Regulation of glycogen synthase kinase 3beta functions by modification of the small ubiquitin-like modifier

Abstract

Modification of the Small Ubiquitin-like Modifier (SUMO) (SUMOylation) appears to regulate diverse cellular processes, including nuclear transport, signal transduction, apoptosis, autophagy, cell cycle control, ubiquitin-dependent degradation and gene transcription. Glycogen synthase kinase 3beta (GSK 3beta) is a serine/threonine kinase that is thought to contribute to a variety of biological events, including embryonic development, metabolism, tumorigenesis, and cell death. GSK 3beta is a constitutively active kinase that regulates many intracellular signaling pathways by phosphorylating substrates such as beta-catenin. We noticed that the putative SUMOylation sites are localized on K(292 )residueof (291)FKFPQ(295) in GSK 3beta based on analysis of the SUMOylation consensus sequence. In this report, we showed that the SUMOylation of GSK 3beta occurs on its K(292) residue, and this modification promotes its nuclear localization in COS-1. Additionally, our data showed that the GSK 3beta SUMO mutant (K292R) decreased its kinase activity and protein stability, affecting cell death. Therefore, our observations at first time suggested that SUMOylation on the K(292) residue of GSK 3beta might be a GSK 3beta regulation mechanism for its kinase activation, subcellular localization, protein stability, and cell apoptosis.

Keywords: GSK 3β; SUMOylation; cell apoptosis; kinase activation; protein stability; subcellular localization.

Fig. (1). GSK 3β functional domain and SUMOylation

The Glycogen synthase kinase 3β (GSK 3β) functional domains (its protein kinase and FRAT/Axin binding domain) and the putative SU-MOylation site (K²⁹² in ²⁹¹FKFPQ²⁹⁵) is indicated (A). GSK 3β SUMO mutant (K292R) was constructed by site directed mutagenesis. GSK 3β SUMO mutant (K292R) was inserted into GST fusion (for bacteria) and Ha fusion (for cell line) expression vectors. (B) GSK 3β wild type (wt) protein that was purified from E. coli was incubated with a SUMOylation assay kit (See Material and method). For the negative control, the same assay conditions were used without ATP (right lane). A western blot of the same sample was performed with GSK 3β monoclonal antibody to monitor the protein amount in the experiment (at bottom). SUMOylated GSK 3β, as several high molecular weight protein bands, was indicated. (C) A western bolt of the immunopurified GSK 3β from COS-1 was performed using the SUMO-1 specific antibody. SUMOylation of GSK 3β was detected as high molecular weight protein bands, as indicated (left lane). For the negative control, an unrelated mouse antibody was used (right lane). To monitor the total protein amount to be used in the cell lysates, the western blot was per-formed with actin monoclonal antibody (bottom). (D) Confocal microscopic analysis of endogenous GSK 3β wt (green color) and SUMO-1 (red color). GSK 3β was detected in both the cytoplasm and nuclear region. The SUMO-1 modification proteins were mainly detected in the nuclear region (yellow color). All the figures in this article represent results from three experiments repeated independently.

Fig. (2). SUMOylation Site in GSK 3β

(A) The purified GST-GSK 3β wt or GST-GSK 3β SUMO mutant (K292R) fusion protein was used as the substrate protein in the SUMOy-lation assay as described in the Materials and Methods section. The SUMOylation of GSK 3β wt was detected as a high molecular weight protein band (left lane), whereas its SUMO mutant was totally inhibited, as shown (right lane). (B) Ha –GSK 3β wt or Ha –GSK 3β SUMO mutant was transfected to COS-1 cells and immunoprecipitated with Ha mouse monoclonal antibody. The immunoprecipitants were sub-jected to the western bolt with SUMO-1, as described in the Materials and Methods section. The SUMOylation of GSK 3β wt was indicated as several high molecular weight protein bands (left lane), whereas its SUMO mutant was totally inhibited (right lane). To monitor the GSK 3β protein expression, the immunoprecipitants were subjected to the western bolt with GSK 3β polyclonal antibody (bottom).

Fig. (3). Confocal microscopic analysis of GSK 3β wt or its SUMO mutant

Confocal microscopic analysis of transfected Ha –GSK 3β wt (A), Ha –GSK 3β SUMO mutant (K292R) (B) was performed to determine whether it merged with SUMO-1 (red color). All Ha –GSK 3β constructs were shown as green color. The transfected Ha -GSK 3β wt (de-tected in both the cytoplasm and the nucleus) merged (yellow) with SUMO-1 in the nucleus (A). The transfected Ha –GSK 3β SUMO mutant was detected in the cytoplasm, but not in the nucleus (B). The SUMO-1 modification proteins were mainly detected in the nuclear region (B middle lane). GSK 3β SUMO mutant in which the SUMOylation site was eliminated was not merged with SUMO-1 in the nucleus (B right lane).

Fig. (4). Kinase activity of GSK 3β wt or its SUMO mutant

The immunopurified Ha -GSK 3β wt or its SUMO mutant (K292R) protein with Ha Ab from COS-1 was immunoblotted with GSK 3β (A) or Anti- GSK 3β Tyr 216 phospho Ab polyclonal antibody (B). The relative optical density (OD), as determined by image analysis with the Fuji Image Quant software, is indicated below. The GSK 3β kinase activity was measured using human Tau protein as a substrate (C). S422 residue phosphorylation of human Tau protein was detected with its specific antibody. The relative GSK 3β activity by image analysis with the Fuji Image Quant software is indicated below. Results shown are one of five repeated experiments

Fig. (5). Protein stability of GSK 3β wt and its SUMO mutant

Ha –GSK 3β or GSK 3β SUMO mutant (K292R) was transfected into COS-1 cells and the cells treated with cyclohexamide. The GSK 3β proteins were chased for the indicated time periods. Ha–GSK 3β proteins were immunoprecipitated with a polyclonal anti- Ha antibody and subjected to SDS-PAGE followed by western blotting with a monoclonal GSK 3β antibody (A). To monitor the protein amount, an equal amount of cell lysate was subjected to western blotting with an actin antibody. Results shown are one of five repeated experiments. Quantifi-cation of the pulse-chase experiment is shown in (B) by image analysis with the Fuji Image Quant software.