Inhibition of adipocyte lipolysis by vaspin impairs thermoregulation in vivo

Abstract

Altered activity of brown adipose tissue (BAT) contributes to obesity, insulin resistance, and cardiovascular disease. BAT secretes endocrine factors ("batokines") that regulate thermogenesis. We identify the serpin vaspin as a batokine that modulates adrenergic control of lipolysis and thermogenesis. Adipocyte-specific vaspin overexpression in mice reduces BAT activation and impairs thermoregulation during cold exposure or fasting. Mechanistically, vaspin binds low-density lipoprotein receptors (LRP1, LDLR, vLDLR), inhibiting adrenergic signaling and lipolysis in brown and white adipocytes by modulating phosphodiesterase activity and endocytic lipid uptake. Gene set enrichment analyses in human subcutaneous adipose tissue and in vitro studies confirm vaspin's anti-lipolytic effects in humans. Overall, vaspin emerges as a regulatory BATokine that fine-tunes BAT thermogenic activity to limit excessive energy expenditure and preserve metabolic balance.

Fig. 1. VasTg mice show impaired thermoregulation.

A Serum levels of human vaspin (SERPINA12) measured by ELISA of VasTg and WT mice (n = 6/6). B, C Rectal body temperature during acute cold exposure in fed (B), n = 9/9) and fasted (C), n = 8/7) VasTg and WT mice. D Serum NEFA levels of VasTg and WT mice after acute (4 h) cold exposure (n = 4/4). E Blood glucose levels of VasTg and WT mice before and after acute (6 h) cold exposure (n = 4/4). F Serum levels of human vaspin (SERPINA12) of VasTg housed at 30 °C, 23 °C, for 6 h at 8 °C and for one week at 8 °C measured by ELISA (n = 7/6/6/5). G Representative H&E-stained sections of BAT (20x; and magnifications) from VasTg and WT mice housed at 23 °C and 8 °C for 7 days (close ups from indicated area and 2 additional animals per group). H mRNA expression of thermogenic and metabolic genes in BAT of VasTg and WT mice after cold exposure (24 h; n = 4–10). Gene expression is relative to controls, normalized to 36b4. I mRNA expression of genes related to shivering thermogenesis in muscle of VasTg and WT mice after cold exposure (24 h; n = 9 per group). Gene expression is relative to controls, normalized to Hprt. J KEGG pathway functional enrichment analysis of DEGs. The vertical axes represent the KEGG pathways significantly enriched; the horizontal axis indicates -log10(p-value). K–N Rectal body temperature (K), thermal images from BAT and tail (L), BAT surface (M) and tail surface (N) temperatures, as well as blood glucose levels (O) in fasted vaspin-treated (intraperitoneally, 1 mg/kg) and control C57BL/6 N mice (n = 7/8 per group) before and after acute (6 h) cold exposure. Data are shown as mean ± SEM. WT or control samples are indicated as black circles, VasTg samples as orange circles and Vaspin-treated samples as green circles. Statistical significance was evaluated by two-way ANOVA with Šídák’s (B, C, J, K, O) or Dunnett’s (F) post-hoc test or uncorrected Fischer’s LSD (H, I), or unpaired two-sided t-tests (A, D, M, N). *p value < 0.05, **p value < 0.01, ***p value < 0.001.

Fig. 2. Vaspin induces changes in cellular (phospho-) proteome and inhibits activation of brown adipocyte metabolism and thermogenesis.

A Experimental design (created in BioRender, https://BioRender.com/sxodmnd). Differentiated immortalized brown adipocytes (imBA) were exposed to recombinant vaspin (0.5 µg/mL) for 30 min or 6 h, respectively. Proteome and phosphoproteome were created from proteins of the same samples. B, C KEGG pathway functional enrichment analyses of affected phosphorylation sites (after 30 min, B and protein abundances (after 6 h, (C). The vertical axis represents KEGG pathways significantly enriched, and the horizontal axis indicates -log10 (p-value). Significantly enriched KEGG pathways are labeled in red (increased) and blue (decreased). For details, please see “Methods” section. D Western blot analysis of basal and CL-induced PKA-activation in vaspin treated (0.2 – 1 µg/ml) imBA cells. E–G Western blot analysis and (F) quantification of basal and CL-induced PKA-activation as well as (G) free fatty acid release (NEFA, G), n = 6 per condition) in vaspin treated differentiated imBA. H–J Western blot analysis and (I) quantification of basal and CL-induced PKA and HSL activation as well as (J) free fatty acid release (NEFA), n = 6 per condition) in differentiated primary brown adipocytes from VasTg and WT mice. K, L Time-resolved OCR (K) of differentiated primary brown adipocytes from VasTg and WT mice measured by Seahorse (representative experiment, n = 5/5) and (L) quantification of basal respiration, ATP production, proton leak, acute response to FSK, maximum and spare respiratory capacity and non-mitochondrial respiration. M Time-resolved ECAR corresponding to the OCR shown in (K). N Energetic profile (OCR / ECAR plot from K and M) of differentiated primary brown adipocytes from VasTg and WT mice. O Lactate release from ISO and FSK treated primary brown adipocytes from VasTg and WT mice (n = 5/5). WT or control samples are indicated as black circles, VasTg samples as orange circles and Vaspin-treated samples as green circles. Data are presented as mean ± SEM of at least two (D, G) or three (E, F, H–N) independent experiments. Statistical significance was evaluated by two-way ANOVA with Šídák’s (L, O) or Tukey’s (F, G, J) post-hoc test or uncorrected Fischer’s LSD (I). *p value < 0.05, **p value < 0.01, ***p value < 0.001.

