Paracrine role of endothelial IGF-1 receptor in depot-specific adipose tissue adaptation in male mice

Abstract

During recent decades, changes in lifestyle have led to widespread nutritional obesity and its related complications. Remodelling adipose tissue as a therapeutic goal for obesity and its complications has attracted much attention and continues to be actively explored. The endothelium lines all blood vessels and is close to all cells, including adipocytes. The endothelium has been suggested to act as a paracrine organ. We explore the role of endothelial insulin-like growth factor-1 receptor (IGF-1R), as a paracrine modulator of white adipose phenotype. We show that a reduction in endothelial IGF-1R expression in the presence of high-fat feeding in male mice leads to depot-specific beneficial white adipose tissue remodelling, increases whole-body energy expenditure and enhances insulin sensitivity via a non-cell-autonomous paracrine mechanism. We demonstrate that increased endothelial malonate may be contributory and that malonate prodrugs have potentially therapeutically relevant properties in the treatment of obesity-related metabolic disease.

Fig. 1. Reduction in murine endothelial IGF-1R expression improves whole body insulin sensitivity and energy expenditure in the setting of high fat feeding.

A Schematic representation of the generation of tamoxifen-inducible endothelial cell specific IGF-1R knockdown mice (ECIGF-1R^KD). Created in BioRender. Luk, C. (2023) https://BioRender.com/x35a412. B Schematic representation of experimental protocol. C Quantification of body weight from 2-week high fat fed (HFD) control and ECIGF-1R^KD mice. (n = 7 and 8). D Quantification of wet organ weight from 2-week HFD control and ECIGF-1R^KD mice (n = eWAT 26 and 13, Liver 17 and 16 and Heart 7 and 11). E Glucose tolerance over time for 2-week HFD control and ECIGF-1R^KD mice (n = 21 and 11). F Area under the curve (AUC) analysis for glucose tolerance for 2-week HFD control and ECIGF-1R^KD mice (n = 21 and 11). G Insulin tolerance test for 2-week HFD control and ECIGF-1R^KD mice (n = 17 and 10). p = 0.04 and 0.01 for 60 min and 120 min respectively. H The area under the curve analysis for insulin tolerance tests for 2-week HFD control and ECIGF-1R^KD mice (n = 17 and 10). p = 0.03. I Energy expenditure in 2-week HFD fed control and ECIGF-1R^KD mice over 24 h period (n = 4 and 6). The dark cycle is shown in grey. J Average energy expenditure in 2-week HFD fed control and ECIGF-1R^KD mice during specific time periods (n = 4 and 6). p = 0.04 and 0.03 for full day and dark cycle respectively. K Quantification of plasma adiponectin levels from 2-week HFD control and ECIGF-1R^KD mice (n = 10 and 11). p = 0.05. L Quantification of Brown adipose tissue adiponectin gene expression from 2-week HFD control and ECIGF-1R^KD mice (n = 7 and 8). p = 0.02. M Quantification and representative images of western blots for p473 AKT, AKT and B-actin from muscle from 2-week HFD fed control and ECIGF-1R^KD mice (n = 6 and 9). p = 0.04. Data shown as mean ± SEM, data points are individual mice. Metabolic parameters were measured by indirect calorimetry, ANOVA testing was performed using calrapp.org. p < 0.05 taken as statistically significant using student unpaired two-tailed t-test and denoted as *.

Fig. 2. Reduction in murine endothelial IGF-1R expression prevents deleterious remodelling of epididymal white adipose tissue in the setting of high fat feeding.

A Representative images of hematoxylin and eosin (H and E) stained epididymal white adipose tissue (eWAT) from 2-week high fat feeding (HFD) control and tamoxifen-inducible endothelial cell specific IGF-1R knockdown mice (ECIGF-1R^KD) mice (Scale bar = 200 µm). B Quantification of adipocyte size in eWAT from 2-week HFD control and ECIGF-1R^KD mice (n = 9 and 19). p = 0.02. C Quantification of eWAT adipocyte size distribution from 2-week HFD control and ECIGF-1R^KD mice (n = 9 and 19). p = 0.05, 0.006, 0.001, 0.01, and 0.001 for ranges ≤1000, 6001–7000,7001–8000, 8001–9000 and >9000 respectively. D Representative images of isolectin B4 (IB4, Red) and LipidTox (Green) stained eWAT from 2-week HFD control and ECIGF-1R^KD mice (Scale bar = 100 µm). E Quantification of eWAT vascularisation from 2-week HFD control and ECIGF-1R^KD mice (n = 6 and 14). p = 0.001. F Representative images of 2-week HFD control and ECIGF-1R^KD eWAT tissue explants (Scale bar = 200 µm). G Quantification of eWAT neovascularisation from 2-week HFD control and ECIGF-1R^KD mice (n = 5 and 5). p = 0.008. H Representative images of picro sirius red stained eWAT from 2-week HFD control and ECIGF-1R^KD mice (Scale bar = 200 µm). I Quantification of eWAT collagen deposition from 2-week HFD control and ECIGF-1RKD mice (n = 7 and 9). J Quantification of eWAT crown like structures from 2-week HFD control and ECIGF-1RKD mice per high powered field (HPF) (n = 5 and 6). K Quantification of eWAT upregulated gene expression from 2-week HFD control and ECIGF-1R^KD mice (Ucp1 n = 8 and 13, Vegfa 8 and 16, Cited1 6 and 6). p = 0.03, 0.02 and 0.05 for Ucp1, Vegfa and cited 1 respectively. L Quantification of basal and forskolin stimulated extracellular glycerol concentration from eWAT from 2-week HFD control and ECIGF-1R^KD mice (n = 3 and 3 mice and 5 explants per mouse). M Quantification of lipolytic capacity from eWAT from 2-week HFD control and ECIGF-1R^KD mice (n = 3 and 3 mice and 5 explants per mouse). P = 0.05. N Quantification of respiration from eWAT normalised to mitochondrial content from 2-week HFD control and ECIGF-1R^KD mice (n = 7 and 8). p = 0.05. Data shown as mean ± SEM, data points are individual mice. p < 0.05 taken as statistically significant using student unpaired two tailed t-test and denoted as * (**p ≤ 0.01). For respirometry data two-way ANOVA corrected for multiple comparisons by controlling the False Discovery Rate using two-stage step-up method of Benjamani, Krieger and Yekutieli was used.

