QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 4 May 2005
doi: 10.1242/dev.01845

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sirbu, I. O. Articles by Duester, G. Search for Related Content PubMed PubMed Citation Articles by Sirbu, I. O. Articles by Duester, G.

Shifting boundaries of retinoic acid activity control hindbrain segmental gene expression

¹ OncoDevelopmental Biology Program, Burnham Institute, 10901 North Torrey Pines Road, La Jolla, CA 92037, USA
² Unité d'Expression Génétique et Maladie URA 1644, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15, France
³ Unité de Biologie du Développement URA 2578, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15, France

View larger version (71K): [in a new window]   Fig. 1. Boundaries of Cyp26a1, vHnf1 and Irx3 expression in the early mouse hindbrain. Anterior is oriented towards the top in all panels, and all embryos are wild type (WT). (A-E) Cyp26a1 mRNA at E7.75-E8.0 is localized anteriorly with its posterior extent to presumptive rhombomere 2 (r2) of the hindbrain, as determined by whole-mount in situ hybridization. Shown is an E7.75 embryo stained for expression of Cyp26a1 (A), and an E8.0 embryo double-stained for expression of Cyp26a1 and Krox20, a known marker for r3 (B). Also shown is an E7.75 embryo double-stained for expression of Cyp26a1 and Hoxb1, a known marker for r4 (C); neither gene is expressed in r3. Double-staining for Cyp26a1 and Epha2 (a marker for presumptive r4 and the adjacent mesoderm) at E7.75 (D) and E7.9 (E) demonstrates a space between the two expression domains where presumptive r3 lies; expression of Epha2 in the mesoderm adjacent to r4 overlaps the lateral regions of presumptive r3 from this viewpoint. (F,G) Expression of vHnf1 is localized posteriorly with its anterior border at the r4/r5 boundary Shown are embryos double-stained for expression of vHnf1 and Krox20 at E8.0 when Krox20 is expressed only in r3 (F) and at E8.25 when Krox20 is expressed strongly in r3 and weakly in r5 (G). There is a gap between Krox20 and vHnf1 expression at both stages, indicating that vHnf1 expression extends to r5 but not into r4. (H) Double staining for Epha2 and vHnf1 shows that vHnf1 is expressed in r5 up to the r4/r5 boundary. (I) Expression of Irx3 at E8.25 is localized anteriorly with a posterior border at the r4/r5 boundary when compared with the expression pattern of vHnf1 (G). (J) Double staining for Irx3 and vHnf1 expression at E8.0 provides evidence that these two expression domains meet at r4/r5.

View larger version (101K): [in a new window]   Fig. 2. Boundaries of hindbrain RA activity along the anteroposterior axis. All embryos are homozygous for the RARE-lacZ RA-indicator transgene; stages are indicated on each panel. Anterior is oriented towards the right in A-D (lateral views) and towards the top in E-O (dorsal views). (A-C) Embryos were double-stained for RARE-lacZ expression (ß-galactosidase activity), which is observed posteriorly, followed by Cyp26a1 expression (in situ hybridization), which is observed anteriorly with its posterior border at r2. RARE-lacZ expression is first seen at E7.5 (B) and the r2/r3 division marked in C indicates the anterior extent of RARE-lacZ, which reaches r3 at E7.6. (D,E) Raldh2 expression detected posteriorly in the paraxial mesoderm at E7.6. (F-K) Dorsal view of embryos double-stained for expression of RARE-lacZ and Cyp26a1 from E7.4-E7.9. RARE-lacZ expression has reached r3 by E7.6 (H), but begins to clear from r3 by E7.9 (K). (L-N) Double-stained embryos showing RARE-lacZ expression (white arrows indicate its anterior extent at the r4/r5 boundary) and expression of either Krox20 in r3 at E8.0 (L), Krox20 in both r3 and r5 at E8.25 (M) and Hoxb1 in r4 at E8.5 (N).

