Identification and Biochemical Characterization of PGC-1beta-Interacting Proteins

Abstract

Nuclear receptors (NRs) are ligand-activated transcription factors. They regulate key biological processes including homeostasis of lipophilic molecules, organogenesis, development, and reproduction. The ability of NRs to regulate gene transcription is dependent upon their ability to directly binding to specific enhancer elements within the promoters of their target genes. While NRs directly bind to DNA, they lack the capacity to modify chromatin, unwind DNA, and recruit transcriptional machinery. All of these aforementioned processes represent key steps in the regulation of gene transcription. Therefore, the full functional activity of NRs depends upon their ability to interact with protein cofactors, termed coactivator and corepressor proteins. Metabolic syndrome is an emerging global epidemic characterized by a complex disorder that affects both glucose and lipid metabolism. Peroxisome proliferator-activated receptor-gamma (PPARg) co-activator-1alpha (PGC-1a) is a coactivator protein that is mainly expressed in tissues that require a high oxidative capacity such as fat, muscle, brain, and liver. PGC-1a plays an important role in regulating key metabolic processes involved in energy homeostasis including gluconeogenesis, mitochondrial biogenesis, adaptive thermogenesis, and b-oxidation of fatty acids. PGC-1a accomplishes this feature through its ability to interact with selected NRs and certain other transcription factor types. PGC-1b is the closest homolog of PGC-1a and also shares some similarities with PGC-1a in amino acid sequence, expression pattern, interacting protein partners, target genes, and biochemical function. However, these two coactivator proteins also display distinct proerties. For example, PGC-1b mediates hepatic lipogenic program but not gluconeogenesis, whereas PGC-1a plays a key regulatory role in hepatic gluconeogenesis but has no effects upon lipogenesis. We therefore hypothesized that the key biological and functional differences between these two coactivator proteins is due to their distinct protein interaction profile. A comparison of the primary amino acid sequence of PGC-1b and PGC-1a revealed that both proteins contain three NR-interaction motifs (-LXXLL-) with the first two being highly conserved in their N-termini, respectively. However, PGC-1b contains an additional and unique stretch of amino acid sequence that separates the second and third NR-interaction motif that is absent from PGC-1a. We therefore fused the cDNA encoding this N-terminal region of PGC-1b to the GAL4-DNA binding domain (GAL4-NBT3) and used this protein as `bait' to screen cDNA expression libraries from liver, brain, and embryo using the yeast two hybrid system to identify novel PGC-1b-interacting proteins. Three potential PGC-1b-interacting proteins were identified. The first is a NR protein called COUP-TF1. The second PGC-1b-interacting protein we identified is a protein phosphatase regulatory protein termed PP4R1. The third protein we identified as a PGC-1b-interacting protein is a member of the SWI/SNF family of proteins called BAF60a. We further confirmed the interaction of PGC-1b/PP4R1 and PGC-1b/BAF60a in vitro using the GST pull-down assay. In vivo analysis using immunohistochemical methods further confirms the ability of PP4R1 and BAF60a to co-localize with PGC-1b. The identification of PP4R1 and BAF60a as PGC-1b-interacting proteins will contribute to an increased understanding of the repertoire of PGC-1b-interacting proteins. This research will also likely contribute to understanding the potential role of these proteins in regulating PGC-1b-mediated coactivation of transcription.