Visualisation of cerebrospinal fluid flow patterns in albino Xenopus larvae in vivo

Abstract

Background: It has long been known that cerebrospinal fluid (CSF), its composition and flow, play an important part in normal brain development, and ependymal cell ciliary beating as a possible driver of CSF flow has previously been studied in mammalian fetuses in vitro. Lower vertebrate animals are potential models for analysis of CSF flow during development because they are oviparous. Albino Xenopus laevis larvae are nearly transparent and have a straight, translucent brain that facilitates the observation of fluid flow within the ventricles. The aim of these experiments was to study CSF flow and circulation in vivo in the developing brain of living embryos, larvae and tadpoles of Xenopus laevis using a microinjection technique.

Methods: The development of Xenopus larval brain ventricles and the patterns of CSF flow were visualised after injection of quantum dot nanocrystals and polystyrene beads (3.1 or 5.8 μm in diameter) into the fourth cerebral ventricle at embryonic/larval stages 30-53.

Results: The fluorescent nanocrystals showed the normal development of the cerebral ventricles from embryonic/larval stages 38 to 53. The polystyrene beads injected into stage 47-49 larvae revealed three CSF flow patterns, left-handed, right-handed and non-biased, in movement of the beads into the third ventricle from the cerebral aqueduct (aqueduct of Sylvius). In the lateral ventricles, anterior to the third ventricle, CSF flow moved anteriorly along the outer wall of the ventricle to the inner wall and then posteriorly, creating a semicircle. In the cerebral aqueduct, connecting the third and fourth cerebral ventricles, CSF flow moved rostrally in the dorsal region and caudally in the ventral region. Also in the fourth ventricle, clear dorso-ventral differences in fluid flow pattern were observed.

Conclusions: This is the first visualisation of the orchestrated CSF flow pattern in developing vertebrates using a live animal imaging approach. CSF flow in Xenopus albino larvae showed a largely consistent pattern, with the exception of individual differences in left-right asymmetrical flow in the third ventricle.

Figure 1

Sites for the injection of polystyrene beads into the ventricles of anesthetised Xenopus larvae. A glass capillary filled with an aqueous suspension of red beads (2.9 μm in diameter) was inserted into the lateral ventricle (a, b) or the fourth ventricle (c, d) for injections. Arrows indicate the injection points. (b, d) magnified views of (a, c), respectively. Scale bars, 1 mm (a, c) and 0.5 mm (b, d).

Figure 2

Morphogenesis of the cerebral ventricles during X. laevis development at stages 38-53, visualised by Qdot565 nanocrystals. The injection into the fourth cerebral ventricle of anesthetised embryos labelled the fourth cerebral ventricle almost exclusively in stage 38 (a) and stage 43 (b) embryos. The third cerebral ventricle (III) forms an oval shape by stage 43, and the midbrain ventricle (M) is visible between the fourth and third ventricles by stage 46. The lateral ventricles (L) bud off the third ventricle at stages 46 to 48 (c, d). All ventricles can be recognised clearly at stage 50 (bright-field; e, fluorescence; f). The morphology of the ventricles in stage 53 larvae is different from those of early larval stages (g). The central canal of the spinal cord is also labelled (h). All panels are dorsal views with anterior at the top. Scale bars, 1 mm. ca: cerebral aqueduct.

Figure 3

Visualisation of fluid flow by injecting polystyrene beads into the cerebral ventricles of anaesthetised albino X. laevis larvae at stage 47-49. Serial multi-coloured dots represent the trajectories of each bead from 0.7 to 5.3 seconds. The brightness of the video frames are inverted, and the trajectories of the beads were visualised by red - > orange - > yellow - > green - > blue - > purple, in turn, using a combination of Image-J software and its plug-in "Color Footprint Rainbow". (a) The beads diffused into the third cerebral ventricle mainly along the left wall of the ventricle just after the injection. (b, c) The beads moved with left-right asymmetrical circulation in the third ventricle (b) and exited through the cerebral aqueduct along the wall of the third ventricle (c). In many individuals, a left-right asymmetry in bead movements in the third ventricle was observed (see Table 1 and Table 2). (d) Within the lateral ventricles, the beads move along the distal walls toward the medial walls creating a semi-circular movement. (e-h) In the cerebral aqueduct, the beads moved rostrally in the dorsal region [(e) and (g) magnified view of (e)] and caudally in the ventral region [(f) and (h) magnified view of (f)]. (i, j) In the fourth ventricle [(i) dorsal side; (j) ventral side], beads tended to concentrate toward the centre and move to the anterior in the dorsal region, whereas beads tended to flow toward the sides and move to the posterior in the ventral region. cp: choroid plexus; lv: lateral ventricle; ca: cerebral aqueduct (aqueduct of Sylvius). All panels are dorsal views (see Additional file 1: movie 1, Additional file 2: movie 2, Additional file 4: movie 4, Additional file 5: movie 5, Additional file 6: movie 6, Additional file 7: movie 7, Additional file 8: movie 8, and Additional file 9: movie 9). Duration of recording; (a) 5.3 sec; (b) 1.0 sec; (c) 0.9 sec; (d) 1.2 sec; (e) 0.9 sec; (f) 0.7 sec; (i) 1.5 sec; (j) 2.6 sec. Scale bars, 0.1 mm.

Figure 4

Visualisation of bead movements injected into the fourth ventricle within the fourth ventricle in stage 39 (a-c) and stage 42 (d-f) anaesthetised non-feeding embryos. Anterior is to the top. (a) Bright-field image of a stage 39 embryo (dorsal view). Orange box indicates the area of recording in (b). (b) Trajectories of the beads (diameter = 5.8 μm) within the fourth ventricle are shown in a separate embryo (dorsal view). The trajectories show the existence of some flow, but a discrete pattern was not apparent. (c) Magnified view of (b). Note the non-directional bead movements. (d) Bright field image of a stage 42 embryo (dorsal view). (e) Trajectories of the beads (diameter = 5.8 μm) within the fourth ventricle in the same embryo (dorsal view). Note that a radial CSF flow, from the centre of the ventricle toward the wall, is generated at stage 42. (f) Magnified view of (e). See Additional file 10: movie 10. Duration of recording; (b) 2.7 sec; (e) 0.9 sec. Scale bars, 0.5 mm (a, d) and 0.1 mm (b, e).

Figure 5

Injection of oil into the Xenopus larval ventricles to estimate their volumes. A living stage 46 larva after injecting 87 nL of sesame oil into the brain ventricles. The yellow sesame oil fully fills the cerebral ventricle of the larva, as observed by the morphology of the ventricles (a, dorsal view; b, lateral view). Scale bars, 0.5 mm.