Tissue- and nuclear receptor-specific function of the C-terminal LXXLL motif of coactivator NCoA6/AIB3 in mice

Abstract

Although the LXXLL motif of nuclear receptor (NR) coactivators is essential for interaction with NRs, its role has not been assessed in unbiased animal models. The nuclear receptor coactivator 6 (NCoA6; also AIB3, PRIP, ASC-2, TRBP, RAP250, or NRC) is a coactivator containing an N-terminal LXXLL-1 (L1) and a C-terminal L2. L1 interacts with many NRs, while L2 interacts with the liver X receptor alpha (LXRalpha) and the estrogen receptor alpha (ERalpha). We generated mice in which L2 was mutated into AXXAL (L2m) to disrupt its interaction with LXRalpha and ERalpha. NCoA6(L2m/L2m) mice exhibited normal reproduction, mammary gland morphogenesis, and ERalpha target gene expression. In contrast, when treated with an LXRalpha agonist, lipogenesis and the LXRalpha target gene expression were significantly reduced in NCoA6(L2m/L2m) mice. The induction of Cyp7A1 expression by a high-cholesterol diet was impaired in NCoA6(L2m/L2m) mice, which reduced bile acid synthesis in the liver and excretion in the feces and resulted in cholesterol accumulation in the liver and blood. These results demonstrate that L2 plays a tissue- and NR-specific role: it is required for NCoA6 to mediate LXRalpha-regulated lipogenesis and cholesterol/bile acid homeostasis in the liver but not required for ERalpha function in the mammary gland.

FIG. 1.

Generation of NCoA6 L2m mutant ES cell lines and mice by homologous recombination. (A) The knock-in strategy for generation of the NCoA6-L2m allele. The WT NCoA6 allele (WA), targeting vector (TV), targeted mutant allele with the neo expression cassette (TA), and mutant allele postexcision of the neo cassette (L2m) are sketched. The locations of major restriction sites used for subcloning or Southern blotting and PCR primers P19, P20, and P21 used in genotyping analysis are indicated. Exons 8 to 10 are indicated by black rectangles. The LoxP sites are indicated by black triangles. The region used as a Southern blot probe is also indicated. L2, the LXXLL-2 motif in WT allele; L2m, the AXXAL-2 mutant in the targeting vector or in the mutant allele. (B and C) Identification of targeted ES clones. Southern blotting was performed first with HindIII-digested ES cell DNA (B) and then with BstEII-digested ES cell DNA (C). The 9-kb band in panel B and the 13.4-kb band in panel C are from the targeted NCoA6 allele with neo (nL2m), while the 7-kb and the 11.4-kb bands in panels B and C are from the WT NCoA6 allele (+). (D) Southern blot assay. Mouse tail DNA was digested with BstEII. The 13.4-kb band is from the NCoA6 nL2m allele. The 11.4-kb band is either the WT NCoA6 allele or the L2m allele without neo. (E) PCR-based genotype analysis. DNA samples were isolated from tail tips. The 550-bp band was amplified from the nL2m allele using primers P19 and P20. The 260-bp band was amplified from the L2m allele using primers P21 and P20. The 220-bp band was amplified from the WT allele (+) using primers P20 and P21.

FIG. 2.

Analysis of L2m mutant expression in knock-in mutant mice. (A) Sequence analysis of NCoA6 L2m mRNA. The expected cDNA sequences and peptide sequences of WT and L2m are listed on the top left. Total RNA samples were prepared from the livers of WT (+/+), NCoA6^+/L2m, and NCoA6^L2m/L2m mice and were reverse transcribed into cDNA as PCR templates. The DNA fragment spanning the L2m mutation region was amplified by PCR, and the PCR product was sequenced. The sequence data of WT (+/+), NCoA6^+/L2m, and NCoA6^L2m/L2m samples are presented. Note that NCoA6^+/L2m mice show both WT and L2m sequences. (B) RPA. Liver RNA samples (20 μg) were prepared from WT (+/+), NCoA6^+/L2m, and NCoA6^L2m/L2m mice. The RPA was performed with a ³²P-labeled antisense riboprobe complementary to a common region of NCoA6 WT and L2m mRNAs. Analysis of β-actin mRNA levels served as a loading control. C. Western blot assay. Cell nuclear lysates were prepared from the livers of WT (+/+), NCoA6^+/L2m, and NCoA6^L2m/L2m mice. Fifty μg of nuclear protein from each sample was separated by SDS-PAGE, and the blot was assayed using antibodies against NCoA6 and PARP. The nuclear protein PARP served as a loading control.

