescent dyes can be used to specifically label intracellular structures. SYTOX Green binds to cell nuclei and can be visualized in the green channel, while YOYO-1 and BOBO-3 Iodide bind to the nuclear envelope and can be visualized in the green or red channel, respectively (Figure 5A). These dyes can be used to highlight certain embryonic structures, such as the serosa scar (Figure 5B, first column), the posterior ventral serosa cells (Figure 5B, second column) or the serosa-amnion-germband tissue tri-layer that emerges during serosa window closure (Figure 5B, third to fifth column). Dual-color staining allows the visualization of two intracellular structures in the same embryo, for example the nuclear envelope together with the actin cytoskeleton, which can be stained with Alexa Fluor-conjugated Phalloidin dyes. For example, YOYO-1 Iodide can be combined with Phalloidin 546 (Figure 5C), while BOBO-3 can be combined with Phalloidin 488 (Figure 5D).
It is also convenient to fix and stain transgenic embryos that already express a certain fluorescent protein, since the fixation procedure quenches the intrinsic fluorescence signal only marginally. For example, embryos from the AGOC{ATub'H2B-mEmerald} #4 transgenic line, in which the nuclei are detected in the green channel, can be stained with Phalloidin 546 so that the actin cytoskeleton becomes visible in the red channel (Figure 6).
Figure 1: Comparison of Conventional Widefield, Confocal Laser Scanning and Light Sheet-based Fluorescence Microscopy. Conventional widefield epi-fluorescence microscopy uses xenon arc/mercury-vapor lamps with appropriate filters or high-power light emitting diodes as light sources. For each acquired two-dimensional image, the whole specimen is illuminated with a non-diffraction-limited light cone. In confocal laser scanning fluorescence microscopy, a diffraction-limited Gaussian laser beam is scanned through the specimen and fluorescence is detected incoherently point-by-point. Similarly to conventional widefield epi-fluorescence microscopy, the whole specimen is illuminated for each acquired two-dimensional image. In light sheet-based fluorescence microscopy, a diffraction-limited Gaussian laser beam illuminates the specimen perpendicularly to the detection axis. Fluorescence is detected either plane-by-plane or line-by-line. During the acquisition of a two-dimensional image, only a small volume centered on the focal plane of the detection objective experiences the effects of the excitation light. The remaining volume of the specimen is not illuminated, does not contribute to out-of-focus blur, does not suffer from photo-toxic effects and does not lose fluorophores due to photo-bleaching. Please click here to view a larger version of this figure.
Figure 2: The Principle of Light Sheet-based Fluorescence Microscopy. In LSFM, illumination and detection are split into at least two optical paths. At least one illumination arm is used to generate the light sheet (cyan), while at least one detection arm is equipped with an appropriate filter and a camera for the parallel detection of the fluorescence signal (green). LSFM has two canonic implementations, the single (or selective) plane illumination microscope (blue background) and the digital scanned laser light sheet microscope (red background). By moving the specimen and the light sheet relative to each other, three-dimensional images are acquired. Information along multiple directions is collected by rotating the specimen around the y-axis, which is preferentially oriented along gravity. Please click here to view a larger version of this figure.
Figure 3: Three Different Mounting Techniques used in the Recording of the Embryonic Development of Tribolium with Light Sheet-based Fluorescence Microscopy. (A) The agarose column refers to a steel cylinder with a small pin that pushes the agarose column, in which the embryo is embedded, out of a glass capillary. The agarose hemisphere is mounted on a steel pipe. The embryo is attached to the pole of the hemisphere with a small amount of agarose. The cobweb holder is a sheet of metal with a slotted hole mounted on top of the steel cylinder. The embryo is glued to a thin agarose film that spans the slotted hole. (B) Schemes of the three mounting techniques inside a light sheet microscope. (C) The three mounting techniques applied within a DSLM. (D) Exemplary images from embryos of the Glia-blue transgenic line recorded with the three mounting techniques. The embryos, which are at the onset of germband retraction, are shown with their posterior end orientated towards top to match the transmission light images. FM, fluorescence mode; ZA, z maximum projection with intensity adjustment. Please click here to view a larger version of this figure.
