esrra is expressed in cartilaginous regions during zebrafish embryogenesis.

(A–L) Embryos at the indicated stages were subjected to in situ hybridization to analyse expression of esrra, sox9a, sox9b and col2a1. Arrows in (E–L) indicate pharyngeal arches where cartilage development occurs. Note that expression of esrra is largely overlapping with that of sox9a, sox9b and col2a1. (M–R) Embryos were injected with either MOctrl or MOesrra, raised to 48 hpf or 72 hpf as indicated, and analysed by in situ hybridization for expression of myod and acta2. myod expression in somites (arrows) is drastically decreased as muscle becomes differentiated from 48 hpf to 72 hpf in control embryos (compare M and O), while it sustains in MOesrra-injected embryos (compare N with P). In contrast, robust expression of acta2, a differentiated muscle marker, in control embryos is almost lost in MOesrra-injected embryos (compare Q with R). In addition, expression of myod in the pharyngeal arches (red asterisks) does not change from 48 hpf to 72 hpf in MOesrra-injected embryos, while myod in control embryo is expressed in a different subset of cells at 72 hpf as compared to that in 48 hpf. Embryos are shown in lateral views with anterior to the left except I–L where embryos are shown in ventral views.

The structure of craniofacial cartilage is disorganised upon knockdown of esrra.

(A–H) Embryos were injected with either MOctrl or MOesrra and subject to alcian blue staining at 3- or 4 dpf as indicated. Craniofacial cartilage, especially ceratohyal (CH) and ceratobranchial (CB) cartilage, are abnormally developed in MOesrra-injected embryos when compared to control embryos. Note that the structure of neural cranium is well organised although smaller in size in MOesrra-injected embryos (compare C and G with D and H). (I–K) sox10:GPF (I,J) or fli1:GFP (K,L) transgenic embryos were used to confirm the cartilage defects upon knockdown of MOesrra. (M–P) Human ESRRa mRNA (hESRRa) was co-injected with MOesrra into 1-cell stage of embryos which were subject to alcian blue staining. Embryos co-injected with hESRRa and MOesrra show partial restoration of CB (asterisk in O) and repress disorientation of CH induced by MOesrra alone (compare O with N). Misexpression of hESRRa at 100 pg does not impair cartilage development (P, see the text for detail). Embryos are shown in ventral view with anterior to the left.

ESRRa regulates development of neural crest cells for cartilage development.

(A–J) Embryos at 1-cell stage were injected with either MOctrl or MOesrra and analysed for expression of dlx2a or dhand by in situ hybridization. Expression of dlx2a at 24 hpf shows a similar pattern between control and MOesrra-injected embryos while expression of both dlx2a and dhand at 30 hpf shows a slight decrease in the branchial arches (white arrows in D and F) in MOesrra-injected embryos. From 2 dpf to 3 dpf, expression of dlx2a becomes largely restricted to pharyngeal cartilage in control embryos (arrows in I), while it is significantly disorganised in the pharyngeal arches of MOesrra-injected embryos (arrow in J). (K–N) Expression of myod from 2 dpf to 3 dpf shows a significant change as the pharyngeal regions in control embryos undergo growth and differentiation (compare K to M). However, expression of myod at 3 dpf remains strikingly similar to that at 2 dpf (compare L and N), except few additional elements being developed in MOesrra-injected embryos. Embryos are shown in either lateral (A–J) or ventral views (K–N).

ESRRa regulates expression of genes essential for cartilage development.

(A–N) Embryos at 1-cell stage were injected with either MOctrl or MOesrra, raised and analysed at the indicated stages for expression of sox9a, sox9b,sox5, sox6 and col2a1 by in situ hybridization. At 2- and 3 dpf, expression of sox9 (A–H) and col2a1 (M and N) is specifically decreased in the branchial arches in MOesrra-injected embryos, although it remains comparable in other expression domains as compared to that in control. Expression of sox5 and sox6 at 2 dpf is also substantially decreased in MOesrra-injected embryos as compared to that in control (I–L). White arrows or bars indicate branchial arches and insets shown are magnified views of the branchial regions.

ESRRa regulates survival but not proliferation of cartilaginous cells.

(A–F) 1-cell stage of embryos were injected with either MOctrl or MOesrra, raised and subjected to acridine orange stain to determine apoptotic cells at the stages indicated. Moesrra-injected embryos display a significant number of apoptotic cells in the head (arrows) and body trunk (insets) at the all observed stages as compared to controls. A combination of MOesrra and MOp53 does not suppress cell apoptosis. (G,H) sox10:GFP embryos were injected similarly to A–F, raised to the indicated stages, and processed to determine proliferation of cartilaginous cells by phosphorylated histone H3 immunostaining (pH3, red signal) in pharyngeal regions. White arrows indicate proliferating chondrogenic cells marked by both green and red. Scale bar is 10µm. (I) The total number of proliferating cells in the pharyngeal arches is similar between control and MOesrra-injected embryos (6.5+/1.3 vs. 6.8+/1.0 cells in average, respectively; n = 27). Also, the number of proliferating cartilaginous cells (doubly positive for both green and red) is also similar between control and MOesrra-injected embryos (1.5+/1.0 vs. 1.0+/1.2 cell in average, respectively; n = 27).

ESRRa directly binds ESRRE consensus elements located upstream of sox9b.

(A) Consensus DNA sequences to where ESRRa is known to bind (ESRRE consensus) are identified upstream of both sox9a and sox9b. In addition, esrra also contains an ESRRE consensus in the proximal promoter region (esrra 70 b). (B) Embryos at 1-cell stage were injected with hESRRa mRNA, raised and collected at 36 hpf for chromatin immunoprecipitation using anti-ESRRa antibody. Significant binding of hESRRa is detected at the ESRRE consensus sites located upstream of sox9b (8.8~8.5 kb denoted as sox9b-1and 3.6 kb as sox9b-2), but not detected at the site found upstream of sox9a (6.5 kb). hESRRa also binds to the proximal promoter of esrra itself. Statistical analysis of pair-wise samples was performed using IBM SPSS Statistics 22 software. Values with p < 0.05 were considered to be statistically significant (indicated by asterisks). (C–F) Embryos were injected with MOesrra together with sox9a or sox9b mRNA, and subject to alcian blue staining at 4 dpf. White arrows indicate correctly-oriented ceratohyal cartilage, and white asterisks point to ceratobranchial arches. Note that overexpression of sox9b efficiently rescues defective cartilage induced by esrra knockdown, while overexpression of sox9a only partially rescues proper formation of ceratohyal cartilage.