FIG. 3.

NCoA6^L2m/L2m mice exhibit normal steroid hormone-regulated mammary gland development. (A) Comparison of mammary gland morphogenesis between WT and NCoA6^L2m/L2m mice. The whole-mounted mouse mammary glands were prepared from WT and NCoA6^L2m/L2m female littermates at different developmental stages, including the 40-day-old virgin stage, gestation day 15, and lactation day 3. Five mice were examined for each stage per genotype group, and the morphologies of representative pairs are presented. LN, mammary gland lymph node. (B) Comparison of estrogen-stimulated mammary ductal growth in WT and NCoA6^L2m/L2m mice. WT and NCoA6^L2m/L2m female littermates (five in each group) were ovariectomized at the age of 17 days and then treated with either placebo or slow-releasing 17β-estradiol (E2) pellets for 15 days. The whole-mounted mammary glands were stained with carmine alum. Mammary glands with representative morphologies are presented. (C) Real-time RT-PCR analysis of PR mRNA expression. Intact WT and NCoA6^L2m/L2m female adult virgin mice without any treatment or WT and NCoA6^L2m/L2m female mice that were ovariectomized and treated with placebo or E2 were used. Each group had five mice. RNA samples were prepared from the third pairs of mammary glands for real-time RT-PCR analysis. The relative expression levels of PR mRNA were normalized to the 18S RNA levels. No statistical differences (P > 0.05) were observed between WT and NCoA6^L2m/L2m groups. (D) Western blot analysis. Fifty μg of protein isolated from the mammary glands of three pairs of WT and L2m littermates on gestation day 14 were used for Western blotting, and the blots were reacted with antibodies against NCoA6, ERα, a common region of PRA and PRB, and β-actin. β-Actin served as a loading control. Band intensity was quantified by densitometry. The average band intensities showed no statistical differences between the WT and NCoA6^L2m/L2m groups.

FIG. 4.

Differential effects of the L2m mutation on ERα and LXRα. (A) Western blot analysis. WT and NCoA6^L2m/L2m primary MEFs were transfected with LXRα expression plasmids and LXRE-tk-Luc reporter plasmids and treated with either vehicle or 10⁻⁶ M T0901317. Cell lysates with 50 μg protein were used in a Western blot analysis using antibodies against NCoA6, LXRα, and β-actin. The relative levels of endogenous NCoA6 in WT MEFs and NCoA6-L2m in NCoA6^L2m/L2m MEFs were comparable when normalized to β-actin. The LXRα levels were also similar regardless of T0901317 treatment. (B) Coimmunoprecipitation of NCoA6 and ERα. WT and L2m primary MEFs were transfected with ERα and treated with 17β-estradiol (E2; 10⁻⁸ M). Cell lysates were subjected to IP using NCoA6 antibody (α-NCoA6) or control IgG. Immunoprecipitates were separated by SDS-PAGE. Immunoblotting (IB) analysis was performed using NCoA6 and ERα antibodies. (C) ERα transfection assay. Primary WT and NCoA6^L2m/L2m MEFs were transfected with ERα, ERE-Luc, and pCMV-RL (Renilla luciferase) plasmids and treated with vehicle (V) or E2 for 24 h. The luciferase activity was assayed and normalized to RL activity. Data were obtained from three independent experiments using different batches of MEFs derived from littermate embryos. Data from three assays are presented as means ± standard deviations (SD). (D) Coimmunoprecipitation of NCoA6 and LXRα. WT and L2m primary MEFs were transfected with LXRα and treated with T0901317 (T; 10⁻⁶ M). Cell lysates were subjected to IP with NCoA6 antibody (α-NCoA6) or control IgG. Immunoprecipitates were subjected to IB analysis using NCoA6 and LXRα antibodies. (E) LXRα transfection assay. WT and NCoA6^L2m/L2m primary MEFs were transfected with LXRα, LXRE-tk-Luc reporter, and pCMV-RL plasmids and treated with vehicle (V) or T0901317 (T) for 24 h. The luciferase activity was assayed and normalized to RL activity. Data from three independent experiments are presented as means ± SD. ***, P < 0.001 by unpaired t test. (F) In vivo hepatocyte transfection assay for LXRα transactivation. LXRα, RXRα, LXRE-tk-luc, and pCMV-RL plasmids were mixed with the in vivo transfection reagent and injected into the tail vein of WT and NCoA6^L2m/L2m mice (n = 5 in each group). After 2 h, mice were orally administrated vehicle (V) or T0901317 (T). Mice were sacrificed 24 h post-plasmid injection. Luciferase activity in liver lysates was assayed and normalized to RL activity. Data are presented as means ± SD. ***, P < 0.001 by unpaired t test.

