Pineal progenitors originate from a non-neural territory limited by FGF signalling

Abstract

The embryonic development of the pineal organ, a neuroendocrine gland on top of the diencephalon, remains enigmatic. Classic fate-mapping studies suggested that pineal progenitors originate from the lateral border of the anterior neural plate. We show here, using gene expression and fate mapping/lineage tracing in zebrafish, that pineal progenitors originate, at least in part, from the non-neural ectoderm. Gene expression in chick indicates that this non-neural origin of pineal progenitors is conserved in amniotes. Genetic repression of placodal, but not neural crest, cell fate results in pineal hypoplasia in zebrafish, while mis-expression of transcription factors known to specify placodal identity during gastrulation promotes the formation of ectopic pineal progenitors. We also demonstrate that fibroblast growth factors (FGFs) position the pineal progenitor domain within the non-neural border by repressing pineal fate and that the Otx transcription factors promote pinealogenesis by inhibiting this FGF activity. The non-neural origin of the pineal organ reveals an underlying similarity in the formation of the pineal and pituitary glands, and suggests that all CNS neuroendocrine organs may require a non-neural contribution to form neurosecretory cells.

Keywords: Chick; Circumventricular organs; Endocrine; FGF; Neural plate border; Neurulation; Otx; Pineal; Placode; Zebrafish.

Fig. 1.

The flh/noto/NOT1-expressing pineal progenitor region extends into the PPR. (A-H) Dual-colour in situ hybridisation for indicated genes on bud stage zebrafish embryos (anterior is leftwards). White asterisk in A marks pitx3-positive anterior PPR; black asterisks mark irx1b-positive posterior PPR. (D,F,H) Magnified views of the boxed areas in C,E,G, respectively; arrows indicate a putative pineal progenitor region; nc marks the neural crest in F,H. (I) In situ hybridisation for flh/noto (red) and immunohistochemical detection of DLX3 (green) in an early bud stage zebrafish embryo. (J) Magnification of boxed area in I showing partial overlap between flh/noto and DLX3 expression. (K) Transverse section of late bud stage embryo stained as in I and J at the level of the diencephalon. Arrow indicates the overlap between flh/noto and DLX3-expressing cells in the PPR (PrePlac.). All zebrafish wild-type double staining experiments were performed with at least 20 embryos, two independent experiments each. Images are representative of most or all embryos for each set of markers. (L) HH8.5 chick embryo stained using dual-colour in situ hybridisation for SOX2 (red) and NOT1 (blue; anterior is upwards). Arrowhead indicates level of section in M. (M) Transverse section through the neural tube of an embryo stained as in L at the level of the posterior forebrain (arrowhead in L). There is NOT1 staining in non-neural ectoderm overlying the SOX2-positive neural tube (n=3; two independent experiments). (N) Schematic colour-coded representation of gene expression patterns shown in A-K (Di, diencephalon; Hypoth, hypothalamus; Mes, mesencephalon; Tel, telencephalon). Scale bars: in A, 50 µm for A-C,E,G; in D,H, 15 µm for D,F,H,J; in I, 50 µm; 15 µm in K; 500 µm in L; 25 µm in M.

Fig. 2.

Non-neural ectoderm contributes to the pineal organ. (A-C) Small patches of cells (white arrows) were labelled using uncaging of fluorescein (Fluo, red) at bud stage. Embryos were fixed immediately and subjected to in situ hybridisation for irx1b (blue) to demonstrate precision of labelling (anterior is leftwards). A, n=6; B, n=4; C, n=7. (D-F) Lateral views (anterior is leftwards) of the diencephalon of embryos labelled as in A-C, incubated for 24 h and labelled for expression of otx5 (red) marking the definitive pineal organ (red arrowheads). Cells labelled by uncaging (blue) in the otx5-positive pineal organ are labelled in E and F. D, n=11; E, n=11; E, n=19/23. (G-I) Sequential images of time-lapse recording of an embryo from late bud to the 26-somite stage in which Kaede-expressing cells have been converted to red fluorescence in the ectoderm flanking the diencephalon. The trajectories of three representative cells (blue, red and green arrowheads in G) are shown in red, green and blue in H,I. (J) Dorsal view showing converted cells in the pineal organ at 22 hpf. See also Movies 1 and 2. (K) Lateral view, anterior towards the left, showing converted cells in the ectoderm and pineal organ, but not in the brain proper, at 30 hpf. Four out of 12 embryos had no converted cells in the brain proper; all four of those showed converted cells in the pineal organ. E, eye; L, lens; T, telencephalon. Dotted lines in J and K outline the pineal organ. Scale bars: 50 µm; bar in A applies to A-C,G-I; bar in D applies to D-F,J,K.