Figure 4: Live Imaging of Tribolium Embryos. (A) An embryo of the AGOC{ATub'H2B-mEmerald} #1 line at the transition from germband retraction to dorsal closure. This line exhibits an enhancer trap expression pattern. The fluorescence signal is mainly found in the serosa, the yolk sac and a few neuronal cell clusters. The embryo is shown over a period of 15 h with an interval of 5 h. The detail images show cell clusters in the head appendages. Initially, the clusters are still covered by serosal cells (first column), upon serosa rupture, the cells follow along with the turning movement of the head. (B) Same embryo shown during dorsal closure for 1:30 h with an interval of 0:30 h. The detail images show the migration of the ruptured serosa over the yolk sac. (C) Top row: An embryo of the AGOC{ATub'H2B-mEmerald} #4 line at the transition from gastrulation to germband elongation. Middle and lower row: Single planes are shown in two depths over a period of 10:30 h with an interval of 1:30 h. (D) An embryo of the Glia-blue line during germband retraction shown over a period of 50:30 h with an interval of 07:30 h. The detail images show the morphogenetic reorganization of the glial cells in the left head lobe. ZA, z maximum projection with intensity adjustment; PA single plane with intensity adjustment. Please click here to view a larger version of this figure.
Figure 5: Fixation, Staining, and Iimaging of WT Embryos during Germband Elongation. (A) Embryos stained with either SYTOX Green, i.e. a fuzzy nuclear stain, or YOYO-1 Iodide or BOBO-3 Iodide, i.e. two nuclear envelope stains. (B) YOYO-1 Iodide-stained embryo. Details show the serosa scar (first column), the nuclear envelope of the large posterior-ventral serosa cells (second column) or the three-layered tissue organization of the serosa, the amnion and the actual embryonic tissue (third to fifth column). (C) Dual-color staining with YOYO-1 Iodide and Phalloidin 546. (D) Dual-color staining with Phalloidin 488 and BOBO-3 Iodide. ZA, z maximum projection with intensity adjustment. Please click here to view a larger version of this figure.
Figure 6: Images of Fixed and Stained Transgenic Embryos from the AGOC{ATub'H2B-mEmerald} #1 line. This transgenic line expresses mEmerald-labeled histone2B (H2B) under control of the alpha-tubulin 1 promoter, which marks the nuclei/chromatin ubiquitously throughout the whole embryonic development. The embryo was further stained with Alexa Fluor 546 Phalloidin, which labels the actin cytoskeleton/f-actin. ZA, z maximum projection with intensity adjustment. Please click here to view a larger version of this figure.
criterion | fluorescence microscope type | ||
conventional wide field | confocal laser scanning | light sheet-based | |
objective lens setup | one objective lens, medium-high NA | one objective lens, high NA | two perpendicularly arranged objective lenses, illumination: low NA, detection: high NA |
illumination light source | white light source and appropriate filters (e.g. mercury short-arc lamps) or LED-based with clean-up filters |
laser with clean-up filter (e.g. diode or DPPS lasers) |
laser with clean-up filter (e.g. diode or DPPS lasers) |
illumination per two-dimensional image | whole sample | whole sample (typically two-dimensional Gaussian beam scan) |
only focal plane (DSLM: one-dimensional Gaussian beam scan) |
detection | parallel (typically CCD or sCMOS camera) |
sequential (photomultiplier tube) |
parallel (typically CCD or sCMOS camera) |
imaging speed | fast (milliseconds per image) | slow (seconds per image) | fast (milliseconds per image) |
out-of-focus fluorescence | no discrimination | blocked nearly completely by the pinhole (rejection efficiency depends on diameter) |
nearly non-existent (only due to scattering) |
spatial dimensions | 2 (x / y) |
3 (x / y / z) |
3 (x / y / z) |
further degrees of freedom | 2 (time, fluorescence channel) |
2 (time, fluorescence channel) |
3 (time, fluorescence channel, direction) |
lateral resolution | r (0.6 λ / NA) |
0.66 r (0.4 λ / NA) |
r (0.6 λ / NA) |
optical sectioning capability | No | Yes (discrimination in the detection pathway) |
Yes (discrimination in the illumination pathway75) |
availability and price | predominantly commercial, low cost | predominantly commercial, expensive | commercial, expensive custom-built, moderate54,59,60,75 |
live imaging publications covering Tribolium embryos | six44,76,77,78,79,80 | seven23,77,79,81,82,81 (spinning disk confocal)83,84,85 |
four23,57,65,86 |
Table 1 – Characteristics of Fluorescence Microscopy Techniques used in Tribolium live Imaging.
