DEV02829 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure 1 -

  Fig. S1. Whole-mount ISH, TUNEL staining and cell proliferation analysis of Hesx1^−/− mutants. (A) ISH with Atx. The rostral limit of the Atx expression domain appears anteriorised in the Hesx1^−/− mutant (right) as compared with the wild-type embryo (left) (arrowheads). (B) ISH with Pax6. Expression of Pax6 in the dorsal region of the diencephalon and pretectum (bracket) appear expanded in the Hesx1^−/− mutant (right) as compared with the wild-type littermate (left). (C-J) TUNEL staining of wild-type embryo (C,E,G,I) and Hesx1^−/− mutant (D,F,H,J). Cell death around the forebrain-midbrain boundary (arrowheads) is enhanced in the Hesx1^−/− mutant (D) as compared with the wild-type littermate (C). (E,F) Frontal views of the embryos depicted in C,D, showing enhanced apoptosis in the forebrain-midbrain boundary and decreased cell death in the roof plate of the telencephalon (arrowhead). (G,H) Transverse sections of wild-type embryo (G) and Hesx1^-/- mutant (H) showing that cell death occurs within the neuroepithelium. White line in C indicates the plane of section. Cell death within the anterior neural plate is undistinguishable between genotypes at 8.5 dpc (I,J). (K,L) No significant differences are observed in cell proliferation between wild-type (K) and Hesx1^-/- (L) embryos at early somite stages.

• Supplemental Figure 2 -

  Fig. S2. Genetic fate mapping of Hesx1-expressing cells in Hesx1^Cre/+;R26^{Cond-lacZ/Cond-lacZ} embryos. All embryos were X-Gal stained. (A) Frontal view of a 7.5-dpc embryo showing staining in the visceral endoderm of the extraembryonic region (arrowhead). Staining is beginning to appear in the anterior foregut (arrows). (B) Dorsal view of a 1- to 2-somite embryo showing strong lacZ expression in the anterior neural plate. (C) Fronto-ventral view of the embryo depicted in B, showing lacZ expression in the foregut pocket (arrowhead) and anterior neural plate (arrow). (D) Lateral view of a 6- to 7-somite embryo. lacZ expression is observed in the anterior neural plate (arrowhead), anterior foregut (black arrow) and in scattered cells of the yolk sac (white arrow). (E) 8- to 10-somite embryo, lateral view. lacZ-positive cells are observed in the AFB region. Arrowhead indicates the boundary between anterior and posterior forebrain. (F-H) Lateral (F), frontal (G) and ventral (H) views of the head of a 10.5-dpc embryo. The vast majority of lacZ-positive cells are located within the AFB: telencephalon (black arrow) and eyes (black arrowhead). Only scattered lacZ-positive cells are observed in the posterior forebrain: dorsal diencephalon (white arrow) and pretectum (white arrowhead). Note the absence of lacZ-positive cells in the midbrain (red arrow). Additionally, X-Gal staining is observed in the olfactory placodes (white arrowhead in G) and the developing Rathke’s pouch (arrowhead in H). (I) lacZ expression is also detectable in the liver bud (arrowhead). (J-M) Frontal sections of the 10.5-dpc embryo depicted in F-I. dd, dorsal diencephalon; dt, dorsal telencephalon; ey, eye; ht, heart; hy, hypothalamus; lb, liver bud; op, olfactory placode; rp, Rathke’s pouch; vd, ventral diencephalon; vt, ventral telencephalon.

• Supplemental Figure 3 -

  Fig. S3. Analysis of Cre expression in Hesx1^Cre/Cre, Hesx1^Cre/− and Hesx1^Cre/+ embryos. (A-C) ISH with Hesx1 (A) and Cre (B,C) on Hesx1^+/+ and Hesx1^Cre/Cre embryos at the 10- to 12-somite (A,B) and 6- to 7-somite stages (C). The Cre expression domain appears slightly smaller than the Hesx1 expression domain, but it is restricted to the ventral forebrain. (D,E) ISH with Cre on Hesx1^Cre/+ (D) and Hesx1^Cre/− (E) embryos. Cre expression is restricted to the ventral forebrain in both genotypes. Staining was detectable only after five days in NBT/BCIP staining solution. (F) qRT-PCR of Hesx1^Cre/+ and Hesx1^Cre/− embryos at the 2-, 5- and 8-somite stage. At the 2-somite stage, Cre expression levels are similar in both genotypes, but lower in Hesx1-deficient embryos at the 5- and 8-somite stages. Error bars represent ± s.e.m. Significant differences are observed only at the 5-somite stage. Gapdh was used as an internal control. Cre expression in the Hesx1^Cre/+ embryo was assigned a value of 1.0 for each stage. At least two embryos were analysed for each stage. The graph summarises the results of five repeats.

• Supplemental Figure 4 -

  Fig. S4. X-Gal staining of Hesx1^Cre/−;R26^Cond-lacZ/+ and Hesx1^Cre/+;R26^Cond-lacZ/+ embryos. Cryosections of Hesx1 ^Cre/−;R26^Cond-lacZ/+ (A-D) and Hesx1^Cre/+;R26^Cond-lacZ/+ (E) embryos were stained at 16.5 dpc. X-Gal-positive cells colonise the nasal septum (A,B) and the follicle primordium of the vibrissae (C,D) (arrowheads). A,B were photographed at 50× magnification and C,D at 400×. (E) lacZ expression in the pituitary is strong in the anterior and intermediate lobes (arrow), but hardly detectable in the posterior lobe (arrowhead). (F) X-Gal staining of Hesx1^Cre/−;R26^Cond-lacZ/+ mutant at the 3-somite stage. No significant differences are observed as compared with Hesx1^Cre/+;R26^Cond-lacZ/+ embryos (see Fig. S2B).

