An LXR-NCOA5 gene regulatory complex directs inflammatory crosstalk-dependent repression of macrophage cholesterol efflux

Abstract

LXR-cofactor complexes activate the gene expression program responsible for cholesterol efflux in macrophages. Inflammation antagonizes this program, resulting in foam cell formation and atherosclerosis; however, the molecular mechanisms underlying this antagonism remain to be fully elucidated. We use promoter enrichment-quantitative mass spectrometry (PE-QMS) to characterize the composition of gene regulatory complexes assembled at the promoter of the lipid transporter Abca1 following downregulation of its expression. We identify a subset of proteins that show LXR ligand- and binding-dependent association with the Abca1 promoter and demonstrate they differentially control Abca1 expression. We determine that NCOA5 is linked to inflammatory Toll-like receptor (TLR) signaling and establish that NCOA5 functions as an LXR corepressor to attenuate Abca1 expression. Importantly, TLR3-LXR signal crosstalk promotes recruitment of NCOA5 to the Abca1 promoter together with loss of RNA polymerase II and reduced cholesterol efflux. Together, these data significantly expand our knowledge of regulatory inputs impinging on the Abca1 promoter and indicate a central role for NCOA5 in mediating crosstalk between pro-inflammatory and anti-inflammatory pathways that results in repression of macrophage cholesterol efflux.

Keywords: LXR; NCOA5; atherosclerosis; inflammation; quantitative mass spectrometry.

Figure 1. Promoter enrichment-quantitative mass spectrometry (PE-QMS) identifies transcriptional regulators of Abca1 expression

1. RT–qPCR of Abca1 transcripts following 1 μM T0901317 stimulation of LXR^+/+ and LXR^−/− primary BMMs. Fold changes are shown relative to LXR^+/+ 0 h. Error bars represent ± SEM for n = 4–8 (**P = 0.001 at 8 h and **P = 0.00004 at 16 h versus LXR^−/−; **P = 0.0003 at 0 h versus LXR^+/+); bottom, proposed model explaining the ligand-dependent induction of Abca1. Following ligand stimulation, LXR–corepressor complexes are exchanged for LXR–coactivator complexes, which promote gene expression. The LXR–cofactor interactions responsible for the attenuation of expression beginning after 16 h remain unknown.

2. Overview of the PE-QMS experimental strategy to identify LXR ligand-stimulated and LXR binding-dependent interactions with the Abca1 promoter. Nuclear extracts were prepared from RAW 264.7 macrophages treated with control or 1 μM T0901317 (depicted as a black circle) for 18 h. The Abca1 promoter region was used to isolate gene regulatory complexes under the indicated conditions. Immobilized templates were washed, and bound proteins eluted and digested. Peptides were purified, then identified, and quantified by mass spectrometry.

3. Heat map of the PE-QMS experiment. Ligand-dependent subsets were quantified using label-free and isotope labeling approaches and are indicated on the left. Colors represent Log2 relative abundance ratios (T0901317/vehicle) of identified proteins. LXR binding-dependent subsets were quantified using an isotope labeling approach and are indicated on the right. Changes in binding > 1.5-fold were employed as cutoffs for depicted subsets. ND = not determined. Full dataset is available as Supplementary Table S1.

Figure 2. Identification of NCOA5 as an LXR repressor

1. Protein–protein interaction network showing connectivity between the putative ligand-stimulated and LXR binding-dependent transcriptional regulators of Abca1 (shown in orange). Edges represent protein–protein interactions. NCOA5 node and edge are outlined red to indicate that this physical association is found in mouse interaction networks.

2. Luciferase reporter assays from RAW 264.7 macrophages expressing reporter alone, or together with full-length SND1, SART1, or HMBOX1. LXR ligand stimulations with 1 μM T0901317 were performed for 18 h. A diagram of the reporter construct is shown above. Fold changes are shown relative to vehicle-stimulated reporter alone (first lane). Error bars represent ± SEM for n = 9 (**P = 0.001, *P = 0.011 for SND1, *P = 0.046 for SART1).

