Nr6a1 controls Hox expression dynamics and is a master regulator of vertebrate trunk development

Abstract

The vertebrate main-body axis is laid down during embryonic stages in an anterior-to-posterior (head-to-tail) direction, driven and supplied by posteriorly located progenitors. Whilst posterior expansion and segmentation appears broadly uniform along the axis, there is developmental and evolutionary support for at least two discrete modules controlling processes within different axial regions: a trunk and a tail module. Here, we identify Nuclear receptor subfamily 6 group A member 1 (Nr6a1) as a master regulator of trunk development in the mouse. Specifically, Nr6a1 was found to control vertebral number and segmentation of the trunk region, autonomously from other axial regions. Moreover, Nr6a1 was essential for the timely progression of Hox signatures, and neural versus mesodermal cell fate choice, within axial progenitors. Collectively, Nr6a1 has an axially-restricted role in all major cellular and tissue-level events required for vertebral column formation, supporting the view that changes in Nr6a1 levels may underlie evolutionary changes in axial formulae.

Fig. 1. Nr6a1 expression level positively correlates with the trunk vertebral number and is regulated by known mediators of the trunk-to-tail transition.

A Nr6a1 is expressed broadly at embryonic day (E)8.5, with expression becoming progressively excluded from the posterior growth zone starting E9.5. By E10.5 Nr6a1 exhibits tissue-specific expression, followed by complete tissue clearance by E12.5. B Violin plot illustrates the Log₂ normalised count, which was derived from the raw counts from each cells divided by their size factors, of Nr6a1 in cells of the E6.5 to E8.5 mouse embryo, revealing robust expression within neuromesodermal progenitors (NMP), caudal mesoderm and caudolateral epiblast. Data derived from ref. [41]. C Expression level of Nr6a1 within the tailbud positively correlates with the number of trunk vertebrae that form later in development. Quantitative PCR analysis of Nr6a1 within tailbud tissue isolated from Gdf11 and miR-196 single and compound mutant embryos, represented as a fold change compared to wildtype (defined as 1). The number of biologically independent samples assessed: WT E9.5 n = 3, E10.5 n = 4; Gdf11^+/− E9.5 n = 3, E10.5 n = 3; miR-196a2^−/−;b^−/− E9.5 n = 3, E10.5 n = 4; Gdf11^+/−; miR-196a1^−/−;a2^−/−;b^−/− E9.5 n = 3, E10.5 n = 4; Gdf11^−/− E9.5 n = 3, E10.5 n = 2; Gdf11^−/−; miR-196a1^−/−;a2^−/−;b^−/− E9.5 n = 3, E10.5 n = 2; Error bar represents standard deviation. Source data are provided as a Source Data file. The number of additional thoraco-lumbar vertebrae observed across the allelic deletion series is indicated. D Whole mount in situ hybridisation reveals that the quantitative increase of Nr6a1 in E10.5 Gdf11^−/− tailbud tissue relative to wildtype reflects a lack of timely clearance from the posterior growth zone. Consistent spatial changes were observed in 2 biologically independent samples/genotype. Red arrow indicates the neural tube.

Fig. 2. Nr6a1 is required for trunk elongation and timely activation of sacro-caudal identity.

A Skeletal preparation of E18.5 embryos following conditional deletion of Nr6a1 revealed a dose-dependent reduction in the number of thoraco-lumbar elements. The number of biologically independent samples assessed: Nr6a1^+/+ n = 8; Nr6a1^+/− n = 10; Nr6a1^−/− n = 14. B In WT skeletons, removal of pelvic bones allowed visualisation of transverse process morphology characteristic of lumbar (L) and sacral (blue S) elements. In two representative TCre^ERT2;Nr6a1^flx/flx embryos, positioning of the normal sacral region is defined by the presence of rudimentary pelvic bones (blue S) and lumbar elements that have assumed sacral identity marked (red S). Red asterisk indicates lateral fusions between two adjacent elements. LHS = left-hand side; RHS = right-hand side. C Quantification of thoraco-lumbar and sacro-caudal vertebral number following conditional deletion of Nr6a1 alleles. Raw data is presented in the upper plots. Mean differences relative to wildtype (+/+) are presented in the lower plots as bootstrap sampling distributions. The mean difference for each genotype is depicted as a black dot and 95% confidence interval is indicated by the ends of the vertical error bar. D Whole mount in situ hybridisation for Hoxd11 in E10.5 embryos revealed a rostral shift in the anterior limit of expression in TCre^ERT2;Nr6a1^flx/flx embryos compared to control. Ectopic Hoxd11 expression in TCre^ERT2;Nr6a1^flx/flx embryos is marked with a red line in lateral view. In dorsal view, the normal anterior limit is marked with a red dotted line, with ectopic Hoxd11 expression restricted to the paraxial mesoderm. Consistent spatial changes were observed in 3 biologically independent samples/genotype.