dye | staining | dilution |
SYTOX Green | nuclei (fuzzy) | 1:10,000 (1:100 in ddH2O and 1:100 in 1% (w/v) BSA in PBS) |
YOYO-1 Iodide | nuclear envelope | 1:10,000 (1:100 in ddH2O and 1:100 in 1% (w/v) BSA in PBS) |
BOBO-3 Iodide | nuclear envelope | 1:10,000 (1:100 in ddH2O and 1:100 in 1% (w/v) BSA in PBS) |
Alexa Fluor 488 Phalloidin | actin cytoskeleton | 1:100 (in 1% (w/v) BSA in PBS) |
Alexa Fluor 546 Phalloidin | actin cytoskeleton | 1:100 (in 1% (w/v) BSA in PBS) |
Table 2 – Staining Solutions.
agarose column (option 1) |
hemisphere (option 2) |
cobweb holder (option 3) |
|
rationale | embryo is embedded in an agarose column that is pushed out of a capillary | posterior end of embryo is glued to pole of an agarose hemisphere | embryo is glued to thin agarose film spanning a slotted hole |
sample holder | steel cylinder with pin | steel pipe | steel cylinder with slotted hole |
directions (orientations) | unlimited (any) | unlimited (any) | four (0°, 90°, 180°, 270°) |
required skill level | moderate | high | low |
agarose around the embryo | high | low | moderate |
number of embryos per session | up to three | one | up to six |
embryo retrieval | tricky | easy | easy |
disposable equipment | glass capillaries | – | – |
recommended for | short-term live imaging, stained embryos | long-term live imaging | long-term live imaging, stained embryos |
references | three23,87,88 | two57,65 | this publication |
Table 3 – Mounting Methods Advantages and Disadvantages.
aberrational phenotype induction | rationale | step | reference (methodological) |
embryonic RNA interference | double-stranded RNA is injected into embryos during the syncytial blastoderm stage to knock-down one or more specific genes | inject embryos after Step 4.8. | two33,81 |
parental RNA interference | double-stranded RNA is injected laterally into female pupae between the abdominal segments 3 and 4, which leads to gene knock-down in the progeny | take the pupae from Step 1.5. | two36,37 |
recessive embryonic lethal mutant lines | egg-laying cultures that carry a recessive embryonic lethal mutation heterozygous yield about 25% homozygote mutant progeny | cross mutant line into imaging line during Step 1.5. | three39,51,89 |
application of extrinsic factors | certain bioactive or toxic factors are extrinsically applied to the embryos by adding them to the imaging buffer | add factor to imaging buffer during Step 2.2. | two83,85 |
Table 4 – LSFM Live Imaging of Aberrational Phenotypes
Supplementary Table 1 – Metadata and Parameter for the Long-term Live-imaging Datasets DS0001-0003. Please click here to download the file.
Supplementary Movie 1 – Live Imaging of a Tribolium Embryo from the AGOC{ATub'H2B-mEmerald} #1 Transgenic Line.
This line exhibits an enhancer trap expression pattern. The fluorescence signal is primarily found in the serosa, the yolk sac and a subset of neuronal cells. The embryo is shown along four orientations from 00:00 h to 66:00 h with an interval of 00:30 h between the time points. The movie starts at the beginning of germband retraction and ends once dorsal closure is completed. Frame rate is five frames per second. ZA, z maximum projection with intensity adjustment. Please click here to download the file.