3. Luciferase reporter assays from RAW 264.7 macrophages expressing reporter alone, or together with full-length NCOA5 or an NCOA5 mutant lacking the NH₂-terminus. LXR ligand stimulations with 1 μM T0901317 were performed for 18 h. Diagrams of the reporter constructs are shown above. Fold changes are shown relative to vehicle-stimulated reporter alone (first lane). Error bars represent ± SEM for n = 6–12 (**P < 0.001, *P = 0.035).

4. RT–qPCR of Abca1 transcripts following infection of primary BMMs with Ncoa5 or control retrovirus. LXR ligand stimulations were performed with 1 μM T0901317 for 18 h. Fold changes are shown relative to vehicle-stimulated control. Error bars represent ± SEM for n = 4 (*P = 0.029).

Figure 3. NCOA5 functions as an LXR corepressor

1. In vitro pulldown assays using recombinant GST-LXRα to isolate the indicated in vitro translated NCOA5 constructs. Assays were performed for 2 h in the presence or absence of 2 μM T0901317. Samples were immunoblotted as indicated. Note the requirement of the NCOA5 NH₂-terminus for interacting with LXRα.

2. Immunoprecipitation of Protein C-tagged LXRα from stable RAW 264.7 macrophage nuclear extracts stimulated with 1 μM T0901317 or vehicle control for 18 h, followed by immunoblotting for endogenous NCOA5 or over-expressed LXRα. Protein immunoblots of nuclear extracts are shown below.

3. NCOA5 ChIP time course from primary BMMs stimulated with 1 μM T0901317 or vehicle control. qPCR was performed for the Abca1 proximal LXRE. Note the ligand-stimulated association of NCOA5. Error bars represent ± SEM for n = 4–9 (*P = 0.02).

4. NCOA5 ChIP assays from LXR^−/− BMMs stimulated with 1 μM T0901317 or vehicle control for 18 h. qPCR was performed for the Abca1 proximal LXRE. Error bars represent ± SEM for n = 5–6 (**P = 0.0003 for vehicle and **P = 0.002 for T0901317).

Figure 4. Attenuation of Abca1 transcription by NCOA5

1. RT–qPCR of Abca1 expression following 1 μM T0901317 treatment in primary BMMs infected with a non-silencing or Ncoa5-specific shRNA. Fold changes are shown relative to shControl 0 h. Note the elevated Abca1 expression at 24–30 h following loss of NCOA5. Error bars represent ± SEM for n = 4–6 (**P = 0.0007 at 24 h and **P = 0.004 at 30 h, *P = 0.02).

2. Quantification of ABCA1 immunoblots from primary BMMs infected as in (A). Representative immunoblot is shown in Supplementary Fig S7E. LXR ligand stimulation with 1 μM T0901317 or vehicle control was performed for 8 h or 30 h. Error bars represent ± SEM for n = 2 (**P = 0.007).

Figure 5. NCOA5 mediates TLR3–LXR signal crosstalk by antagonizing Abca1 expression and function

1. A, B RT–qPCR of Abca1 expression from primary BMMs infected with non-silencing or Ncoa5-specific shRNAs. Ligand stimulations were performed for 4 h with vehicle control, 1 μM T0901317 alone or together with 6 μg/ml PolyIC (A) or 10 ng/ml LPS (B). Fold changes are shown relative to vehicle-stimulated shControl. Note only the loss of TLR3-mediated repression following Ncoa5 silencing. Error bars represent ± SEM for n = 4–10 (**P = 0.0003 for A, **P = 0.0004 for shControl in B, **P = 0.0002 for shNcoa5 in B versus T0901317).

