A practical staging atlas to study embryonic development of Octopus vulgaris under controlled laboratory conditions

Background Octopus vulgaris has been an iconic cephalopod species for neurobiology research as well as for cephalopod aquaculture. It is one of the most intelligent and well-studied invertebrates, possessing both long- and short-term memory and the striking ability to perform complex cognitive tasks. Nevertheless, how the common octopus developed these uncommon features remains enigmatic. O. vulgaris females spawn thousands of small eggs and remain with their clutch during their entire development, cleaning, venting and protecting the eggs. In fact, eggs incubated without females usually do not develop normally, mainly due to biological contamination (fungi, bacteria, etc.). This high level of parental care might have hampered laboratory research on the embryonic development of this intriguing cephalopod. Results Here, we present a completely parameter-controlled artificial seawater standalone egg incubation system that replaces maternal care and allows successful embryonic development of a small-egged octopus species until hatching in a laboratory environment. We also provide a practical and detailed staging atlas based on bright-field and light sheet fluorescence microscopy imaging for precise monitoring of embryonic development. The atlas has a comparative section to benchmark stages to the different scales published by Naef (1928), Arnold (1965) and Boletzky (2016). Finally, we provide methods to monitor health and wellbeing of embryos during organogenesis. Conclusion Besides introducing the study of O. vulgaris embryonic development to a wider community, this work can be a high-quality reference for comparative evolutionary developmental biology.

out into a stalk and in octopods, many stalks are interwoven and glued together with material secreted by the female oviducal glands to form a string or festoon (Fig. 1a) [4,5]. Octopuses that lay eggs that hatch out as planktonic paralarvae generally produce thousands of small eggs, reaching 500,000 in O. vulgaris [6]. Fertilization is achieved during spawning whereafter the string is attached to a substrate in the den [3,6]. During embryonic development, cephalopod eggs generally increase in volume, although this phenomenon is more pronounced in decabrachian eggs compared to octopod eggs [7]. In O. vulgaris eggs, this swelling process affects egg width and wet weight whereas length is nearly unaffected [8].
The embryonic development of cephalopods can roughly be separated in three periods. The first one includes maturation and fertilization of the oocyte, discoidal meroblastic cleavage to form the blastodisc and division to complete the blastoderm. The gastrulation or second period comprises the formation of the germinal layers with establishment of endoderm and extraembryonic yolk epithelium and the start of epiboly followed by concentrations of mesoderm. The organogenesis or third period begins with an elevation of blastodisc folds that prelude the appearance of the first organ primordia that will give rise to the typical dibranchiate topology and then, linear growth will eventually form a fully developed hatchling [2]. The last stages of development (maturation) are more difficult to compare between cephalopods, since species that produce large eggs generally hatch out as juveniles that are miniature adults, while small egg-embryos hatch out as small planktonic paralarvae. The latter still have to go through major morphological changes to attain the juvenile form, such as the development of the arm-crown complex, swimming control, the chromatophore system and horizontal pupillary response [9][10][11]. Furthermore, taxon specific features that arise in cuttlefish (e.g. cuttlebone) or squid (e.g. tentacles) embryos are absent from octopus and thus not discussed here.
Octopuses (e.g. Octopus, Eledone and Tremoctopus) undergo double reversion during embryonic development [12,13]. The first reversion or blastokinesis takes place at Stage VII in O. vulgaris, when the extraembryonic yolk epithelium just completed closure at the vegetative pole and is realized by a change of direction of the ciliary beat of the yolk envelope [12]. In this process, the embryo migrates from the micropyle to the stalk side of the egg, which takes 7 to 36 h depending on water temperature [12,14]. While positioning at the stalk side might protect embryos better from predators and would reduce mechanical stress during organogenesis (Nande, personal communication), failure of turning does not impact embryonic development. The second reversion at Stage XIX then positions the embryo for smooth hatching [12]. The physiological and morphological factors that trigger hatching in cephalopods are still unknown [5,15], but hatching starts with stretching mantle movements that rupture the apex of cells in the hatching gland or organ of Hoyle at the dorsal tip on the mantle [16,17]. These glandular cells store proteolytic Fig. 1 Octopus vulgaris eggs and the embryonic morphological body axes. a A string of O. vulgaris eggs. Scale bar represents 500 μm. b The morphological axes in cephalopod embryos correspond to the axes in other mollusks. In this orientation, the location of the funnel is posterior, the embryonic mouth is anterior, the arms are ventral and the mantle is dorsal enzymes that dissolve the chorion locally, making the egg integument permeable to water, which increases the osmotic pressure within the perivitelline space [5,[18][19][20]. Afterwards, the mantle is extruded due to a release of pressure and the Kölliker organs (hard bristle-like structures spread over the skin) make sure that the embryo does not slip back into the chorion so it can move freely from the egg during hatching [9,20,21].
Due to breeding season limitations as well as geographical spread, different cephalopod species are being researched around the world. In addition, the release of several cephalopod genomes as well as transcriptomic information over the last years now allows molecular and functional studies on these enigmatic creatures [22][23][24][25][26]. In combination with novel genome editing technologies, this opens interesting opportunities to interrogate in vivo gene function. However, in O. vulgaris, progress in these fields has been hampered by the absence of protocols to maintain egg clutches without maternal care in standardized laboratory conditions. Furthermore, to fully evaluate the impact of genetic change on development, an updated description of embryonic development using modern imaging technologies is valuable. Additionally, there is a need for a standardized, fully-illustrated staging system allowing easy comparison of embryonic development between different cephalopod model species. We acknowledge the inevitable generalization introduced by comparing embryonic stages and refer to species-specific morphological descriptions of S. officinalis, Euprymna scolopes, Todarodes pacificus, Loligo pealei, L. gahi and O. vulgaris [2,[27][28][29][30][31]. Although cephalopod egg and thus hatchling size and consequently the embryonic development duration greatly vary, morphogenetic processes are similar.
We thus provide detailed bright-field and light sheet fluorescence microscopy (LSFM) images of all developmental stages to be used in the laboratory as a staging atlas. Furthermore, this work describes a standardized standalone tank system that should facilitate any laboratory on small-egged cephalopods, regardless of access to fresh seawater. We also supply validation assays for checking the health of embryos at different stages.

