Clinical and functional heterogeneity associated with the disruption of retinoic acid receptor beta

Abstract

Purpose: Dominant variants in the retinoic acid receptor beta (RARB) gene underlie a syndromic form of microphthalmia, known as MCOPS12, which is associated with other birth anomalies and global developmental delay with spasticity and/or dystonia. Here, we report 25 affected individuals with 17 novel pathogenic or likely pathogenic variants in RARB. This study aims to characterize the functional impact of these variants and describe the clinical spectrum of MCOPS12.

Methods: We used in vitro transcriptional assays and in silico structural analysis to assess the functional relevance of RARB variants in affecting the normal response to retinoids.

Results: We found that all RARB variants tested in our assays exhibited either a gain-of-function or a loss-of-function activity. Loss-of-function variants disrupted RARB function through a dominant-negative effect, possibly by disrupting ligand binding and/or coactivators' recruitment. By reviewing clinical data from 52 affected individuals, we found that disruption of RARB is associated with a more variable phenotype than initially suspected, with the absence in some individuals of cardinal features of MCOPS12, such as developmental eye anomaly or motor impairment.

Conclusion: Our study indicates that pathogenic variants in RARB are functionally heterogeneous and associated with extensive clinical heterogeneity.

Keywords: Dystonia; Global developmental delay; Microphthalmia; Retinoic acid; Retinoic acid receptor beta.

Figure 1.. Localization of RARB variants

Schematic representation of the position of bi-allelic (blue) and dominant (novel in black, previously reported in green) coding variants along the RARB protein. Also shown are the positions of the 12 α-helices (H1-H12) and the three β-turns (B1-B3) of the protein. DBD, DNA binding domain; LBD, ligand binding domain.

Figure 2.. Transcriptional response of RARB variants to RA ligands

(A) Human embryonic kidney HEK293 cells were transfected with Gal4 fusion plasmids of wild-type human RARB or the indicated genetic variants in the presence of UAStkLuc reporter luciferase gene construct. Cells were then treated with 1 μM atRA, 1 μM 9-cis RA, or vehicle (DMSO; 1/1,000, v/v) for 16 hr. Luciferase values were normalized to β-galactosidase activity and expressed as a fold response compared to vehicle-treated cells set at 1.0 for each mutant. Empty Gal4-transfected cells were used as a negative control. Data (mean ± SEM) were derived from at least four independent experiments performed in triplicate. *, p < 0.05; **, p < 0.01 versus wild-type RARB response to each respective RA ligand. (B) The same assay as in (A) was used to assess the impact of increasing concentrations of atRA on the transcriptional response of loss-of-function variants. *, p < 0.05; **, p < 0.001 versus untreated mutant RARB response. (C) Similar as in (B) except that GOF variants were tested to decreased concentrations of atRA. *, p < 0.05; **, p < 0.001 versus wild-type RARB treated in the same condition.

Figure 3.. Dominant-negative effects of RARB variants

(A) Endogenous response of HEK293 cells to atRA. Cells were transfected with a RAREtkLuc reporter luciferase gene construct or an empty tkLuc reporter as a negative control, and then treated with 1 μM atRA or vehicle (DMSO; 1/1,000, v/v) for 16 hr. Luciferase values were normalized to β-galactosidase activity and expressed as a fold response compared to vehicle-treated RAREtkLuc-transfected cells set at 1.0. (B to F) HEK293 cells were seeded in 24-well plates and transfected as in A with a RAREtkLuc reporter in the presence of increasing concentrations of plasmids encoding wild-type RARB or each indicated variant. Luciferase values were normalized to β-galactosidase activity and dominant-negative activity was determined as the % change of endogenous RA response determined as in A and set at 100%. Data (mean ± SEM) were derived from at least four independent experiments performed in triplicate. *, p < 0.05; **, p < 0.005 versus wild-type RARB-transfected cells in the same conditions. (G) The C98,101A mutation in the DNA binding domain abolished RARB response to RA. RAR triple KO MEFs were transfected or not (control) with wild-type or C98,101A mutated RARB in the presence of the RAREtkLuc reporter and treated with RA (1 μM atRA, 16h). Luciferase values were expressed as fold (mean ± SEM) compared to untreated cells. Data were derived from at least four independent experiments performed in triplicate. *, p < 0.01. (H) DBD integrity is indispensable for dominant-negative activity of LOF variants. HEK293 cells were transfected with each indicated LOF variant in the context or not of the C98,101A mutation, and treated as in G in the presence of the RAREtkLuc reporter. Dominant negative activity was determined as the % change of endogenous RA response set at 100%. Data (mean ± SEM) were derived from at least four independent experiments performed in triplicate. *, p < 0.05.

Figure 4.. Molecular dynamics simulation of RARB variants structural changes

(A) Root mean square deviation traces over 100 ns all-atom molecular dynamics (MD) trajectories have been performed for p.(Met290Arg) and p.(Leu402Pro) variants and compared to wild-type RARB. Results show that all receptor forms were stable under simulation. (B) The p.(Met290Arg) variant results in repacking of side chain residues in the binding pocket: wild-type structure used to set up simulation (blue, PDB ID 4DM8) versus a frame from the end of the mutant simulation (orange). Arg290 extends towards the ligand pocket, pushing aside Phe295, a key contact residue. (C) The p.(Leu402Pro) variant changes the shape of coactivator binding surfaces, removing important hydrophobic contacts to the co-activator helix (top): wild-type structure used to set up simulation (blue, PDB ID: 4DM8) versus a frame from the end of the mutant simulation (grey).