View larger version (60K): [in a new window]   Fig. 3. RA generated by Raldh2 is required for Cyp26c1 expression in r4. Anterior is oriented towards the right in all panels except B, where anterior is towards the top. (A,B) Lateral (A) and dorsal (B) views of E7.75 Raldh2–/– (–/–) and wild-type (WT) embryos double stained for expression of RARE-lacZ (ß-galactosidase activity) observed posteriorly followed by staining for Cyp26c1 mRNA (whole-mount in situ hybridization) observed anteriorly. The Raldh2–/– embryo lacks RARE-lacZ expression, but Cyp26c1 expression at this stage (head mesoderm) is not affected. (C-E) Cyp26c1 expression in dorsal view of wild-type embryos at E7.75 (C), E7.9 (D) and E8.0 (E) showing its initial expression in r4, with the remaining expression occurring in head mesoderm. (F-I) Dorsal view of double staining for expression of Raldh2 (posterior) and Cyp26c1 (anterior) in Raldh2–/– and wild-type embryos at E8.25 (F-G) and E8.5 (H-I). Raldh2–/– embryos (marked by a large reduction in Raldh2 mRNA) lack the r4 domain of Cyp26c1 expression while retaining the r2 expression domain normally observed by E8.5. Raldh2–/– embryos also exhibit an abnormal posterior extension of Cyp26c1 expression in head mesoderm at E8.5 marked by an asterisk (I). The anterior extent of Raldh2 mRNA in wild-type embryonic mesoderm is somite 1 (s1), which is adjacent to rhombomere 8 (r8) at E8.25 (F).

View larger version (75K): [in a new window]   Fig. 4. Requirement of RA for induction of Hoxb1 and vHnf1, and for repression of Irx3. Anterior is towards the right (A-D,G-J,N-O) or towards the top (E-F,K-M). (A-C) Hoxb1 expression normally exhibits an anterior extension into the hindbrain, but this extension is eliminated in Raldh2–/– embryos. Arrowheads indicate the anterior extent of Hoxb1 expression, which is at the level of the node in a wild-type embryo at E7.5 (A), and anterior to the node in the hindbrain at E7.75 (B) and E8.0 (C). The anterior expansion of Hoxb1 expression into the hindbrain is eliminated in an Raldh2–/– embryo (C). (D) Double in situ hybridization at E8.0, showing a lateral view of vHnf1 expression posteriorly in the hindbrain and Cyp26a1 expression anteriorly. The Raldh2–/– embryo lacks vHnf1 expression, but Cyp26a1 is not affected by a loss of RA. (E,F) A dorsal view of vHnf1 expression at E8.25, showing a complete loss of vHnf1 mRNA in an Raldh2–/– embryo. (G,H) vHnf1 expression at E8.5 is normally present in the anterior spinal cord, as well as the posterior hindbrain up to the otic vesicle (ov), but an Raldh2–/– embryo lacks vHnf1 mRNA in these neural tissues. (I,J) Neural expression of Hoxb1 at E8.5 is normally limited to r4 in the hindbrain and to the posterior spinal cord, but an Raldh2–/– embryo lacks the r4 expression domain and instead exhibits a weak expression domain shifted to the most posterior region of the hindbrain (asterisk). (K,L) Double-staining for expression of Cyp26c1 (anterior) and Hoxb1 (posterior) at E8.25 demonstrates that both overlap in r4 of wild-type embryo, but that the Raldh2–/– embryo lacks expression of both in the region where r4 should have developed (arrowhead). The brackets indicate a region where the Raldh2–/– embryo retains expression of Hoxb1 in the most posterior region of the hindbrain (asterisk), whereas this domain of Hoxb1 expression is normally lost by E8.25. (M) Double staining for expression of Irx3 and Raldh2 in E7.9 wild-type and Raldh2–/– embryos demonstrates the normal posterior border of Irx3 expression at r4 separated from the Raldh2 domain, whereas the mutant exhibits a loss of Raldh2 expression and an expansion of Irx3 expression into the posterior hindbrain (bracket and asterisk). (N,O) Irx3 expression at E8.25 is normally present in the anterior hindbrain with a posterior limit at r4 (arrowhead), but an Raldh2–/– embryo exhibits a posterior expansion of Irx3 expression to the hindbrain/spinal cord junction (bracket and asterisk).

View larger version (57K): [in a new window]   Fig. 5. vHnf1 is required to define r4/r5 gene expression boundary in mouse embryos. (A,B) Dorsal and lateral views of Hoxb1 expression in wild-type and conditional vHnf1–/– embryos. The normal r4/r5 boundary of Hoxb1 expression is indicated in wild-type embryos. vHnf1–/– embryos exhibit expansion of Hoxb1 expression posterior to the r4/r5 boundary in the hindbrain (brackets).

