Feedback regulation of retinaldehyde reductase DHRS3, a critical determinant of retinoic acid homeostasis

Abstract

Retinoic acid is crucial for vertebrate embryogenesis, influencing anterior-posterior patterning and organogenesis through its interaction with nuclear hormone receptors comprising heterodimers of retinoic acid receptors (RARα, β, or γ) and retinoid X receptors (RXRα, β, or γ). Tissue retinoic acid levels are tightly regulated since both its excess and deficiency are deleterious. Dehydrogenase/reductase 3 (DHRS3) plays a critical role in this regulation by converting retinaldehyde to retinol, preventing excessive retinoic acid formation. Mutations in DHRS3 can result in embryonic lethality and congenital defects. This study shows that mouse Dhrs3 expression is responsive to vitamin A status and is directly regulated by the RAR/RXR complex through cis-regulatory elements. This highlights a negative feedback mechanism that ensures retinoic acid homeostasis.

Keywords: homeostasis; metabolism; negative feedback; nuclear hormone receptors; retinoic acid receptor; retinoids; transcriptional regulation; vitamin A.

Fig. 1

Dhrs3 expression in tissues from adult mice and embryos. Adult male mice and pregnant dams (for embryo analysis) were maintained on a vitamin A‐sufficient (VAS) diet. Tissue Dhrs3 mRNA expression was measured by RT‐qPCR and its expression was normalized to the expression of housekeeping genes and represented as a percent of the expression observed in adult liver. After the adult liver, Dhrs3 is most highly expressed in lung, kidney, adipose tissue, and testes, with lower levels in other tissues and embryos. Values are presented as mean ± SD (n = 4).

Fig. 2

Changes in the expression of Dhrs3 in response to different levels of vitamin A in the diet. Adult male mice were maintained for 16 weeks from weaning on chow (15 IU vitamin A/g diet), VAS (4 IU vitamin A/g diet), or VAD (vitamin A‐deficient) diet. The expression of Dhrs3 in tissues of mice maintained on diets with varied vitamin A content was examined by RT‐qPCR normalized to the expression seen in mice maintained on chow. For each tissue, the levels of Dhrs3 expression were normalized to the expression of housekeeping genes (Hprt and B2m) and represented as fold change in comparison to the levels of expression in the same tissue from mice maintained on chow. Comparisons between genotypes and groups were made using two‐way ANOVA followed by post hoc tests. The symbols indicate statistical significance: chow versus VAD diet using *P < 0.05, ***P < 0.001; VAS versus VAD diet using ###P < 0.001. Data are shown as mean ± SD (n = 10).

Fig. 3

Dhrs3 is a direct target of RAR/RXR. MC3T3 were pretreated with 5 μg·mL⁻¹ CHX or DMSO for one hour following which gene expression was induced with 50 nm TTNPB (a stable pan‐RAR agonist) in the presence or absence of CHX. After 3 h, total RNA was extracted, reverse transcribed, and gene expression analysis was performed by RT‐qPCR and normalized to the expression of housekeeping genes and represented as a fold change in comparison to the levels of expression in cells treated with DMSO vehicle control. (A) Dhrs3 expression levels in MC3T3 cells treated with DMSO alone, TTNPB alone, CHX alone, or TTNPB in the presence of CHX (TTNPB + CHX). CHX has a modest impact on TTNPB‐induced expression of Dhrs3 suggesting that its expression can be induced by RAR/RXR in the absence of translation. (B) Similar results were obtained for Rarb a known direct target of RAR. Statistical analysis was performed using one‐way ANOVA with Holm–Sidak post hoc tests. Data are represented as the relative fold change ± standard deviation (SD) versus DMSO‐only treated control (n = 3); *P < 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Fig. 4

Regulation of Dhrs3 by RAR‐controlled cREs. (A) Diagram showing the exon structure of the principal isoform of mouse DHRS3 (302 amino acids, CCDS18912.1), transcript ENSMUST00000154208.8 on mouse chromosome 4 (144619647‐144654779). Five potential RARα‐binding regions identified by ChIP‐qPCR are depicted as blue boxes beneath the genomic map. (B) ChIP‐qPCR results show fold enrichment of DNA fragments copurified with anti‐RARα antibody for each targeted Dhrs3 region and for a known RAR‐binding site in Rarb, relative to control IgG. Notable enrichment is observed in regions 1, 2, and 5 of Dhrs3, and the known Rarb cRE. (C) Detailed view of upstream regions 1, 2 and 5 from (A), highlighting the sequence and location of cRE1, cRE2 and cRE5 including potential RAREs predicted by the TRANSFAC and JASPAR databases. Distances from the transcription start site are indicated by the map of exon 1 at bottom. The insets display the sequence conservation of cRE1, cRE2 and cRE5 in mouse in human genomes, aligned with several RARE consensus motifs, including a previously proposed DR5* RARE in cRE2. DRs in RARE motifs are presented in capital letter and spacers are in lowercase letters, and TSA sequence are underlined. (D) eRNA levels expressed from cRE1, 2 and 5 as measured by RT‐qPCR after 3 h of treatment with 100 nm TTNPB (grey bars) or DMSO control (black bars) in MC3T3‐E1 subclone 4 cells. Expression levels are normalized to Hprt and B2m mRNA. Statistical analysis was conducted using Student's t‐test. Data are presented as relative fold change ± standard deviation (SD) (n = 3); *P < 0.05, **P ≤ 0.01, ***P ≤ 0.001.