Results
The small, yolky eggs of O. vulgaris are roughly 2.5 mm long and 1 mm wide. Octopus embryos are described to develop poorly without maternal care [2,32]. However, we have found that O. vulgaris embryos can develop without maternal care in artificial oxygenated seawater at continuous strong flow rate and dim light. The standalone system ensured a continuous flow in the tanks resulting in an oblique orientation and soft swirling of the strings, likely mimicking the jet flow the mother normally provides (Fig. 2). The embryos developed highly synchronous within the string and hatched after approximately 1 month at 19°C. We provide a summary table with key characteristics of each stage to allow consistent staging of O. vulgaris embryos ( Table 1). As the developmental stages presented by Naef are based on days of development rather than on morphological characteristics and contain considerable gaps in development, we split some events and added '.1' or '.2' in such cases. For all descriptions presented, the morphological axes of the embryo are used (Fig. 1b). According to these axes, the location of the funnel is posterior, the embryonic mouth anterior, the arm crown ventral and the mantle dorsal.

Cleavage, gastrulation and epiboly
The germinal disc is restricted to the animal pole of the egg, at the micropyle side, which is opposite from the stalk. Meroblastic, bilaterally symmetrical cleavage and subsequent formation of the blastodisc takes place over the first 24-48 h after fertilization, depending on water temperature. The first three cleavages are incomplete and generate eight equally sized blastomeres in octopods ( Fig. 3a-d), which differs from decapods where the two dorso-medial cells are more narrow compared to the ventro-medial cells [2]. Further cell proliferation results in the formation of the blastodisc at Stage I ( Fig. 3e). At Stage II, formation of the blastula is completed (Fig. 3f), followed by the onset of epiboly at Stage III, characterized by lateral expansion of the blastoderm over the yolk by cell division (Fig. 3g). The blastodisc, which can be found at the very top of the yolk at Stage II starts to grow and expand over the yolk, generating a cap-like structure by Stage IV (Fig. 3h). At Stage V, a quarter of the yolk is covered by the embryonic cap (Fig. 3i). Using bright-field imaging, the embryo looks uniform at this stage. However, using light sheet microscopy and DAPI as a nuclear stain, the embryo proper with its densely packed nuclei can be clearly distinguished from the extraembryonic ectoderm with larger nuclei spaced further apart ( Fig. 3i-i'). At Stage VI, the germinal disc covers half of the yolk mass ( Fig. 3j-j'). From this stage onwards, the embryo slowly rotates clockwise when observed from the micropyle side of the egg, along its longitudinal axis (Additional file 1 shows a movie of embryo rotation accelerated to 8x original speed at Stage XI) [12,14]. By the end of Stage VII.1, the embryo and yolk envelope (extraembryonic) cover 3/4th of the yolk, followed by complete closure at the vegetative pole, ready for the first reversion (Fig. 3k).

