Evolution of the essential gene MN1 during the macroevolutionary transition toward patterning the vertebrate hindbrain

Abstract

The tight link between brain and skull formation is a fundamental aspect of vertebrate evolution and embryogenesis. Their developmental synchronization is essential for structural and functional integration. The brain and skull shape coevolution is evident along the vertebrate phylogeny; however, the genetic basis underlying their close evolutionary and developmental relationship remains little explored. Here, we reveal the evolution and function of the MN1 gene that was previously found to be associated with significant shape variation in the mouse skull and the formation of cranial bones. We show that the vertebrate MN1 gene evolved from an ancestral deuterostome sequence. In vertebrates, the MN1 gene structure, synteny, and spatiotemporal expression pattern are remarkably conserved, indicating that the gene carries out a core function. Using a newly generated mouse knock-out model, we demonstrate in vivo that Mn1 integrated into an ancient molecular machinery and controls the expression of the Cyp26 genes in the developing hindbrain, thereby tuning the retinoic acid levels and patterning of the developing central nervous system. This study thus showcases the emergence of a novel gene function from an ancestral sequence and its role in generating a macroevolutionary innovation. The data expand our knowledge of brain and skull codevelopment and coevolution and highlight the role of this regulatory loop in craniofacial human syndromes.

Keywords: MN1; co-evolution and co-development of brain and skull; gene evolution; hindbrain patterning; retinoic acid.

Fig. 1.

Assessment of the evolutionary origin of the MN1 gene. (A) Summary of MN1 genes identified in Bilateria. Colored boxes indicate the presence of homologous MN1 genes, white boxes the absence of MN1, and dashed-colored boxes presumptive distant homologs. (B) Group-scale synteny analysis. Dendrogram indicating the relative evolutionary relationships between species and events during deuterostome evolution (Left). Chromosomal-level synteny analysis of representative species covering major deuterostome lineages (Right). The human MN1 genomic region spanning 50 upstream and downstream genes was highlighted in red. Asterisks denote the chromosomes in which MN1 homologs are located. Chromosomes are scaled by gene rank order.

Fig. 2.

Expression pattern of MN1 during embryogenesis across gnathostome species. MN1 expression during mouse (A), chicken (B), small-spotted catshark (C), and zebrafish (D) development. Note the sparse Mn1 expression signal in migrating cranial neural crest cells (CNCCs) laterally to the cephalic portion of the neural tube at E8.5. In chicken, MN1 expression can be observed in the presumptive hindbrain at HH7, before cranial neural crest cell migration (HH9-10). MN1 transcripts are not detected in the trunk or tail regions of any of the analyzed catshark embryos. Zebrafish possess two mn1 paralogs (mn1a and mn1b) with spatially distinct expression profiles. Both mn1 paralogs are expressed in the CNS and partially overlap in the developing hindbrain during the early stages (14 to 24 hpf). In the consecutive stages (33 to 48 hpf), mn1a exhibits more restricted expression within the facial mesenchyme while mn1b remains expressed in the CNS. (Scale bar, 500um.) At least three embryos were assayed per species and stage.

Fig. 3.

Gene expression analysis in WT and Mn1^−/− embryos. (A) Overview of the RNA-seq experimental design. Mn1^+/+ and Mn1^−/− embryos were collected from Mn1^+/- intercrosses. n = 4 per genotype and embryonic stage were included in the analysis. The outliers were removed (SI Appendix, Fig. S15), and a minimum n = 3 per condition were further considered in bioinformatic analyses. At E12.5, brain and craniofacial structures were microdissected and processed independently to account for differences between nervous and mesenchymal tissues. (B) Volcano plots of statistically significant DEGs. Dots indicate genes significantly upregulated (red) or downregulated (blue) in Mn1^−/−. The most relevant genes are highlighted, with genes involved in retinoic acid degradation surrounded by a dashed box and Mn1 marked by a black arrowhead. (C) HCR images confirm the Mn1 expression differences between WT and Mn1^−/− embryos. (Scale bar, 500 μm.) At least three embryos were assayed per genotype and developmental stage.

Fig. 4.

Cyp26 gene expression is downregulated in the Mn1 mutants. Schematic representation and HCR of Cyp26 gene expression domains in the developing hindbrain in E8.5 (Top) and E9.5 (Bottom) Mn1 mutant embryos compared to their WT littermates. The cartoons represent WT (Left side) and Mn1 knock-out (Right side) embryos. White arrowheads point at the main differences in gene expression in r2-r5 segments. Note the decreased Cyp26b1 expression in r2/r5 and the absent Cyp26a1,c1 expression in r2/r4 in the Mn1 mutants. (Scale bars, 100 μm.) At least three embryos were assayed per genotype and developmental stage.

Fig. 5.

Hox gene expression during hindbrain segmental patterning is affected in Mn1 mutants. Schematic representation and HCR of HoxA and HoxB gene expression domains in the developing hindbrain in E8.25 (Top), E8.5 (Middle), and E9.5 (Bottom) Mn1 mutant embryos compared to their WT littermates. The Left side of the cartoons represents WT and the right side Mn1 knock-out embryos. White arrowheads point at the main differences in the rhombomeric segments. Note the reduced expression of HoxA2 and HoxB1 in the significantly smaller r3 segment in the Mn1 mutants. (Scale bar, 100 μm.) At least three embryos were assayed per genotype and developmental stage.

Fig. 6.

Summary of the Mn1 mutation effect in the developing hindbrain. (A) Schematic representation of the mouse developing hindbrain showing the main differences in the expression domains of genes involved in the RA/Hox program between WT and Mn1 mutants. (B) The interactions between signaling components known to orchestrate hindbrain segmentation and the suggested Mn1 integration into the ancient RA/Hox program. AP, anteroposterior; Mb, midbrain.