FIG. 5.

LXRα-mediated lipogenesis in the liver of NCoA6^L2m/L2m mice is impaired. (A) Real-time RT-PCR analysis of T0901317-induced gene expression in the liver. RNA samples were prepared from the livers of vehicle-treated (V) or T0901317-treated (T) WT and NCoA6^L2m/L2m male mice (five in each group). Relative concentrations of the LXRα, LXRβ, SREBP-1c, FAS, LPL, SCD-1, and Cyp7A1 mRNAs were measured by real-time RT-PCR, and their relative expression levels were normalized to the 18S RNA concentrations in the respective samples. Data are presented as means ± standard deviations (SD). *, P < 0.05; **, P < 0.01; ***, P < 0.001. (B) Hematoxylin and eosin-stained liver sections prepared from vehicle- or T0901317-treated WT and NCoA6^L2m/L2m mice. A higher number of unstained intracellular vacuoles (lipid droplets, indicated by green asterisks) were visible in WT liver sections compared with NCoA6^L2m/L2m liver sections after T0901317 treatment (right upper panel). Bar, 50 μm. (C) Oil Red O-stained liver cryosections prepared from vehicle- or T0901317-treated WT and NCoA6^L2m/L2m mice. The Oil Red O-stained lipid droplets were much larger in WT liver sections compared with that in NCoA6^L2m/L2m liver sections after T0901317 treatment (right upper panel). Bar, 25 μm. D. NCoA6^L2m/L2m mice have lower serum and hepatic triglycerides after T0901317 treatment. WT and NCoA6^L2m/L2m mice (five each) were treated with vehicle (V) or T0901317 (T). Triglycerides were measured by the enzymatic method. Data are presented as means ± SD. *, P < 0.05; ***, P < 0.001.

FIG. 6.

Reduction in LXR target gene expression and fecal bile acid excretion in NCoA6^L2m/L2m mice fed a high-cholesterol diet. (A to C) Relative expression levels of Cyp7A1, LXRα, and LXRβ mRNAs in the liver. RNA was prepared from the livers of WT and NCoA6^L2m/L2m mice (five in each group) fed either a normal diet (ND) or a high-cholesterol diet (HCD). Relative Cyp7A1, LXRα, and LXRβ mRNA levels were measured by real-time RT-PCR and normalized to the 18S RNA. Data are presented as means ± standard deviations (SD). ***, P < 0.001. (D) NCoA6^L2m/L2m mice have reduced fecal bile acid excretion. Feces were collected from WT and NCoA6^L2m/L2m mice (five in each group) fed ND or HCD. Total fecal bile acid was extracted, measured, and normalized to body weight. Data are presented as means ± SD. *, P < 0.05; **, P < 0.01. (E to G) Relative expression levels of the ABCA1, ABCG5, and ABCG8 mRNAs in the intestine. RNA was prepared from the small intestine of WT and NCoA6^L2m/L2m mice (five in each group) fed ND or HCD. Relative ABCA1, ABCG5, and ABCG8 mRNA levels were measured by real-time RT-PCR and normalized to the 18S RNA. Data are presented as means ± SD. *, P < 0.05; **, P < 0.01.

FIG. 7.

NCoA6^L2m/L2m mice fed a high-cholesterol diet exhibit impaired cholesterol homeostasis. (A) Photograph of livers of WT and NCoA6^L2m/L2m mice fed either a normal diet (ND) or a high-cholesterol diet (HCD). The livers of NCoA6^L2m/L2m mice fed HCD were bigger and yellower than other livers. (B) H&E-stained liver sections prepared from WT and NCoA6^L2m/L2m mice following the paraffin-embedding procedure. Many more unstained intracellular vacuoles (lipid droplets, indicated by green asterisks) were visible in NCoA6^L2m/L2m liver sections compared with other liver sections (right lower panel). CV, central vein. Bar, 25 μm. (C) Oil Red O-stained liver sections prepared from ND- or HCD-fed WT and NCoA6^L2m/L2m mice by using a cryostat. The Oil Red O-stained lipid droplets are larger in livers of NCoA6^L2m/L2m mice fed HCD (right lower panel). Bar, 50 μm. (D and E) Levels of hepatic (D) and serum (E) cholesterol in WT and NCoA6^L2m/L2m mice fed ND or HCD. Data are presented as means ± standard deviations (n = 5). *, P < 0.05; **, P < 0.01.