$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Explants allow for quantification of large-scale shape changes in 3D. To illustrate this, we explanted somite four (N = 3) from early zebrafish embryos obtained from a cross between Utr::mCherry (Tg(actb2:mCherry-Hsa.UTRN); e119Tg) and H2B::GFP (Tg(h2az2a:h2az2a-GFP); kca6Tg) heterozygous lines. Utrophin is an actin-binding protein, and the Utr::mCherry line shows the distribution of filamentous actin structures. This transgene was used here as a marker for cell outlines. H2B::GFP is a fluorescently-tagged histone and consequently marks the distribution of the chromatin, providing an effective description of nuclear location and shape, as well as mitotic figures that highlight cell division.
We assessed the success of the explant protocol during imaging. We observed that in a somite damaged during dissection either the integrity of the somite tissue is compromised, with many cells dissociating and extruding from the explanted somite, and/or many cells dying, which can be noted by the presence of fragmented nuclei in the nuclear channel. Successful explants remained healthy for 4-6 h after which changes in somite integrity were observed with cells dissociating from the explant and dying.
In the utrophin channel, we performed manual segmentation of explants in MATLAB (R2018b) on z-slices spaced every 10 µm. We developed a custom algorithm in MATLAB for segmentation, which is freely available for download (https://github.com/sundar07/SomSeg). In the algorithm, the file, frame and z-slice to be segmented can be set, followed by a user prompt, which allows for manual drawing of an outline around the region of interest. This was repeated for multiple z-slices and plotted, which showed rounding up of explants over time in 3D (Figure 3A). In addition, complementary information on tissue shape using the nuclear channel was also obtained. For this, we used Mastodon (version 1.0.0-beta-19, https://github.com/mastodon-sc/mastodon), a FIJI15 plugin, to obtain nuclei centroid positions through spot detection. We first converted the tif files from the microscope to xml/hdf5 format in FIJI, following which a new project was opened using the Mastodon plugin. In the plugin, we chose the spot detection option, where we defined a region of interest that covered the explant and used the difference of Gaussian detector with a diameter of 5 µm and a quality factor of 25 for detecting spots. We then transferred the nuclei centroid positions to MATLAB and used an in-built function (convhull) to obtain a convex hull (Figure 3B), which characterized the geometry of the explant. This similarly showed rounding of explants over time (Figure 3B). The nuclei data additionally allows for quantifying cell movements by tracking in 3D and a change in cell number over time. On the other hand, the utrophin channel allows for quantification of changes in cell shapes as explants become rounder. Together, these parameters are valuable for characterizing intrinsic material properties of somites, which aids in developing effective physical descriptions of tissue-scale shape changes.
Explants also allow for characterizing contact stresses with neighboring tissues in a quantitative manner. To illustrate this, we manually isolated two somites and placed them in close proximity and observed their dynamics (N = 2). For this experiment, the orientation of the somites with respect to the in vivo body axes was not tracked. Interestingly, over time, the explanted somites adhered to each other along one surface, while the free surfaces i.e., regions away from the contact site rounded up (Figure 4). This suggests that adhesive forces overcome stresses generated by surface tension at contact sites in explants. This can be further characterized by following shape changes as described above and by quantifying contact angles between the two tissues over time in 3D. Thus, the explants provide an attractive system for quantifying competing forces in action that lead to specific tissue shapes, the implications of which can then be explored in vivo.

Figure 1: Preparation of single-somite explants. Zebrafish embryo (A) is first dechorionated (B), followed by removal of skin and yolk (C). Somite-containing region of the embryo is then selected by removing rest of the tissues (D, E). Regions around the somite of interest are then serially removed (F-H) to ultimately isolate a single somite (I), followed by time-lapse imaging using a light-sheet microscope (middle z-section of somite shown here) (J). Abbreviations: S = somite; N = notochord; LPM = lateral plate mesoderm; A = anterior; P = posterior. Dashed lines indicate cut positions. Scale bar = 50 µm in all panels. Please click here to view a larger version of this figure.

Figure 2: Assembly of light-sheet imaging chamber. (A-D) A FEP membrane strip (dimensions in (B)) is wrapped around the membrane holder, fitted on to the imaging chamber and the membrane is glued to the imaging chamber. (E-F) The following day, the membrane holder is removed and an imaging mold (dimensions in (E)) is placed in the center of the imaging chamber in low-melting agarose. The entire unit is kept at 4 °C for 30 min, following which the mold is removed and the chamber with the trough is used for imaging explants. Please click here to view a larger version of this figure.

Figure 3: 3D analysis of tissue shapes. (A) Cell outlines were visualized with fluorescently-tagged utrophin (Utr::mCherry) and nuclei were visualized with fluorescently-tagged histone (H2B::GFP). The outlines of the explant were manually segmented at multiple depths using MATLAB and shown here. (B) Nuclei (red) were detected in the same explant using Mastodon, a FIJI plugin, and subjected to a convex hull (cyan), which informs on the geometry of the explant. Note rounding up of explants is evident in both analyses. Please click here to view a larger version of this figure.

Figure 4: Explanted somites adhere to each other. Two somites isolated from embryos and manually placed in close proximity tend to adhere over time (N = 2). Multiple z-slices with a frame interval of 2 min were acquired using a light-sheet microscope and middle sections of somites from selected time points are shown here. Cell outlines were visualized with fluorescently-tagged Utrophin (Utr::mCherry). Scale bar = 25 µm. Please click here to view a larger version of this figure.