Supplementary Movie 2 – Live Imaging of a Tribolium Embryo from the Glia-blue Transgenic Line.
The embryo is shown along four orientations from 00:00 h to 66:00 h with an interval of 00:30 h between the time points. The movie starts at the beginning of germband retraction and ends once dorsal closure is completed. Frame rate is five frames per second. ZA, z maximum projection with intensity adjustment. Please click here to download the file.
Quality control
In live imaging assays, the preparation and recording procedure must be non-invasive, i.e. neither the mechanical and chemical handling (collection, dechorionation, mounting onto the sample holder) nor the integrated energy load during the observation should affect the viability of the specimen. For studies that characterize WT development, it is recommended to only use data from experiments in which the embryo survives the recording process, is retrieved successfully and develops into a healthy adult. The adult should not show any morphological aberrations, should be fertile and its progeny should also be fertile. There are three major reasons for low survival rates, especially in long-term live imaging assays:
Firstly, the embryo surface was covered extensively with agarose during Step 7, which probably hinders ion and gas exchange and constricts the embryo mechanically. Affected embryos, especially when imaged during early embryogenesis, typically develop inconspicuously for several hours, arrest suddenly and lose their fluorescence signal quickly. With some experience, the surface fraction that is embedded in agarose by using the hemisphere technique can be kept below one-third, while with the cobweb holder technique, at most half of the surface is covered by a very thin agarose layer. Secondly, the integrated energy load applied to the embryo exceeds the tolerance. The values given in Step 8.2 serve as a rule of thumb, but the following criteria should be considered: (i) embryos during early embryonic development do not endure laser irradiation as well as embryos during late embryonic development, (ii) although the integrated energy load might be identical, higher laser power with either lower exposure time or longer temporal intervals appears to be more stressful than lower laser power with a high exposure time or shorter temporal intervals, (iii) higher wavelengths are typically better tolerated and (iv) for transgenic lines, the tolerance limit appears to be line-specific, whereas fluorophore expression strength seems to be inversely proportional to the laser irradiation tolerance. Additionally, the design of the fusion protein and/or the intracellular localization also play a role. Thirdly, the embryo is damaged during the mounting process. The most crucial factor here is that the incubation time in the sodium hypochlorite solution is as short as necessary to completely remove the chorion. Longer incubation times might harm the embryos. Care should also be taken to not puncture or squeeze the embryos with the brush and to keep the mounting agarose at an ambient temperature. Damaged embryos can be identified during Step 8.4 before the actual imaging process commences, so that damaged embryos can be replaced with viable ones.
By following those guidelines, typically around 85-90% of the embryos survive the live imaging procedure and hatch. Around 90% of the hatched embryos develop into healthy and fertile adults. The survival ratio appears to be independent from imaging temperature and mounting techniques (when the agarose column is used for imaging periods of not more than half a day), but slightly decreases with elevated levels of laser irradiance. As an additional control, especially during the initial characterization of custom-made transgenic lines for fluorescence live imaging or the framework establishing phase as precursor for a comprehensive experimental sequence, it is recommended to run non-imaged WT and transgenic embryos in parallel to the imaged embryos, which are incubated at the same temperature as the imaged embryo, and compare their development time and morphology65.
Furthermore, studies that characterize aberrational phenotypes (Table 4) via live imaging should also run a non-affected control embryo in parallel. It is recommended to keep the treatment of both embryos as similar as possible, but this has to be considered individually for each experimental strategy:
Knock-down via RNA interference.
For embryonic RNA interference33,81, embryos that derive from the same egg-laying culture should either be injected with functional or control double-stranded RNA. The differently injected embryos should then be imaged simultaneously in the same LSFM by using either the agarose column or the cobweb holder technique. For parental RNA interference36,37, the imaging culture should be split into two subcultures. Female pupae of one subculture should be injected with the functional double-stranded RNA, while the other subculture should be injected with control double-stranded RNA. Embryo collection should be performed in parallel, and both the experimental and the control embryo can be imaged simultaneously in the same microscope by using the agarose column or cobweb holder technique.
Recessive embryonic lethal mutant lines.