Fig. 3. Reduction in murine endothelial IGF-1R expression alters the endothelial secretome and reveals a role for malonate in modulating white adipose function.

A Schematic representation of conditioned media experimental protocol. Created in BioRender. Luk, C. (2023) https://BioRender.com/x03i976. B Quantification of human primary adipocyte gene expression after 24 h treatment with conditioned media from primary murine endothelial cells isolated from 2-week high fat feeding (HFD) control and tamoxifen-inducible endothelial cell specific IGF-1R knockdown mice (ECIGF-1R^KD) mice (n = 4 and 4). p = 0.05, 0.01, 0.05, 0.03 and 0.01 for Ucp1, CIDEA, PPARGC1A, CYCS, TMEM26. C Quantification of human primary adipocyte gene expression after treatment with boiled conditioned media from primary murine endothelial cells isolated from 2-week HFD fed control and ECIGF-1R^KD mice (n = 4 and 4). p = 0.01 and 0.007 for Ucp-1 and Cidea respectively. D Volcano plot of small molecule analysis of the aqueous and lipid fractions of conditioned media from primary murine endothelial cell from 2-week HFD fed control and ECIGF-1R^KD mice (n = 4 and 4 per genotype). Red dots are significantly different between the genotypes. E Quantification of 3T3-L1 adipocyte gene expression of Ucp1 after upregulated metabolite stimulation (n = 3 and 3 for all treatment groups apart from PG, Quinolinic acid and shikimic acid n = 4 and 4 and malonic acid n = 7 and 5). p = 0.004. F Quantification of 3T3-L1 adipocyte gene expression of Cidea after upregulated metabolite stimulation (n = 3 and 3 per treatment group expect malonic acid n = 7 and 6). p = 0.001. G Quantification of gene expression in 3T3-L1 adipocytes after 24 h 10 mM malonic acid stimulation (n = 5 and 5 per gene except, Adipoq, Cd137, Ppara, Vegfa n = 4 and 4, Tbx1 N = 5 and 4, Cidea N = 7 and 6 and Ucp-1 n = 7 and 5). P = 0.03, 0.01, 0.00.1,0.004, 0.02, 0.01, 0.001, 0.0001, and 0.003 for Adipoq, Cd137, Cidea, Cited1, Fgf21, Ppargc1a, Pparg, Tmem26, Ucp1 and Vegfa respectively. Data shown as mean ± SEM, n is an individual experiment. p < 0.05 taken statistically significant using student unpaired two tailed t-test and denoted as * (p ≤ 0.01 and is denoted as **). For 3D an independent parametric Welch t-test, assuming two-group variances may differ was used. The statistical significance was determined by P-value with no multiple corrections.

Fig. 4. Therapeutic potential of malonate on adipose tissue function.

A Structural representation of malonic acid and Diacetoxymethyl malonate (MAM) used in vitro and in vivo. B Quantification of gene expression in human primary adipocytes after 24 h 25 µM MAM stimulation (n = 3 per treatment group). p = 0.04. C Quantification of respirometry in human primary adipocytes after 24 h 25 µM MAM stimulation (n = 4 per treatment group). p = 0.03. D Schematic representation of location of intra-epididymal adipose delivery of MAM (16 mg/kg). Created in BioRender. Luk, C. (2023) https://BioRender.com/b50h998. E Quantification of beiging gene expression from epididymal white adipose tissue (eWAT) from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). F Quantification of lipolysis gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). G Quantification of vascularisation gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). p = 0.03. H Quantification of adipokine production gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). p = 0.01 for Lep. I Quantification of mitochondrial function gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). p = 0.05, 0.01, 0.005 and 0.03 for Cycs, Ppargc1a, ppara, ppary respectively. J Quantification of lipid storage gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). p = 0.04. K Quantification of lipogenesis gene expression from eWAT from obese mice treated with MAM 16 mg/kg (n = 4 per treatment group). p = 0.04. Data shown as mean ± SEM, n is an individual experiment. p < 0.05 taken statistically significant using student unpaired two tailed t-test and denoted as * (p ≤ 0.01 and is denoted as **). For respirometry data a two-way ANOVA corrected for multiple comparisons by controlling the False Discovery Rate using two-stage step-up method of Benjamani, Krieger and Yekutieli, *p < 0.05.