Fig. 3.

The PPR specifier genes dlx3b/dlx4b in conjunction with flh/noto are required and sufficient for pineal progenitor specification. (A-D) 24 hpf zebrafish embryos injected with the MOs indicated in B-D and stained for the pineal marker otx5 in red (lateral views, anterior is leftwards). There is reduced otx5 staining in B and C, and a complete absence in D. Scale bar: 50 µm. MO injection experiments were performed three times independently, injecting each MO or combination of MOs in parallel into 40-50 embryos from the same parents. Images are representative for each experimental condition. Fifty percent of the embryos injected with dlx3b/4b MOs looked as in C, the other 50% had a range of even weaker otx5 staining. (E,F) Injection of dlx3 and flh mRNAs at the one-cell stage results in ectopic induction of the pineal marker otx5 (blue) after 24 h of incubation (n=21/32; control embryo shown in E; two independent experiments). Inset in F is an image focusing on the ectopically induced otx5-expressing cells that are marked with the arrow in F. Scale bar: 50 µm. (G,H) Lateral views of embryos injected at the one-cell stage with hs:flh/noto and hs:dlx4b, heat-shocked at shield stage and stained by in situ hybridisation for otx5 (red) after 24 h of incubation. There is sparse ectopic induction of otx5 in superficial cells (arrowheads). Scale bars: 50 µm (bar in A applies to A-D; bar in E applies to E-H). (I) Quantification of otx5 induction in cells injected with hs:gfp alone, with hs:flh/noto and hs:dlx4b, or with hs:flh/noto alone (two independent experiments; six embryos were selected for statistical analysis; data are mean±s.d.). P<0.0001 (GFP versus flh+dlx), P<0.0001 (GFP versus flh), P=0.041 (flh+dlx versus flh).

Fig. 4.

Otx function is required for pinealogenesis. Wild-type zebrafish embryos are shown in A,C,E,G,I and otxH MO-injected embryos in B,D,F,H,J (anterior is leftwards, lateral views are shown in A-H, dorsal views in I,J). Asterisks mark the location of the pineal organ in A-H. (A,B) The pineal organ is morphologically reduced in otxH embryos at 48 hpf. (C,D) Immunohistochemical staining for opsin at 48 hpf; staining is absent from the pineal area in D. (E,F) Double in situ hybridisation for otx5 and shh (both in blue) at 24 hpf. otx5-positive pineal progenitors are absent in F. Arrows indicate the zona limitans intrathalamica; this signalling centre is reduced in F. (G,H) In situ hybridisation for otx5 at the eight-somite stage, including dorsal views of the pineal region shown in the insets. (I,J) In situ hybridisation for flh/noto at bud stage; flh/noto-positive pineal progenitors are absent in J (arrow in J indicates persistent flh/noto staining in notochord). Each marker was tested in two independent experiments, comparing at least 30 control and otxH MO-injected embryos for each marker and stage. Images are representative of all analysed embryos. Scale bar: 50 µm.

Fig. 5.

Wild-type cells can rescue pineal progenitor specification cell non-autonomously in otxH morphants. (A,B) Representative fluorescent images of wild-type zebrafish embryos containing cells grafted from an otxH embryo (green) and stained by in situ hybridisation for otx5 (red) at 24 hpf (lateral views, anterior is leftwards). Inset in B shows a dorsal view of the pineal region in more detail. otx5 is expressed in otxH cells that are adjacent to wild-type cells (arrow). Two independent experiments were performed grafting 13 and 19 embryos, respectively. (C) Bright-field image of an otxH embryo containing grafted wild-type cells and stained by in situ hybridisation for otx5 (red) at 24 hpf (lateral view, anterior is leftwards). (C′) Fluorescent images of the pineal region of the embryo shown in C in more detail; grafted wild-type cells are green. There is non-cell-autonomous rescue of otx5 expression (arrow). Inset in C′ shows a magnified dorsal view of the otx5-positive pineal organ in C′. Two independent experiments were performed grafting 17 and 14 embryos: 15/17 and 11/14 showed cell-non-autonomous pineal rescue. Scale bars: 50 µm in A,C′; 50 µm in insets (bar in the inset in C′ applies to the insets in both B and C′).

