SIRT1 mediates FOXA2 breakdown by deacetylation in a nutrient-dependent manner

Abstract

The Forkhead transcription factor FOXA2 plays a fundamental role in controlling metabolic homeostasis in the liver during fasting. The precise molecular regulation of FOXA2 in response to nutrients is not fully understood. Here, we studied whether FOXA2 could be controlled at a post-translational level by acetylation. By means of LC-MS/MS analyses, we identified five acetylated residues in FOXA2. Sirtuin family member SIRT1 was found to interact with and deacetylate FOXA2, the latter process being dependent on the NAD+-binding catalytic site of SIRT1. Deacetylation by SIRT1 reduced protein stability of FOXA2 by targeting it towards proteasomal degradation, and inhibited transcription from the FOXA2-driven G6pase and CPT1a promoters. While mutation of the five identified acetylated residues weakly affected protein acetylation and stability, mutation of at least seven additional lysine residues was required to abolish acetylation and reduce protein levels of FOXA2. The importance of acetylation of FOXA2 became apparent upon changes in nutrient levels. The interaction of FOXA2 and SIRT1 was strongly reduced upon nutrient withdrawal in cell culture, while enhanced Foxa2 acetylation levels were observed in murine liver in vivo after starvation for 36 hours. Collectively, this study demonstrates that SIRT1 controls the acetylation level of FOXA2 in a nutrient-dependent manner and in times of nutrient shortage the interaction between SIRT1 and FOXA2 is reduced. As a result, FOXA2 is protected from degradation by enhanced acetylation, hence enabling the FOXA2 transcriptional program to be executed to maintain metabolic homeostasis.

Figure 1. FOXA2 is an acetylated protein.

(A) LC-MS/MS spectrum and the deduced peptide sequence of FOXA2 acetylation at lysine K275 (in red). (B) Peptides containing five in vivo FOXA2 acetylation sites which were identified by LC-MS/MS analyses. FLAG-FOXA2 was expressed in human kidney HEK293T cells, purified and digested with proteases. Peptides were separated and enriched by C18 analytical columns and subsequently subjected to LC-MS/MS. Acetylated residues in red and underlined. (C) Schematic representation of the FOXA2 protein structure. Markers illustrate the localization of the five acetylation sites and colored boxes represent functional domains. TAD, Trans Activation Domain. NLS, Nuclear Localization Sequence.

Figure 2. SIRT1 interacts with and deacetylates FOXA2.

(A) Increase in FOXA2 acetylation by NAM. FLAG-FOXA2-transfected HEK293T cells were treated with class I/II HDAC inhibitor TSA (5 µM), class III HDAC inhibitor NAM (20 mM) or both for 16 hours. Acetylation level of immunoprecipitated FLAG-FOXA2 was assessed on Western Blot probed with an anti-acetylated-lysine specific antibody. (B) SIRT1 co-transfection affects FOXA2 acetylation. HEK293T cells were transfected with FLAG-FOXA2 alone or together with MYC-SIRT1 in the presence of 20 mM NAM. Acetylation was assessed as in (A). (C) SIRT1 interacts with FOXA2. HEK293T cells were co-transfected as in (B) followed by co-immunoprecipitation with anti-FLAG M2 beads. Interaction of FLAG-FOXA2 and MYC-SIRT1 was assessed by Western Blot analysis with indicated antibodies. (D) Endogenous FOXA2 and SIRT1 interact in the nucleus. HepG2 cells were subjected to immunofluorescence-based proximity ligation assay . Nuclei were counterstained with DAPI.

Figure 3. Nutrient withdrawal reduces SIRT1 and FOXA2 interaction.

(A) Nutrient withdrawal reduced SIRT1-FOXA2 interaction. HEK293T cells were co-transfected with indicated constructs and starved from glucose and serum for 16 hours. FOXA2 and SIRT1 were co-immunoprecipitated with anti-FLAG-M2 beads and protein levels were assessed by Western Blot analysis probed with the indicated antibodies. Relative interaction is defined as the fraction of FOXA2 that binds to SIRT1. (B) Reduced endogenous interaction of FOXA2 and SIRT1 in the nucleus upon nutrient withdrawal as in A for 2 or 4 hours. Interaction between endogenously expressed SIRT1 and FOXA2 was assessed in HepG2 cells by PLA. Nuclei were counterstained with DAPI.

Figure 4. SIRT1 inhibits FOXA2-driven transcription.