Fig. 3. Nr6a1 function within somitic tissue is restricted to the trunk region.

A Analysis of caudal elements in wildtype and TCre^ERT2;Nr6a1^flx/flx and Nr6a1^flx/flx embryos (n = 13 and n = 8 biologically independent samples, respectively) revealed no difference in vertebrae morphology or number. B Whole mount in situ hybridisation for the somite marker Uncx4.1 revealed a rostral shift in hindlimb positioning in TCre^ERT2;Nr6a1^flx/flx embryos compared to control, somite number is marked in the right panel. Additionally, somite morphology was specifically disrupted in the future thoraco-lumbar region of TCre^ERT2;Nr6a1^flx/flx embryos, returning to normal immediately after the hindlimb bud. Co-expression analysis of Gdf11 (red arrows) revealed no change in expression in the tailbud between genotypes. Consistent expression patterns were observed in 2 biologically independent samples/genotype.

Fig. 4. Nr6a1 suppresses posterior Hox gene expression.

A Quantitative PCR analysis of posterior Hox gene expression within tailbud tissue isolated from TCre^ERT2 controls, TCre^ERT2;Nr6a1^+/flx and TCre^ERT2;Nr6a1^flx/flx embryos at E9.5, n = 3 biologically independent samples per genotype. Statistical analysis of differences between genotypes was performed using an unpaired two-sample Welch’s t test, p values are provided in the figure. Source data are provided as a Source Data file. B Whole mount in situ hybridisation revealed a spatial expansion of Hoxb13 expression in TCre^ERT2;Nr6a1^+/flxembryos compared to wildtype, at both E9.5 and E10.5. Red arrow at E9.5 indicates ectopic Hoxb13 expression. TBM = tail bud mesoderm; NT = neural tube.

Fig. 5. Prolonged Nr6a1 activity is sufficient to increase thoraco-lumbar count and sustain trunk expression signatures.

A Skeletal preparation of E18.5 embryos, dorsal view, revealed an increase in the number of thoraco-lumbar elements in Cdx2P:Nr6a1 embryos (n = 2 independent samples) compared to wildtype (n = 8 independent samples). C = cervical, T = thoracic, L = lumbar, S = sacral and Cd = caudal vertebrae. B Higher magnification, lateral view, of the lumbar region in embryos from (A) revealed the anapophysis (red arrow) normally present on the first 3 lumbar elements in wildtype was now present on 7 of the 8 lumbar elements in Cdx2P:Nr6a1 embryos. Experimental numbers as per 5A. C RNAseq analysis of E10.5 wildtype and Cdx2P:Nr6a1 tailbuds revealed widespread changes in Hox expression downstream of Nr6a1. Results are presented as a log2-transformed fold change in Cdx2P:Nr6a1 samples relative to wildtype, n = 2 for Cdx2P:Nr6a1 and n = 4 for wildtype. Only those Hox genes with a false discovery rate (FDR) < 0.05 are displayed, and are colour-coded based on the axial region where the Hox protein functions. Source data are provided as a Source Data file. D Whole mount in situ hybridisation revealed reduced expression of Hoxd12 within the tail of E10.5 Cdx2P:Nr6a1embryos compared to wildtype. Consistent spatial changes were observed in 2 biologically independent samples/genotype. E Venn diagram overlay of differentially expressed genes (DEGs) in Cdx2P-Nr6a1 (this work) and Gdf11^−/−  tailbud tissue reveals substantial overlap. Source data are provided as a Source Data file. F Analysis of the 151 co-regulated genes identified in (D), presented here as log2-transformed fold change in Cdx2P:Nr6a1 samples relative to wildtype, revealed that 136 genes (90%) displayed the same direction of regulation in both genetically altered backgrounds. Red and blue dots indicate upregulated and downregulated genes in Gdf11^−/− tailbuds, respectively. Source data are provided as a Source Data file.