View larger version (76K): [in a new window]   Fig. 6. Maternal dietary RA supplementation rescues r3/r4 and r4/r5 gene expression boundaries in Raldh2–/– embryos. Embryos were subjected to low-dose maternal dietary RA supplementation from E6.75-E8.0 (A,F-H) or from E6.75-E8.25 (B-E). (A) RARE-lacZ expression in RA-treated wild-type and Raldh2–/– embryos at E8.0. The Raldh2–/– embryo exhibits an anterior border of RARE-lacZ expression similar to that observed in a wild-type littermate, although the intensity of staining was less than that observed in the wild-type embryo. (B-E) RA treatment of wild-type and Raldh2–/– embryos followed by analysis at E8.75 demonstrates that RA administration rescues r4 expression of Hoxb1 (B,C) and Cyp26c1 (D,E) in Raldh2–/– embryos. RA-treated wild-type embryos exhibited relatively normal expression of Hoxb1 and Cyp26c1. All embryos were also simultaneously stained for Raldh2 expression, which appeared only in wild-type embryos as indicated. (F-H) RA treatment of Raldh2–/– embryos followed by analysis at E8.0 results in posterior hindbrain expression of vHnf1 (F-G) and relatively normal r4/r5 restriction of Irx3 (H). Embryos were double stained for Raldh2 expression, which was observed only in wild-type embryos (brackets).

View larger version (88K): [in a new window]   Fig. 7. Teratogenic dose of RA induces anterior shifts in the Cyp26c1 expression domain and the Hoxb1/vHnf1 expression boundary. All embryos were subjected to a 20 mg/kg dose of RA at E7.5, then analyzed 18 hours later at E8.25 (see Figs 2, 3, 4 for untreated controls). (B,C,F,H) Lateral view; (G,I) dorsal view; (A) ventral view; (D,E) anterior view. (A) RARE-lacZ expression is observed to the anterior tip of some embryos (left embryo), but in other embryos RARE-lacZ expression is reduced at the anterior tip (upper arrowhead in right embryo). RARE-lacZ expression in all embryos is greatly reduced posteriorly in the tailbud (lower arrowheads) when compared with the trunk. (B-C) Cyp26a1 expression is observed in the tailbud (arrowhead) similar to untreated embryos (MacLean et al., 2001), whereas Cyp26c1 expression has been shifted to the anterior-most region of the embryo (arrowhead). A lateral view shows that most Cyp26c1 expression is neural, with mesodermal expression being greatly reduced. (D,E) Comparison of anterior expression patterns of RARE-lacZ and Cyp26c1 demonstrates that the abnormal domain of Cyp26c1 expression lies in the region where RARE-lacZ expression first clears anteriorly. (F,G) The anterior domain of Hoxb1 expression (arrowhead) has been shifted further anteriorly from where it should normally reside at r4. (H-I) The anterior border of vHnf1 expression (arrowhead) has been shifted further anteriorly. This abnormal anterior border of vHnf1 expression lies posterior to the abnormal expression domains of Cyp26c1 and Hoxb1.

View larger version (23K): [in a new window]   Fig. 8. Role of shifting RA boundaries during posterior hindbrain segmentation. A model for RA action during mouse hindbrain development is shown. Our findings demonstrate that RA generated by Raldh2 in the paraxial mesoderm travels anteriorly to presumptive r3 and r4 during establishment of Hoxb1 expression, but that this is very transient. Initially, RA forms an early anterior boundary at r2/r3 (next to the r2 border of Cyp26a1 expression) followed soon after by a late anterior boundary at r4/r5 (next to the r4 border of Cyp26c1 expression). RA is therefore present in r3/r4/r5 to directly regulate Hoxb1 induction and repression through previously described 3' and 5' RAREs (Marshall et al., 1994; Studer et al., 1994). We also demonstrate that vHnf1 requires RA for posterior hindbrain expression, and that vHnf1 is needed to limit the posterior extent of Hoxb1 expression to help establish the r4/r5 expression boundary for Hoxb1. Thus, RA acts directly to induce Hoxb1 expression and then RA acts both directly and indirectly (through induction of vHnf1) to restrict Hoxb1 expression to r4. Also shown is a mutual repression between Irx3 and vHnf1 at the r4/r5 boundary, which has been demonstrated in zebrafish (Lecaudey et al., 2004) and is supported by our findings in mouse.