Organogenesis and maturation
At Stage VII.1, the surface of the embryo appears smooth. The first organ primordia can be visualized using DAPI, revealing the prospective arms as patches of dense nuclei close to the yolk envelope ( Fig. 3k-k'). The embryo makes its first reversion at the end of Stage VII. This process takes 7 to 36 h, depending on the incubation temperature [14], in which the embryo migrates over the yolk from the micropyle to the stalk side of the egg and can be observed in different topologies ( Fig. 3mo). At Stage VII.2, primordia become visible by brightfield microscopy as thickenings and depressions that arise from the surface of the embryo (Fig. 3l). The eye placodes, mantle anlage, arm primordia and mouth are the first distinguishable structures (Fig. 3l') and become more discernable towards Stage VIII (Fig. 4), when the mantle rim is elevated.
During the next stages of organogenesis, the organ primordia become more prominent and are clearly distinguishable from the yolk, giving rise to an immature embryo at Stage XVII (Fig. 4, 5, 6 and 7; Additional files 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 and 13 show movies of embryos imaged with LSFM). At Stage IX, the arm buds are clearly separated from one another, the mantle appears more elevated and first yellow pigmentation of the retina is visible. The yolk sac envelope that contains blood lacuna and a network of muscular elements starts to create peristaltic waves of surface contraction at this stage, establishing blood circulation for the early embryo (Additional file 14 shows yolk contraction at Stage XI) [33]. This phenomenon will cease around Stage XVI, when the embryonic heartbeat is well established and when the area of contact between the yolk envelope and the chorion becomes too small [12].
In order to distinguish embryos between Stages IX and XIII, mantle size and the angle relative to the imaginary plane through the eyes, as well as folding of the funnel tube are easily recognizable morphological characteristics (LSFM images in Figs. 4 and 5, funnel in Fig. 6). The shape of the funnel is visible through the chorion, but is easier to observe after dechorionation. At Stage IX, the funnel tube rudiments become visible (Fig.  6a) and fuse at the margins by stage X (white arrow Fig.  6b). At Stage XI, the funnel tube rudiments have grown in size and bend towards the midline (Fig. 6c). Then, at the beginning of Stage XII (Stage XII.1), the funnel starts to form a real tube that is fused at the ventral extremity by Stage XII.2 ( Fig. 6d-e). But, it is at Stage XIII that the formation of the siphon shaped funnel tube is complete (Fig. 6f). In the subsequent events, the position of the mouth on the anterior side changes (Fig. 7, white arrows on LSFM images). The mouth is situated between the first pair of arms on the anterior side from Stage VIII to XIV and is still open to the outside at Stage XV.1. It will start to internalize, becoming encircled by the anterior arms at Stage XV.2. By Stage XVI, the mouth is covered by the arm crown, waiting to take its final position as soon as the outer yolk is reduced.
As the embryos grow, the shape of the mantle goes from depressed towards the middle at Stage VII.2 to flat and perpendicular to the longitudinal axis at Stage X. At Stage XI, the mantle is elevated on the posterior side and thus tilted and clearly grows in size by Stage XII. At Stages XIII and XIV, the length of the mantle equals and exceeds the length of the head in the dorsoventral axis, respectively (Figs. 5, 6 and 7), and at Stage XIV, a heartbeat can be observed at the mantle tip (Additional file 15 shows embryonic heart beat at Stage XVII). From Stage IX to stage XIV, the color of the retina changes from light yellow to dark red/brown. The color of the eye and retina continues to darken during development, Fig. 2 Graphical representation of the standalone system used for egg incubation. a The opaque conocylindrical PVC tank has a water inlet at the top (blue arrow), which creates a circular current and delivers seawater at an exchange rate of 3 L min − 1 . A mechanical filter (1 mm mesh size) is placed at the bottom of the tank in the water outflow. An O. vulgaris egg string is attached to a glass rod and placed on the lateral side of the tank, where it moves constantly by a gentle current generated by the water inflow. The blue arrows indicate the water flow (from top to bottom) and the red horizontal bars indicate the water level. b The standalone system consists of 10 conocylindrical tanks placed on top of a reservoir. Artificial seawater is aerated by the strong water flow pouring into the biological filter in the reservoir and is sterilized by an external UV filter (details provided in Methods section). The total volume in the system is 100 L. The chromatophore pattern (appearance, color and size of chromatophores) is another convenient characteristic to stage O. vulgaris embryos (Figs. 7, 8 and 9). At Stages XV.1 and XV.2, the first chromatophores appear as small yellow dots on the posterior side, next to the funnel and on the mantle, respectively. By Stage XVI, the first chromatophores on the anterior mantle appear. From Stage XVIII.2 onwards, the chromatophores react to changes in light intensity under the microscope (expand under light stimulation and contract in the dark). The ratio of the size of the external yolk sack in relation to the size of the embryo is another measure that can be used for staging (Figs. 7,8 and 9). At Stage XIV, this ratio approximates 1:1 and rapidly decreases to 1:3 at Stage XVI, 1:4 at Stage XVIII.1 to 1:6 at Stage XIX.1. This latter stage is also characterized by the first appearance of ink in the ink sac on the posterior side. The embryo undergoes the second reversion at Stage XIX. We annotate these stages as XIX.1 before and XIX.2 after the second reversion.
At Stages XX.1 and XX.2, the external yolk sack is nearly and completely depleted, respectively (Additional file 16 shows a movie of a Stage XX.2 embryo imaged with LSFM). It has been described that cephalopod embryos are likely slightly sedated in the egg by a tranquillizing factor to prevent premature hatching which can occur at these stages [34]. What precisely induces natural hatching is still unknown, but it is easily triggered by several factors, such as mechanical stimuli, photoperiodicity and sudden changes in light levels or temperature [15]. We observed that natural hatching starts approximately 7 days after the second reversion at 19°C, but is detrimental to the paralarvae in the tank system under continuous flow. Therefore, 7 days post second reversion, we moved the strings from the system to a different tank containing aerated artificial seawater, which induced hatching within minutes.