Fig. 6.

Upregulation of FGF signalling activity in the neural plate border region is responsible for the absence of pineal progenitors in otxH morphant embryos. (A-B′) Representative images of dual colour in situ hybridisation for erm (blue) and flh (red) in wild-type (n=29; A,A′) and otxH (n=32; B,B′) bud stage zebrafish embryos; anterior is leftwards. (A′,B′) The boxed neural plate border region in A,B in more detail. (C-H) Wild-type (C,E,G) and otxH morphant (D,F,H) zebrafish embryos stained for fgf8 (red) and otx5 (blue) at 24 hpf (anterior is leftwards, dorsal towards the top; C,D, n=18). Embryos in E,F were treated with the FGF inhibitor SU5402 (n=25 and n=22, respectively). Embryos in G,H are transgenic for a heat shock-inducible dominant-negative FGF receptor and were heat-shocked at shield stage (n=19 and n=22, respectively). otx5-positive pineal progenitors are rescued by pharmacological inhibition of FGF signalling in F and by heat shock-induced receptor inhibition in H (compare with D). Inset in H shows a sibling from the Tg(hsp:dnfgfr1) cross that did not show rescue and is therefore likely to be from the 25% of offspring that are expected to be negative for the transgene. Results are based on two independent experiments. Scale bar: 50 µm.

Fig. 7.

FGF signalling antagonises pinealogenesis. (A) Bud stage zebrafish embryo (anterior points to the top) containing transplanted fgf8-overexpressing cells in the right side of the neural plate (red), labelled using in situ hybridisation for flh (blue). flh-positive pineal progenitors are present on the right side (arrows; n=7/7). (B,C) Chick embryos (anterior towards the top) at HH8.5 (B) and HH9.5 (C) stained using in situ hybridisation for expression of FGF8 (n=6; two independent experiments). Expression is present in the midbrain-hindbrain boundary in B (arrowhead; the arrow marks expression in the tailbud region). Expression is present in the anterior neural folds (arrow) and midbrain-hindbrain boundary (arrowhead) in C. (D) HH18 chick embryo stained using in situ hybridisation for NOT1 (lateral view of diencephalon, anterior is rightwards). Inset shows same embryo at lower magnification. (E) Chick embryo electroporated at HH10 with FGF8 and eGFP, fixed after 36 h of incubation and stained by in situ hybridisation for expression of NOT1 (blue) and by immunohistochemistry for GFP (red, inset). Arrowheads mark the pineal organ; the arrow in the inset marks the location of electroporated cells at some distance from the pineal organ. NOT1 is downregulated in E compared with D (n=11/15; two independent experiments). (F) Zebrafish embryo (anterior is leftwards, dorsal towards the top) transplanted at 30% epiboly stage with heat shock-inducible dnfgfr-transgenic donor cells (green), heat-shocked at shield stage, fixed at 24 hpf and labelled by in situ hybridisation for otx5 (red). Ectopic pineal progenitors are induced following heat shock (n=9/12). Arrowhead highlights otx5-positive pineal organ; arrow indicates ectopic otx5-positive cells. (G) Lateral view (anterior towards the right) of a chick embryo electroporated at HH8 with dnFGFR and eGFP, and stained after 24 h of incubation by in situ hybridisation for NOT1 (blue) and by immunohistochemistry for GFP (red, inset). Arrowhead indicates NOT1-positive pineal organ; black and white arrows highlight ectopic NOT1-positive cells. (H) Oblique section of embryo along the line indicated in G. Arrowhead indicates pineal organ; arrows indicate clusters of NOT1-expressing cells in superficial ectoderm (n=6/13). (I) Lateral view of chick embryo electroporated at HH8 with dnFGFR and eGFP into the right anterior neural fold, and stained for NOT1 after 36 h of incubation (arrowhead indicates pineal organ). Inset shows GFP fluorescence before in situ hybridisation. There is ectopic induction of NOT1-positive pineal progenitor cells in electroporated neural tube (arrows; n=7/27). Scale bars: 50 µm in A for A-F; in B, 500 µm for B,C; in D, 200 µm for D,E,I; 100 µm in G for G,H.