(A) and (B) SIRT1 co-transfection affects the FOXA2 transcriptional program. HEK293T cells were co-transfected with the indicated constructs and TK Renilla in combination with (A) G6pase and (B) CPT1a reporter constructs. After 48 hours, a luciferase-driven transcriptional assay was performed to assess FOXA2-mediated transcriptional activity on the promoter constructs. (C) SIRT1 knock-down increased FOXA2-driven transcription on the G6pase gene promoter. HEK293T cells were transfected with either control, or SIRT1 siRNA and subsequently co-transfected as in (A). Transcriptional activity was assessed via luciferase assays. (D) SIRT1 affects FOXA2-driven transcription by reducing FOXA2 protein levels. HEK293T cells were transfected as in (A) with FLAG-FOXA2 alone or in combination with increasing amounts of MYC-SIRT1. Empty pcDNA3 vector was co-transfected to balance between samples for the amount of DNA transfected. Transcriptional analysis was determined by luciferase assays. Values shown represent the mean of three independent experiments +SEM. Statistical analysis was performed by two-tailed Student's t-Test. **, P<0.01. Protein levels were assessed by Western Blotting using indicated antibodies. Actin was used as loading control.

Figure 5. SIRT1-mediated deacetylation targets FOXA2 towards proteasomal degradation.

(A) SIRT1 affects FOXA2 protein stability. HEK293T cells were transfected with FLAG-FOXA2 in the presence of 20 mM NAM or vehicle for 16 hours prior to cell lysis. Western Blots were probed for FLAG, and actin as a loading control. (B) Wild type SIRT1 and mutant SIRT1 H363Y similarly interact with FOXA2. HEK293T cells were transfected with FLAG-FOXA2 alone or in combination with either wild-type (wt) or the catalytically impaired MYC-SIRT1-H363Y mutant. FOXA2 and SIRT1 were co-immunoprecipitated with anti-FLAG-M2 beads and protein levels were assessed by Western Blot analysis probed with the indicated antibodies. (C) SIRT1 catalytic activity regulates FOXA2 stability. HEK293T cells were transfected and subjected to co-immunoprecipitation as in (B). Protein expression and FOXA2 acetylation level were assessed by Western Blot using the indicated antibodies. (D) Acetylation-impaired FLAG-12K-R-FOXA2 mutant showed reduced acetylation and protein stability. HEK293T cells were transfected with either FLAG-FOXA2 or FLAG-12K-R-FOXA2 and cultured as in (A). After FLAG-immunoprecipitation, protein levels were assessed by Western Blot probed with the indicated antibodies. (E) FOXA2 is subjected to proteasomal degradation in a time-dependent manner. HEK293T cells were transfected with FLAG-FOXA2 and supplemented with proteasome inhibitor MG132 (20 µM) or vehicle-containing medium for 3 or 6 hours. FOXA2 protein levels were assessed by Western Blot, and actin as a loading control. (F) SIRT1 mediates endogenous FOXA2 breakdown. HepG2 cells were transfected with MYC-SIRT1 and supplemented with 20 µM MG132- or vehicle-containing medium for 3 hours. (G) NAM preserves stability of FOXA2 upon abrogation of protein translation. HEK293T cells were transfected with FLAG-FOXA2 and supplemented 20 mM NAM or vehicle for 16 hours, followed by addition of cycloheximide (5 µg/ml) or vehicle-containing medium for 4 or 8 hours. (E–G) show a representative experiment from three independent experiments, and quantification of FOXA2 protein levels was performed by normalizing to actin protein levels. (H) 12K-R FOXA2 mutant has a higher level of poly-ubiquitylation. HEK293T cells were transfected with FLAG-FOXA2 or FLAG-12K-R-FOXA2. After culture of 20 µM MG132 for 3 hours, FOXA2 was immunoprecipitated with FLAG-M2 beads and immunoprecipitated proteins were assessed by Western Blot probed with the indicated antibodies. (I) Nutrient-dependent regulation of Foxa2 acetylation level in murine livers. Murine Foxa2 was immunoprecipitated from whole liver protein lysates from mice fed ad libitum (AL) or starved (ST) to 25% of ad libitum levels. Western Blots were probed with the indicated antibodies. The asterisk (*) marks a co-eluting acetylated protein of unknown identity.

Figure 6. Model for nutrient-dependent regulation of FOXA2 stability via SIRT1-mediated deacetylation.

Graphical representation of the interaction and deacetylation of FOXA2 by SIRT1 depending on available nutrient levels. In times of nutrient excess (left panel) SIRT1 interacts with FOXA2 in the nucleus, resulting in FOXA2 deacetylation and subsequent proteasomal degradation. When nutrients are limited (right panel), FOXA2 is required to exert its transcriptional program in the nucleus. The interaction of FOXA2 with SIRT1 is diminished, resulting in enhanced acetylation and stability of FOXA2.