Assays to evaluate embryonic fitness
Yolk contraction can be observed from Stage IX to Stage XVI under the stereomicroscope and is a valuable readout to evaluate embryonic survival at early organogenesis stages. Furthermore, upon development of the retina, a "saddle" to discoidal shape of the pigmented layer is typical of high-quality embryos. Frowning or folding of the retina points towards poor health. From Stage XIV onwards, a heartbeat can be recognized in the transparent embryos. Occasionally, small crustaceans can be observed on the strings. Generally, these are part of the natural ecosystem of the string and are not impacting embryonic development. Nevertheless, poor rearing conditions (insufficient flow, dissolved oxygen levels and strings floating or sunken) can trigger strings to overgrow with fungi (white thread-like structures or parts turning pale or pink) or get infected by worms. A final readout of state of the art rearing is the hatching of actively swimming paralarvae that display positive phototaxis, reported for most cephalopod hatchlings [9,35,36].

Comparison of staging scales
Since Naefs staging atlas of O. vulgaris is based on the age of embryos in days rather than stage-specific morphological characteristics, Arnold and Lemaire (later adapted by Boletzky) introduced ten extra stages, focusing on the development of L. pealei (officially renamed Doryteuthis pealeii) and S. officinalis, respectively [27,28,37]. These extra stages mostly cover the period of embryo cleavage (e.g. Arnold and Boletzky Stage 9

Discussion
We introduced a low-cost standalone system that runs on artificial seawater for incubating small-egged Octopus species without maternal care. The feasibility and effectiveness of our system was reflected in a highly synchronous development of embryos within the string and in the production of viable hatchlings.