2. C, D RT–qPCR of Abca1 expression from primary BMMs infected with non-silencing or Ncoa5-specific shRNAs. Ligand stimulations were performed for 4 h unstimulated, or with 6 μg/ml PolyIC (C) or 10 ng/ml LPS (D). Fold changes are shown relative to unstimulated shControl. Error bars represent ± SEM for n = 4–6 (**P = 0.0006 for C, **P = 0.00000001 for shControl in D, **P = 0.000001 for shNcoa5 in D, *P = 0.013 versus T0901317).

3. E NCOA5 ChIP assays from primary BMMs stimulated for 3 h with 1 μM T0901317, 1 μM T0901317 + 6 μg/ml PolyIC, or vehicle control. qPCR was performed for the Abca1 proximal LXRE. Note the increased occupancy only in the presence of both lipid and inflammatory ligands. Error bars represent ± SEM for n = 8–12 (**P = 0.0007 versus T0901317).

4. F NCOA5 ChIP assays from primary BMMs stimulated for 3 h with 6 μg/ml PolyIC or left unstimulated. qPCR was performed for the Abca1 proximal LXRE. Error bars represent ± SEM for n = 5–6.

5. G Cholesterol efflux assays to APOA1 from primary BMMs infected with non-silencing or Ncoa5-specific shRNAs. Agonist stimulations were performed for 6 h. Error bars represent ± SEM for n = 3 (*P = 0.02 versus T0901317).

Figure 6. NCOA5 represses Abca1 expression following TLR3–LXR signal crosstalk by inhibiting RNAPII function

1. A, B Unmodified/pSer5 RNAPII (A) or RNAPII pSer2 (B) ChIP assays from primary BMMs stimulated for 4 h with 1 μM T0901317, 1 μM T0901317 + 6 μg/ml PolyIC, or vehicle control. qPCR was performed for the Abca1 TSS. Error bars represent ± SEM for n = 6–9 (**P = 0.0001 in A, **P = 0.002 in B versus T0901317).

2. C, D RNAPII ChIP assays as in (A, B) but performed from LXR^−/− BMMs. Error bars represent ± SEM for n = 5–9 (**P = 0.000000001, *P = 0.02 versus T0901317).

3. E RT–qPCR of Abca1 expression from LXR^+/+ versus LXR^−/− BMMs. Ligand stimulations were performed for 4 h with vehicle control, 1 μM T0901317, or 1 μM T0901317 together with 6 μg/ml PolyIC. Fold changes are shown relative to vehicle-stimulated LXR^+/+. Error bars represent ± SEM for n = 4 (**P = 0.004, *P = 0.012 versus T0901317).

4. F RNAPII pSer2 ChIP assays from primary BMMs infected with non-silencing or Ncoa5-specific shRNAs. Ligand stimulations were performed for 4 h with vehicle control, 1 μM T0901317, or 1 μM T0901317 + 6 μg/ml PolyIC. qPCR was performed for the Abca1 TSS. Note RNAPII pSer2 returns to baseline occupancy following TLR3 stimulation in shControl but not shNcoa5 BMMs (**P = 0.004 versus baseline). Error bars represent ± SEM for n = 3 from 11 mice.

Figure 7. Model for the NCOA5-mediated repression of Abca1 expression

1. In response to sterol ligand treatment, LXR–RXR heterodimers initially induce the transcription of Abca1 through recruitment and activation of RNAPII (left). Following prolonged sterol ligand treatment, the NCOA5 repressor directly interacts with LXR at the Abca1 promoter. This repressor complex inhibits the recruitment and function of RNAPII resulting in attenuated expression of Abca1 (right). We hypothesize that prolonged sterol ligand treatment induces a regulatory modification on NCOA5 (shown in green), such as phosphorylation, which facilitates its recruitment to LXR.

2. Crosstalk between pro-inflammatory TLR3 and anti-inflammatory LXR pathways promotes the association of NCOA5 with LXR, resulting in the inhibition of RNAPII recruitment and function, and repression of Abca1 gene expression. The regulatory modification of NCOA5 following activation of these two pathways may be similar or distinct to that in (A). We also cannot discount the contribution of additional constituents of this transcriptional complex to the recruitment of NCOA5.