The homeobox protein Prox1 is a negative modulator of ERR{alpha}/PGC-1{alpha} bioenergetic functions

Abstract

Estrogen-related receptor alpha (ERRalpha) and proliferator-activated receptor gamma coactivator-1alpha (PGC-1alpha) play central roles in the transcriptional control of energy homeostasis, but little is known about factors regulating their activity. Here we identified the homeobox protein prospero-related homeobox 1 (Prox1) as one such factor. Prox1 interacts with ERRalpha and PGC-1alpha, occupies promoters of metabolic genes on a genome-wide scale, and inhibits the activity of the ERRalpha/PGC-1alpha complex. DNA motif analysis suggests that Prox1 interacts with the genome through tethering to ERRalpha and other factors. Importantly, ablation of Prox1 and ERRalpha have opposite effects on the respiratory capacity of liver cells, revealing an unexpected role for Prox1 in the control of energy homeostasis.

Figure 1.

Prox1 interacts with and influences the transcriptional activity of ERRα and PGC-1α. (A) Prox1, ERRα, and PGC-1α interact in vivo. Lysates from mouse liver were subjected to immunoprecipitation and immunoblot analyses with the indicated antibodies. (B) Direct interactions between Prox1, PGC-1α, and ERRα. In vitro translated ERRα and PGC-1α were subjected to pull-down analysis with GST-Prox1 fragments. (N) N terminus; (M) middle; (C) C terminus. (C) Prox1 interacts with the DNA-binding domain (DBD) of ERRα. In vitro translated Prox1 and a LxxLL 1/2 mutant were subjected to pull-down analysis with GST-ERRα fragments. (FL) Full-length; (N) N terminus; (LBD) ligand-binding domain. (D) Prox1 interacts with a new functional domain of PGC-1α. In vitro translated Prox1 with GST-PGC-1α fragments. (E) Schematic representation of a potential trimeric interaction between Prox1, ERRα, and PGC-1α. (AD) Activation domain; (RD) repression domain; (RRM) RNA recognition motif; (AF-2) activation function 2; (white bars) LxxLL motifs. (F) Re-ChIP experiments performed in the mouse liver on the Pdk4 promoter using either anti-ERRα or anti-Prox1 antibodies in a serial manner. (G) Effects of wild-type and mutant Prox1 proteins on the transcriptional activity of ERRα and PGC-1α. The Pdk4 promoter–luciferase reporter gene was cotransfected in HepG2 cells with empty vector (−), ERRα, PGC-1α, or a combination of both expression vectors in the presence or absence of wild-type or mutant Prox1. (H) Same assay as in G using the Cs promoter as the reporter gene.

Figure 2.

Genome-wide promoter occupancy of ERRα and Prox1 in mouse liver. (A) Venn diagrams illustrating the overlap in ERRα (red) and Prox1 (green) direct target genes from ChIP-on-chip analyses in the mouse liver. (B) Standard ChIP validation of a subset of ERRα-enriched (red) and Prox1-enriched (green) segments. Occupancy of PGC-1α on these selected DNA segments bound by ERRα, Prox1, or both factors as assayed by standard ChIP is also shown. (C) Representative binding profiles of ERRα (red line) and Prox1 (green line) on specific or common target extended promoters containing either distinct or overlapping binding sites.

Figure 3.

The ERRα bioenergetic regulon. (A) Enrichment of canonical metabolic pathways in the ChIP-on-chip target genes determined to be common (yellow) or specific to either ERRα (red) or Prox1 (green). (NS) Not significant; (*) P < 0.05. (B) ChIP-on-chip direct target genes specific to ERRα (red) or Prox1 (green) or shared by both factors (yellow) involved in metabolic pathways are shown. All genes involved in glycolysis, pyruvate metabolism, and the TCA cycle are targets of ERRα, a cluster of genes defining the ERRα bioenergetic regulon. Genes labeled in black were not identified as being enriched by either ERRα or Prox1.

Figure 4.

Divergent regulation of mitochondrial functions by ERRα and Prox1. (A) Western blot analysis on lysates prepared from the HepG2 knockdown samples is shown with the respective antibodies as indicated. Detection of RPLP was used a control. (B,C) Cellular oxygen consumption (B) and extracellular acidification rates (C) were measured in intact HepG2 cells treated with either control siRNA or an siRNA against ERRα or Prox1. Rates determined following sequential addition of oligomycin, FCCP, and rotenone were taken from an average of two measurements and are expressed as a percentage of the baseline rates. (*) P < 0.05.