Fig. 6. Gdf11 signalling terminates Nr6a1-dependent trunk elongation.

A Skeletal preparation of E18.5 embryos, lateral view, revealed a dose-dependent rescue of Gdf11^−/− thoraco-lumbar expansion following conditional deletion of Nr6a1 alleles. The number of biologically independent samples assessed: Nr6a1^+/+;Gdf11^−/− n = 9; Nr6a1^+/−;Gdf11^−/− n = 5; Nr6a1^−/−;Gdf11^−/− n = 7. T = thoracic, L = lumbar. B Quantification of thoracic, lumbar and sacro-caudal vertebral number across genotypes. Raw data is presented in the upper plots. Mean differences were calculated relative to Gdf11^−/−;Nr6a1^+/+ and presented in the lower plots as bootstrap sampling distributions. The mean difference for each genotype is depicted as a black dot and 95% confidence interval is indicated by the ends of the vertical error bar.

Fig. 7. Nr6a1 and Gdf11 signalling cooperate in the control of Hox cluster transitions.

A, B Differential Hox expression kinetics observed during in vitro differentiation of Nr6a1 and wildtype ESCs. Source data are provided as a Source Data file. A RNAseq analysis of trunk Hox gene expression (Hox6-9) in Nr6a1^−/− (red) and wildtype (blue) cells. In vitro differentiation protocol schematised, analysis timepoints marked with asterisk and data represented as counts per million (CPM). Only those genes showing a false discovery rate (FDR) < 0.05 are visualised. B RNAseq analysis of posterior Hox gene expression (Hox10-13) in Nr6a1^−/− (red) and wildtype (blue) cells. In vitro differentiation protocol schematised, analysis timepoints marked with asterisk and data represented as counts per million (CPM). Only those genes showing a false discovery rate (FDR) < 0.05 are visualised. G = addition of Gdf11.

Fig. 8. Nr6a1 activity regulates neural vs mesodermal fate choice in vitro and in vivo.

A Heightened neural (red) and reduced mesodermal (blue) gene expression signatures were observed following prolonged maintenance of Nr6a1 within the mouse tailbud. RNAseq analysis was presented as a log2-transformed fold change in Cdx2P:Nr6a1 samples relative to wildtype, n = 2 for Cdx2P:Nr6a1 and n = 4 for wildtype. Neural and mesodermal gene lists were selected based on in vivo single-cell RNAseq analysis, and only genes with a false discovery rate (FDR) < 0.05 are displayed. Source data are provided as a Source Data file. B Heightened mesodermal (blue) and reduced neural (red) gene expression signatures were observed in Nr6a1^−/− in vitro-derived NMPs compared to wildtype. In vitro differentiation protocol schematised, analysis time-point (d3) marked with asterisk. RNAseq analysis is presented as a log2-transformed fold change in Nr6a1^−/− samples relative to wildtype, n = 3/genotype. Neural and mesodermal gene lists as per above, and only genes with a false discovery rate (FDR) < 0.05 are displayed. Source data are provided as a Source Data file. C Flow cytometry analysis of Sox2 and T/Bra protein expression within cells of the E9.0 tailbud, collected from WT (n = 4 biologically independent samples) and TCre^ERT2;Nr6a1^flx/flx (n = 9 biologically independent samples) embryos. Single Sox2-positive, single T/Bra-positive, or dual positive cells are presented as a percentage of all single, viable cells. Statistical comparison between genotypes was performed using an unpaired t-test with Welch’s correction, p-values are provided in the figure.

Fig. 9. Schematic summary highlighting the multiple roles of Nr6a1 within the tailbud.

Blocks represent somites, pink = trunk-forming somites and green = tail-forming somites.