• Supplemental Figure 5 -

  Fig. S5. X-Gal-positive cells do not colonise the midbrain in Hesx1^Cre/+;R26^Cond-lacZ/<^b>⁺ or Hesx1^Cre/−;R26^Cond-lacZ/<^b>⁺ embryos. 12.5-dpc brains from Hesx1^Cre/+;R26^Cond-lacZ/+ (A,C) and Hesx1 ^Cre/−;R26^Cond-lacZ/+ (B,D) embryos. Embryos were X-Gal stained first, followed by double ISH with Shh (purple) and Pax6 (red) (A,B), or single ISH with Pax6 (C,D). Note that many more lacZ-positive cells (arrow, light blue) colonise the brain region posterior to the ZLI (arrowheads) (A,B). The region of the embryos depicted in C,D is boxed in A,B. Note that lacZ-positive cells are located within the Pax6 expression domain in the posterior forebrain and pretectum and do not cross the posterior boundary of the Pax6 expression domain. Shh was stained with NBT/BCIP, and Pax6 was stained either with INT/BCIP (A,B) or NBT/BCIP (C,D).

• Supplemental Figure 6 -

  Fig. S6. Generation of the R26-Cond-Hesx1 targeted allele. (A) Top to bottom: pBigT-Hesx1, a plasmid containing a loxP-flanked cassette with PGK-Neo selectable marker and a tpA transcriptional stop sequence, into which the Hesx1 coding region was cloned; pR26PA, containing genomic R26 sequences for homologous recombination; wild-type R26 locus, with the location of the probe indicated (red line); the structure of the targeted locus prior to and after Cre-mediated excision of the loxP-flanked (PGK-Neo, tpA) cassette. (B) Southern blot hybridisation of DNA isolated from wild-type (+/+) and two heterozygous (R26^Cond-Hesx1/+) ES clones after digestion with EcoRV. (C) Representative examples of PCR genotyping of wild-type, R26^Cond-Hesx1/+ and R26^{Cond-Hesx1/Cond-Hesx1} embryos. (D,E) RNA ISH with Hesx1 detects transcripts in the ventral diencephalon and developing Rathke’s pouch in the R26^Cond-Hesx1/+ embryo (left), but they are ubiquitously expressed in the R26^Cond-Hesx1/+;β-Actin-Cre compound embryo (right). Note the failure in the closure of the cranial neural folds in the R26^Cond-Hesx1/+;β−Actin-Cre compound embryo (arrowheads). (F) Coomassie staining (top) and western blot using a specific HESX1 antibody (bottom) of protein extracts isolated from: (1) heads or (2) tails from R26^Cond-Hesx1/+ embryos (10 μg); (3) heads or (4) tails from R26^Cond-Hesx1/+;β-Actin-Cre compound embryos (2 μg). HESX1 protein is approximately 22 kDa.

• Supplemental Figure 7 -

  Fig. S7. HESX1 cannot repress transcription from Wnt1 reporter plasmids. (A) Schematic of the Wnt1 locus. Boxes represent exons, black boxes represent the coding sequence of the gene. Thickened lines beneath correspond to regions upstream (165, 166 and 167) and downstream (2C, 1A, 147 and 39del) of the Wnt1 coding region which contain putative binding sites for HESX1. The grey line for constructs 2C, 1A and 147 corresponds to the sequence shown in D (+8138 to +8249), containing three putative binding sites for HESX1. (B,C) Co-transfection of reporter constructs with a plasmid expressing Hesx1 was unable to significantly repress activation of the reporter in any of the constructs tested. Co-transfection of a positive control reporter construct containing six repeats of a paired homeodomain consensus binding element (P3) showed that HESX1 was able to significantly repress transcription of this reporter by approximately 40% (P=0.002). pRLSV40 Renilla luciferase reporter vector (Promega) was used to control for transfection efficiency. Experiments were performed in triplicate and data presented as mean±s.d. (D) Binding of HESX1 to a putative enhancer element downstream of Wnt1 in vitro. (Left) Sequence of region downstream of Wnt1 (+8138 to +8249), indicated by grey region in A, containing three partial putative paired homeodomain binding sites (W1, W2 and W3, underlined). It has been shown that mutation of W3 causes ectopic activation of the Wnt1 locus in the forebrain of transgenic embryos (Iler et al., 1995). (Right) EMSA shows that HESX1 is capable of binding to two of these sites, W2 and W3 (lower band, monomer; upper band, dimer). No binding was observed to the W1 sequence. In the EMSA assay, ³²P-labelled probes were incubated with either: in vitro translation extract (lane 1); increasing amounts of in vitro translated (IVT) HESX1 (lanes 2 and 3); IVT HESX1 and 100 times excess of unlabeled probe (lane 4). W2 mut (ATTTAGGA) and W3 mut (TGGTTGAGGTAAGGA) ablated binding of HESX1 to the mutant probes, confirming the specificity of binding to these sites in vitro.

  Reference

  Iler, N., Rowitch, D. H., Echelard, Y., McMahon, A. P. and Abate-Shen, C. (1995). A single homeodomain binding site restricts expression of Wnt1 in the developing brain. Mech. Dev. 53, 87-96.