GRK5 is required for adipocyte differentiation through ERK activation

Abstract

Previous studies have identified G protein-coupled receptor (GPCR) kinase 5 (GRK5) as a genetic factor contributing to obesity pathogenesis, but the underlying mechanism remains unclear. We demonstrate here that Grk5 mRNA is more abundant in stromal vascular fractions of mouse white adipose tissue, the fraction that contains adipose progenitor cells, or committed preadipocytes, than in adipocyte fractions. Thus, we generated a GRK5 knockout (KO) 3T3-L1 preadipocyte to further investigate the mechanistic role of GRK5 in regulating adipocyte differentiation. During adipogenic stimulation, GRK5 KO preadipocytes had decreased lipid accumulation and delayed mature adipocyte development compared to wildtype cells coupled with suppressed adipogenic and lipogenic gene expression. Beside GPCR signaling, RNA sequencing and pathway analysis identified insulin-like growth factor 1 (IGF-1) signaling to be one of the top 5 significantly dysregulated pathways in GRK5 KO cells. GRK5 KO cells also displayed decreased insulin-stimulated ERK phosphorylation, a downstream target of insulin-stimulated IGF-1 receptor activation, suggesting that GRK5 acts through this critical pathway to impact 3T3-L1 adipocyte differentiation. To find a more translational approach, we identified a new small molecule GRK5 inhibitor that was able to reduce 3T3-L1 adipogenesis. These data suggest that GRK5 is required for adipocyte differentiation through IGF-1 receptor/ERK activation and may be a promising translational target for obesity.

Fig. 1. The relationship between Grk5 expression and adiposity.

A Eight-week-old male C57BL/6J mice were fed chow or a high fat diet (Envigo #TD 88137, 42% from fat, 0.2% total cholesterol) for 16 weeks and their body weight was measured (n = 5/diet group). B Mice were then fasted for 24 h and their epididymal (Epi) visceral (Vis) white adipose tissue (WAT) and brown adipose tissue (BAT) RNA was extracted and reverse-transcribed into cDNA for real-time PCR quantification of Grk5 normalized to 18 s (endogenous control). C Six-week-old male C57BL/6J mice were fed chow or a high fat diet (Research Diets Inc #D12492, 60% from fat) for 12 weeks and their body composition such as fat mass was quantified by EcoMRI (n = 10/diet group). D Adipocyte fraction and stromal vascular (SV) cell fraction were isolated from the Epi Vis WAT of overnight fasted mice. Both fractions’ RNA was extracted and reverse-transcribed into cDNA for real-time PCR quantification of Grk5 normalized to 18 s (endogenous control). All results are mean ± SEM, presented as the fold change compared to chow-fed mouse group and analyzed using a two-tailed Student’s unpaired t test (A–C), or the fold change compared to chow SV fractions and analyzed using a one-way ANOVA with Sidak multiple comparisons (D).

Fig. 2. The effect of GRK5 deficiency on adipocyte differentiation.

A Cellular proteins of undifferentiated wildtype (WT) control and GRK5 knockout (KO) 3T3-L1 preadipocytes (n = 4/genotype) were harvested and subjected to Western blot using anti-GRK5, anti-GRK2, and anti-α-tubulin antibodies. B After 2 days of growth, proliferation was assessed in undifferentiated WT and GRK5 KO 3T3-L1 preadipocytes (n = 6/genotype). The percentage of EdU-positive cells (pink) was calculated by merging EdU (red) and Hoechst 3342 (blue) staining. C WT and GRK5 KO 3T3-L1 cells (n = 3/genotype) were proliferated for 2 days (Day 0) and then differentiated into adipocytes for 9 days. Day 0, 3, 6, and 9 cells were lipid-extracted to measure triacylglycerol (TAG) mass by a colorimetric assay. Daily Cytation images at 10x magnification were taken during 9 days of adipocyte differentiation. All results are mean ± SEM and analyzed using a two-tailed Student’s unpaired t test (B) and a two-way ANOVA with Sidak multiple comparisons (C).

Fig. 3. The effect of GRK5 deficiency on adipogenic and lipogenic gene and protein expression.

Cellular RNA was extracted from wildtype (WT) control and GRK5 knockout (KO) 3T3-L1 cells (n = 3/genotype) and reverse-transcribed into cDNA for real-time PCR quantification of Acc1, Pparγ, Fasn, Cd36, Fabp4, Dgat2, and Lipin1 normalized to 18 s (endogenous control). Cellular proteins of undifferentiated (Day 0) and differentiated (Day 9) WT control and GRK5 KO 3T3-L1 cells (n = 2/genotype) were harvested and subjected to Western blot using anti-GRK5, anti-PPARγ, anti-CD36, and anti-α-tubulin antibodies. All results are mean ± SEM and presented as the fold change compared to WT at Day 0 and analyzed using a two-way ANOVA with Sidak multiple comparisons.

Fig. 4. The potential underlying mechanisms of GRK5 in adipocyte differentiation.

A Pathway analysis of RNA sequencing data using Limma. B Wildtype (WT) control and GRK5 knockout (KO) 3T3-L1 preadipocytes (n = 2/genotype) were differentiated for 2 days and treated with 1 μg/ml of insulin for 5, 10, and 15 min. Then, the cellular proteins were harvested and subjected to Western blot using anti-GRK5, anti-phosphorylated (p)-ERK, anti-ERK, and anti-GAPDH antibodies.

Fig. 5. The effect of a GRK5 inhibitor on adipocyte differentiation.

A Alignment of the GRK5 (green) and GRK2 (cyan) crystal structures. There is an alpha helix (shown in red) present near the binding pocket in GRK5, not present in GRK2. B Relative docking positions of GRK5-IN-2 in GRK5 and GRK2. C The dose-response curves of GRK5-IN-2 and staurosporine were determined by a GRK5 kinase system and a luminescent ADP detection assay. D Wildtype 3T3-L1 preadipocytes were differentiated and concurrently treated without (DMSO vehicle) or with GRK5-IN-2 (n = 3/dose) for 7 days. Cells were lipid-extracted to measure triacylglycerol (TAG) mass by an enzymatic colorimetric assay. E Day 3 differentiated wildtype 3T3-L1 cell cultures were pretreated without (DMSO vehicle) or with GRK5-IN-2 (40 μM) for 30 min and then treated with 1 μg/ml of insulin plus 0.5 μCi/ml of [1,2-¹⁴C]-acetic acid for 60 and 120 min (n = 3/time point). Cells were lipid-extracted, and TAG, free cholesterol (FC), cholesteryl ester (CE), and phospholipid (PL) were separated using thin layer chromatography. [¹⁴C]-TAG, [¹⁴C]-FC, [¹⁴C]-CE, and [¹⁴C]-PL were quantified by liquid scintillation counting. All results are mean ± SEM and analyzed using a two-way ANOVA with Sidak multiple comparisons.