When working with recessive embryonic lethal mutants39,89,91,92,93, an egg-laying culture of heterozygous mutants, which are normally phenotypically inconspicuous, yields theoretically 25% homozygous mutant progeny (but sometimes less due to practical handling issues89) and 75% of phenotypically inconspicuous progeny (50% heterozygous mutants and 25% WTs). It is recommended to use the cobweb holder technique to mount several embryos from such an egg-laying culture. The homozygous mutants can be identified throughout the course of imaging by the manifestation of the phenotype, while the normally developing embryos serve as controls. The genotype of the imaged specimens can be determined once imaging and quality control are completed.
Application of extrinsic factors.
If the influence of extrinsic factors within the imaging buffer83,85, e.g. bioactive or toxic substances, on development is investigated, parallel imaging together with a control embryo in the same LSFM is typically not possible. Therefore, imaging has to be performed sequentially. Optimally, the time between subsequent experiments is minimized and the embryos derive from the same imaging culture. Alternatively, two identically constructed and operated LSFMs can be used. In this case, embryos from the same egg-laying period should be used, but it is of great importance that the calibration of both microscope is conducted carefully so that the imaging properties are comparable. By using the cobweb holder technique, multiple embryos for each condition can be imaged at once. With the agarose column technique, the extrinsic factors might be hindered from reaching the embryos.
Current limitations of the LSFM live imaging approach
LSFM is a powerful tool to analyze embryonic morphogenesis non-invasively in multiple broad-scaled dimensions. Standard operation procedures, as proposed in this protocol, will support the process to establish light sheet technology as a standard tool in developmental biology. However, the approach is limited by a number of factors on diverse experimental and organizational levels:
Initially, LSFM was developed to analyze one specimen at a time. All current approaches that image two or more embryos simultaneously in the same microscope by 'stacking' the specimen along the y-axis in the agarose column or the cobweb holder addresses this issue only to a certain degree. Multi-well plate- and coverslip-based setups are available94,95,96. However, they suffer from reduced image quality and sacrifice some of the essential benefits from LSFM, such as imaging along multiple directions. Currently, the best option is to work with multiple microscopes in parallel, which becomes reasonably feasible when costly components, such as the laser, are shared60.
With the EFA-nGFP57,65,84, the FNL86, the Brainy58,86, the AGOC{ATub'H2B-mEmerald} #1, the AGOC{ATub'H2B-mEmerald} #4 and the Glia-blue58, only six custom-made transgenic lines suitable for long-term live imaging are currently available. In addition, the HC079 (amnion), G12424 (serosa) and G04609 (heart) enhancer trap lines51 have been characterized with LSFM23. Alternatively, fluorescent embryos can also be generated by mRNA81 or dye injection79, those approaches are relatively straightforward, but also suffer from limitations86. Several more standards are still required such as cytoskeletal-, organelle- and membrane-labeled lines or lines that express fluorophores under certain non-ubiquitous promoters to observe only certain organs or tissues. Ideally, those lines are also available with different fluorescent proteins.
Another challenge is the data volume generated by LSFM. Due to acquisition of information in three spatial dimensions and at least three further degrees of freedom (time, direction, fluorescence channel), the size of live imaging datasets can easily reach many Gigabytes to a few Terabytes. Even after three-dimensional cropping and ZIP-based compression of the TIFF containers, the three exemplary datasets associated with this study have sizes of 31.6, 12.6 and 45.1 Gigabytes, i.e. 89.3 Gigabyte in total. When working with LSFM, it is necessary to have the respective infrastructure for data storage and processing88,97.
The highest image quality is obtained at the surface of the embryo, but with increasing depth, the fluorescence signal becomes more and more blurry. For example, the posterior end of the germband becomes covered by the yolk sac fold during gastrulation and cannot be properly resolved (Figure 4C, first to fifth column). Imaging along multiple directions solves this problem at least partially – high quality information from the surface all around the embryo is available, but the inner regions remain fuzzy. Currently, no satisfying solutions are available, but in the long run, infrared fluorescent proteins98, genetic approaches99 or microscope setups with adaptive optics100 could be considered.