Replacing maternal care
Incirrate octopods and some oceanic squids display parental care during embryonic development [15,38]. As in fertilized eggs [41]. In 1977, Van Heukelem used a glass funnel with filtered seawater to incubate the eggs of O. maya. After adjusting seawater flow, the eggs were maintained slowly tumbling and rubbing against one another in order to keep the egg surface clean and aerated. This author also described that air bubbles interfered with the development of the yolk epithelium and were thus harmful to the embryos [42]. Similarly, our early attempts to incubate egg strings in beakers or tanks with fine air bubbles venting in from the bottom were equally unsuccessful, and yielded embryos that did not manage to partition the inner from the outer yolk sack, leading to incomplete yolk epithelium development and thus, embryo malformation and death. Accordingly, egg strings should not be exposed to air bubbles and aeration of the water is therefore best performed outside of the tanks that house the strings. A second major improvement to our tank system was the combination of a relatively strong water flow and attachment of strings to the lateral side of the tank where the main current is, several centimeters below the water surface, ensuring that the strings were swirling around gently in the water. These adaptations yield a similar condition in which eggs are continuously rubbing against each other, likely functioning as a natural cleaning system. Third, we maintained the eggs in very dim light conditions (0-5 lx) using a 14 L:10D photoperiod, which likely mimics the natural dark environment of egg clutches in the den. To what extent egg maintenance in dim light is absolutely required remains to be studied.

Hallmarks of good quality embryos
Using these conditions, we noted a highly synchronous development within each string, with very little embryonic death or malformation occurring. Whereas embryonic development progress is more difficult to assess before Stage VII.2, after the first reversion, a number of hallmarks can be used to assess vitality of the embryos, such as yolk contractions, and later on heart beating, although these might be irregular at early embryonic stages. Inability to gradually reduce the inner yolk during organogenesis, frowning of the retina and increased presence of particles on the chorion are signs of poor embryo condition, and Octopus insularis eggs to clean them from microorganism contamination prior to individual egg housing in restricted water circulation [43]. Restricted housing without bleaching caused 100% mortality within a few days whereas 67.6% of the bleached embryos survived. Although individual egg housing can be beneficial for certain experiments, it is extremely labor intensive and requires much more space to house the same amount of eggs compared to the system described here.

Developing clear staging criteria
Several hallmarks can be used to easily identify developmental stages in O. vulgaris. In the early embryo, the rate of epiboly demarcates each stage. Afterwards, from Stage IX to Stage XIV, the formation of the funnel, as well as mantle shape and size can be used to differentiate the embryos. From Stage XV onwards, the amount, color intensity and reactivity of the chromatophores increases with embryo development and the size of the outer yolk sack is progressively reduced until it is completely absorbed at hatching (Table 1). When rearing conditions are not ideal, premature hatching occurs and paralarvae hatch out with the outer yolk sack still present, resulting in high mortality rates [44].
In cephalopod research, two different representations of body axes are used at random (i.e. morphological and functional body axes). When adopting the morphological body axes of a cephalopod, the embryonic mouth is anterior and the funnel posterior, the mantle dorsal and the arms ventral. In this setup, the mouth-funnel axis corresponds to the molluscan anterior-posterior axis where the foot is ventral. On the other hand, when using the functional body axes that correspond to the adult convention, the embryonic mouth is dorsal and the funnel ventral, the mantle posterior and the arms anterior. For the sake of comparison, the body axes should be clearly defined in each publication.

Conclusions
The data presented here aimed at facilitating developmental research on cephalopods, and in particular octopus species, under standardized laboratory conditions. We therefore removed potential roadblocks, such as obligatory maternal care and the availability of natural seawater, which we solved by introducing a low-cost standalone tank system that runs on artificial seawater. Given the high fecundity of O. vulgaris females, the high number of eggs from each string and the robustness of the embryos, egg strings from different females can be shipped and shared between laboratories in order to serve the growing community. In the present study, using classical and contemporary imaging technologies, we generated a comprehensive overview of O. vulgaris embryonic development along with a practical illustrated atlas. We documented the different stages of embryonic development and compared them to published literature, allowing practical use and unambiguous staging, which represents a reliable resource for comparative developmental biology in the cephalopod field.

