Quantitative imaging of cell dynamics in mouse embryos using light-sheet microscopy

Abstract

Single/selective-plane illumination, or light-sheet, systems offer several advantages over other fluorescence microscopy methods for live, 3D microscopy. These systems are valuable for studying embryonic development in several animal systems, such as Drosophila, C. elegans and zebrafish. The geometry of the light path in this form of microscopy requires the sample to be accessible from multiple sides and fixed in place so that it can be rotated around a single axis. Popular methods for mounting include hanging the specimen from a pin or embedding it in 1-2% agarose. These methods can be particularly problematic for certain samples, such as post-implantation mouse embryos, that expand significantly in size and are very delicate and sensitive to mounting. To overcome the current limitations and to establish a robust strategy for long-term (24 h) time-lapse imaging of E6.5-8.5 mouse embryos with light-sheet microscopy, we developed and tested a method using hollow agarose cylinders designed to accommodate for embryonic growth, yet provide boundaries to minimize tissue drift and enable imaging in multiple orientations. Here, we report the first 24-h time-lapse sequences of post-implantation mouse embryo development with light-sheet microscopy. We demonstrate that light-sheet imaging can provide both quantitative data for tracking changes in morphogenesis and reveal new insights into mouse embryogenesis. Although we have used this approach for imaging mouse embryos, it can be extended to imaging other types of embryos as well as tissue explants.

Keywords: Cell dynamics; Light-sheet; Mouse embryo culture; Postimplantation; Quantitative.

Fig. 1.

Preparing hollow agarose cylinders for light-sheet microscopy of live mouse embryos. (A) A 1 ml syringe and the end of a cotton swab were used to create hollow agarose cylinders by filling the syringe with 1-2% liquid agarose, inserting the end of the swab into the middle of the syringe and extruding the solidified agarose using the syringe plunger. The cut end of a 1 ml micropipette tip was used to center the swab. (B) The hollow agarose cylinder was removed from the cotton swab stick using a razor blade and placed into a Petri dish with sterile PBS. (C-E) Live mouse embryos (red arrows) were transferred into hollow agarose cylinders (equilibrated in warm culture medium) using forceps. Embryos sink to the bottom of the agarose cylinder when the cylinder is tilted upright. (F) The cotton swab is gently reinserted into the open end of the cylinder, fitting snugly, to provide a way to handle the cylinders and mount them in the light-sheet system.

Fig. 2.

Culturing mouse embryos in hollow agarose cylinders. (A) Embryos (white arrow) were cultured in hollow agarose cylinders filled with culture medium. (B) The agarose cylinder is then submerged in dissection medium. (C-F) E8.5 embryos grown in both agarose cylinders submerged in dissection medium and in a Petri dish in culture medium grew well and underwent partial to full vascular remodeling and embryo turning. Scale bars: 1 mm. (G) Comparison of static embryo culture in a Petri dish versus embryo culture in cylinders showed no significant difference in embryonic growth when judged by a five-point scale as defined in the Materials and Methods. (H) The swab and the agarose cylinder assembled in the sample holder of the Lightsheet Z.1 microscope. (I) Schematic for the light path, the agarose cylinder and the imaging chamber. (J) The stainless steel imaging chamber within the Lightsheet Z.1 microscope with the sample in place.

Fig. 3.

Live imaging of Flk1-myr::mCherry- and Flk1-H2B::eYFP-labeled endothelial cells revealed endothelial cell migrations at the border of the vascular and avascular space. (A,B) Depiction of entire E8.5 embryo showing four key events captured during the time lapse: yolk sac expansion (blue arrows), two regions of EC migrations, proximal to the edge of the yolk as well as distal from the edge (white arrows and white boxes), and sprouting angiogenesis/vessel anastomosis (brown box) at the edge of the yolk sac. (C,D) Select images from time-lapse sequences (4.7 h and 14.5 h) showing EC nuclei (YFP) and membrane (mCherry fluorescent protein) of cultured E8.5 embryo and yolk sac. During the culture, yolk sac and blood vessel expansion occurs around the embryos, as indicated by the blue arrows in C and the final position of the yolk sac, while the avascular spaces are maintained (selected white shapes). (E,F) Endothelial cells close to the edge of the yolk sac show directed migration trajectories as it begins to close (white arrows). E, the same image shown as in C, overlaid with the final migration trajectories to indicate the state of the tissue at 4.7 h. F similarly shows the 14.5 h image as in D, overlaid with the final trajectories. (G,H) Colored maps indicate ECs at 4.7 h (G) and 14.5 h (H) with large displacement lengths (>200 µm, green, yellow and red spheres in G,H). In the region more distal from the edge, ECs showed smaller displacement lengths (<100 µm, blue spheres in G,H). Images were acquired with a 20×, NA=1.0 objective, zoom=0.4, every 10 min for 16.5 h, z-step=2.67 µm, total z-depth=411.18 µm. Scale bar in C: 100 µm.

Fig. 4.

Analysis of dynamic events within the yolk sac and embryo. (A-D) Time-lapse images of E8.5 yolk sacs revealed sprouting angiogenesis (A, white arrows) at the edge of the yolk sac, followed by vessel anastomosis (B-C, blue arrows). Images were acquired with a 20×, NA=1.0 objective, zoom=0.4, every 10 min for 7.5 h, z-step=2.67 µm, total z-depth=411.18 µm. (E-H) Time-lapse, light-sheet microscopy images of paired dorsal aorta fusion (double arrows) showing that the opposing aorta are coming together at two different regions, but not progressively from a single starting point. Images were acquired with 20× Objective, NA=1.0, zoom=0.4, every 8 min for 7 h 36 m, z-step size=1.484 µm, total z-depth=71.22 µm. Scale bar: 100 µm.

Fig. 5.

Single- and multi-angle 3D reconstructions of an E6.5 TCF/LEF-H2B::eGFP mouse embryo. (A-I) 3D transparency projections of a TCF/LEF-H2B::eGFP-labeled E6.5 mouse embryo, posterior view (A,D,G), right lateral view (B,E,H) or transverse view (C,F,I). (A-C) 3D projections taken from a single-angle revealed attenuation of the fluorescence signal (arrows). (D-F) Deeper signals within the embryo were recovered using multi-angle imaging as shown in the 3D transparency projections. (G-I) Surface rendering of the multi-angle images (D-F) revealed the 3D volume of the embryo. Scale bar: 100 µm.

Fig. 6.

Live imaging of TCF/LEF-H2B::eGFP E6.5 embryos using agarose cylinders and 3D tracking of endothelial cells. For all panels, images from select time points from the 3D time-lapse movie are on the left and images from the movie overlaid with Imaris 3D tracking data are on the right. Embryos are oriented with the anterior side on the left. (A-F) Cultured embryos in hollow agarose cylinders show a twofold increase in embryo size over 24 h. Images from select time points are shown (A, 5 h 30 min; B, 8 h 50 min; C, 12 h 10 min; D, 15 h 30 min; E, 18 h 50 min; F, 22 h 10 min), and 3D cell tracks represent cell movements that occurred during the 2.5-h time period before the respective time point. Images from early time points show passive, radial cell movements associated with embryo growth (A-C), while images from later time points (E-F) show distinct antero-lateral migration events of deeper TCF/LEF-H2B::eGFP⁺ cells as gastrulation begins. GFP⁺ endoderm cells on the surface show less movement. Images were taken with a 20× objective, NA=1.0, zoom=0.7, every 10 min for a total of 24 h, z-step size=1.14 µm, total z-depth=115 µm. Scale bars: 50 µm.