Adaptation for other species
By now, LSFM has been used to document the embryonic development of three insect species, i.e. the fruit fly Drosophila melanogaster61,62,101, the scuttle fly Megaselia abdita102 and the red flour beetle Tribolium castaneum57. The experimental framework, standard operating procedures and mounting techniques described here should be easily adaptable to other insect species. Germline transformation protocols for many insects that belong to various orders are available103, including species of economical and/or medical importance, such as the honeybee104, several lepidopterans105,106 or multiple mosquito species107,108,109. With respective transgenic lines suitable for fluorescence live imaging, the embryonic morphogenesis of insects can be characterized while considering their evolutionary lineages and/or ecological niches. About one million insect species have been described and a further twenty million insects species most likely exist110, covering more than 400 million years of evolution2. Only the comparative approach will generate information that in turn provides insights which cannot be obtained by studying the developmental principles of only a single species.
The authors have nothing to disclose.
We thank Sven Plath for technical support. The Glia-blue transgenic line was a kind gift from Gregor Bucher (Göttingen, Germany). The research was funded by the Cluster of Excellence Frankfurt am Main for Macromolecular Complexes (CEF-MC, EXC 115, speaker Volker Dötsch) granted in part to EHKS at the Buchmann Institute for Molecular Life Sciences (BMLS, director Enrico Schleiff) at the Goethe Universität Frankfurt am Main by the Deutsche Forschungsgemeinschaft (DFG).
full grain wheat flour | Demeter e.V. | 1.13E+08 | US: whole wheat flour, UK: whole meal flour |
405 fine wheat flour | Demeter e.V. | 1.13E+08 | US: pastry flour, UK: soft flour |
inactive dry yeast | Flystuff / Genesee Scientific | 62-106 | |
phosphate-buffered saline (PBS), pH 7.4 | Thermo Fisher Scientific | 10010-023 | |
sodium hypochlorite, ~12% active Cl | Sigma Aldrich | 425044-250ML | Caution: sodium hypochlorite is corrosive |
low-melt agarose | Carl Roth | 6351.2 | |
6-well plate | Orange Scientific | 4430500 | |
24-well plate | Orange Scientific | 4430300 | |
glass capillaries, internal Ø 0.46 mm | Brand GmbH + Co KG | 7087 09 | |
SYTOX Green | Thermo Fisher Scientific | 57020 | Staining solution preparation is explained in Table 2 |
YOYO-1 Iodide | Thermo Fisher Scientific | Y3601 | Staining solution preparation is explained in Table 2 |
BOBO-3 Iodide | Thermo Fisher Scientific | B3586 | Staining solution preparation is explained in Table 2 |
Alexa Fluor 488 Phalloidin | Thermo Fisher Scientific | A12379 | Staining solution preparation is explained in Table 2 |
Alexa Fluor 546 Phalloidin | Thermo Fisher Scientific | A22283 | Staining solution preparation is explained in Table 2 |
sieve, 800 µm mesh size | VWR International | 200.025.222-051 | |
sieve, 710 µm mesh size | VWR International | 200.025.222-050 | for growth medium preparation (step 1.1) |
sieve, 300 µm mesh size | VWR International | 200.025.222-040 | |
sieve, 250 µm mesh size | VWR International | 200.025.222-038 | for egg laying medium preparation (step 1.2) |
glass dish, Ø 100 mm × 20 mm | Sigma Aldrich | CLS70165102 | |
cell strainer, 100 µm mesh size | BD Biosciences | 352360 | |
paint brush, head Ø 2 mm | VWR International | 149-2121 | |
syringe, 1.0 ml | B. Braun Medical AG | 9166017V | |
scintillation vials | Sigma Aldrich | M1152-1000EA | |
paraformaldehyde | Sigma Aldrich | 158127 | Caution: paraformaldehyde is toxic and corrosive |
n-heptane ≥ 99% | Carl Roth | 8654.1 | Caution: n-heptane is flammable and toxic |
Triton X-100 | Sigma Aldrich | X100-100ML | Caution: Trition X-100 is corrosive |