Standalone system for egg incubation and embryo maintenance
Live egg strings of O. vulgaris were obtained from breeding females from the Instituto Español de Oceanografía (IEO), Tenerife, Spain, as soon as possible after spawning. The egg strings were attached to a nylon thread and transported in seawater in closed 50 mL falcons at ambient temperature to the Laboratory of Developmental Neurobiology in Leuven, Belgium. Transport time from tank to tank amounted to a maximum of 12 h. Upon arrival in the lab, single strings were placed in a standalone system that consisted of 10 conocylindrical opaque PVC tanks (16 cm diameter, 25 cm height), with a water inlet placed at the top to create a circular current with a water exchange rate of 3 L min − 1 (Fig. 2a). The standalone system continuously circulated aerated artificial seawater (Instant Ocean 40 g L − 1 , supplemented with 8 mg L − 1 Strontium), which was continuously cooled to 19°C, sterilized by UV (Deltec Profi UV sterilizer 39 W type 391), filtered through a mesh (1 mm) in each tank and circulated through a shared biological filter (21 × 21 × 11 cm, MarinePure Block, CER MEDIA) (Fig. 2b). The total volume of the system was 100 L, conductivity 50-55 mS, light intensity between 0 and 5 lx (dusk-dark) with a photoperiod of 14 L:10D and pH was maintained between 8.1-8.3. Each O. vulgaris egg string was attached to a glass rod using the nylon threads and placed on the lateral side of a tank, where it was in constant motion generated by the gentle current from the water inflow (Fig. 2a). The top of the tanks was covered with plastic foil to avoid evaporation and Aluminum foil to block light. After observation of the second reversion, embryos were left undisturbed for 7 days to avoid premature hatching [44] and were then transferred to a separate, opaque tank, containing the same aerated artificial seawater as the standalone system. This allowed observation of paralarval swimming, color change and phototactic behavior right after hatching.

Bright-field imaging
Egg strings were obtained from four different females. Embryos were observed daily and a sample of 20 representative embryos was removed daily from the string for imaging. All observations were based upon embryos reared in the standalone system. At least 4 strings for each female were monitored. Since fertilization was not timed and spawning takes place over several days, different strings of a single female were in different developmental stages, allowing monitoring of subtle changes and transitions during embryonic development. Embryos reared in this system were compared to fixed reference embryos obtained from sibling strings at the laboratory of E. Almansa (IEO) and also to independent reference embryos from the laboratory of E. Vidal (Center for Marine Studies, University of Parana, Brazil). Images were taken with a Zeiss Stereo Discovery.V8 equipped with an AxioCam ICc 3 camera (Carl Zeiss AG, Germany) and represent static stages based on a morphological consensus from different embryos.

Optimized CUBIC clearing protocol
The advanced CUBIC (Clear, Unobstructed Brain/Body Imaging Cocktails and Computational Analysis) protocol adapted from Susaki et al. [45] was used for clearing before light sheet imaging. This method has been used for whole body clearing of many organisms, is passive and thus does not require specialized expensive equipment and training, and also uses hydrophilic reagents, which are safer compared to organic clearing reagents. In short, eggs were fixed overnight in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS)or from Stage XX.1 onwards first submersed in 2% EtOH in seawater (to avoid stress and premature hatching) and then fixed in 2% EtOH, 4% PFA in seawater -and washed in PBS. To anticipate retrieval and convenient manipulation of the cleared embryos, Chinese ink was injected in the yolk before manual dechorionation using forceps. Embryos were incubated in 1/2-destilled-water-diluted ScaleCUBIC-1 (ScaleCUBIC-1: 25 wt% urea, 25 wt% Quadrol, 15 wt% Triton X-100) in an orbital shaker of a hybridization oven at 37°C for 3-6 h and then immersed in ScaleCUBIC-1. After overnight incubation, ScaleCU-BIC-1 was replaced and embryos were further incubated for 3 days with one additional ScaleCUBIC-1 replacement. At this point, the yolk was completely transparent, chromatophores were cleared and the eye pigment of Stage XX embryos was reduced from black to reddish (comparable to live Stage XIII embryos). Embryos were then washed with PBS three times (1 × 2 h, 1x overnight and 1 × 2 h) in the hybridization oven. Afterwards, they were incubated in 1/2-water-diluted ScaleCUBIC-2 (Sca-leCUBIC-2: 25 wt% urea, 50 wt% sucrose, 10 wt% triethanolamine) for 3-6 h (until the samples sunk to the bottom) and then incubated in ScaleCUBIC-2 for 1 day in the hybridization oven. For nuclear staining, DAPI (final concentration 1 μg mL − 1 ) was added to ScaleCU-BIC-1 in the 3 days incubation in ScaleCUBIC-1 step and during washes in PBS.

Light sheet fluorescence microscopy (LSFM)
Stained embryos were glued with their yolk sack on a metal rod and imaged using a Zeiss Z1 light sheet microscope (Carl Zeiss AG, Germany) in low-viscosity immersion oil mix (Mineral oil, Sigma M8410 and Silicon oil, Sigma 378,488, 1:1). Then, 3D reconstructions were generated in Arivis (Vision4D, Zeiss Edition 2.10.5).