Fig. 3. Vaspin inhibition of adrenergic signaling involves LRP1 binding and modulation of PDE activities.

A Regulatory mechanisms of adrenergic signaling and levels of intervention (created in BioRender, https://BioRender.com/sxodmnd). B Cell culture supernatant of human vaspin (SERPINA12) after 3 h starvation (control) and parallel blocking of vaspin secretion using BFA (BFA–after 3 h prior medium change, final–post stimulation with CL) before signal transduction and lipolysis assays in differentiated primary brown adipocytes from VasTg measured by ELISA (n = 3 per condition). C–E Western blot analysis and (D) quantification of basal and CL-induced PKA-activation (n = 3 (basal) and 4 (CL)) as well as (E) free fatty acid release in differentiated primary brown adipocytes from VasTg and WT mice after blocking vaspin secretion (NEFA, n = 5–12 per condition). F BAT Lrp1 expression in WT mice housed at 30 °C, 23 °C or 8 °C (n = 6 per temperature). G–J Knockdown of LRP1: (G) Western blot analysis and (H) quantification of LRP1 expression, (I) basal and CL-induced PKA-activation as well as (J) free fatty acid release (NEFA, n = 10–12 per condition, please see source data file) in vaspin-treated differentiated imBA, with or without siRNA-mediated Lrp1 knockdown. K, L Non-heparin-binding (NHB) vaspin variant: (K) Western blot analysis and (L) quantification basal and CL-induced PKA-activation in vaspin and NHB-treated and control differentiated imBA (n = 3/3). M LRP1 ligand RAP: Western blot analysis of basal and CL-induced PKA-activation in RAP or vaspin treated differentiated imBA. N Blocking of vaspin binding to LRP1 (RAP preincubation) or LRP1 internalization by clathrin-mediated endocytosis (using CPZ): Quantification of basal and CL-induced free fatty acid release with or without vaspin treatment in differentiated imBA (n = 19–20 per condition, please see source data file). O, P Inhibition of Gas signaling using PTX: (O) Western blot analysis and (P) quantification of basal and NE-induced PKA-activation in vaspin-treated differentiated imBA with PTX pretreatment (n = 3/3). Q, R Inhibition of PDE activities using IMBX: (Q) Western blot analysis and (R) quantification of basal and CL-induced PKA-activation in differentiated primary brown adipocytes from VasTg and WT mice (n = 3/3). S–U Inhibition of PDE3 and PDE4 using Cilo and Ro: S) Western blot analysis and (T) quantification of basal and CL-induced PKA-activation in differentiated primary brown adipocytes from VasTg and WT mice with Cilo or Ro pretreatment and controls (n = 3/3/3). U Inhibition of total PDE activities using IBMX, Cilo or Ro: Quantification of basal and CL-induced free fatty acid release with or without vaspin treatment in differentiated imBA (n = 10–20 per condition, please see source data file). V-X ELISA-based analysis of TAMRA-vaspin binding to LDL receptors (V) - LDLR in black, vLDLR in red; LRP receptors (W) - LRP2 in black, LRP4 in red, LRP5 in orange, LRP6 in yellow, LRP10 in green); and class B scavenger receptors (X) - SR-BI in black, CD36 in red. Nonlinear regression analysis was performed to determine EC₅₀ presented in Table 1. Data are presented as mean ± SEM of at least two (B–E, J–L, M–U) or three (G–I) independent experiments. WT or control samples are indicated as black circles, VasTg samples as orange circles, Vaspin-treated samples as green circles, Vaspin NHB-treated samples as blue circles. Statistical significance was evaluated by one-way ANOVA with Šídák’s (B, D, E, L) or uncorrected Fischer’s LSD (N, R, U), or two-way ANOVA with Dunett’s (I) or Tukey’s (F, J, P) post-hoc or uncorrected Fischer’s LSD (T), or unpaired two-sided t-test (H). *p value < 0.05, **p value < 0.01, ***p value < 0.001.

Fig. 4. Vaspin suppresses adrenergic activation and lipolysis of white adipose tissue in mice and humans.

A–E Rectal body temperature (A), thermal images from BAT (B), BAT surface (C) and tail surface (D) temperatures, as well as blood glucose levels (E) in VasTg and WT during a fasting-refeeding cycle (A): n = 8/9, C, D: n = 4/4). F–I Western blot analysis (F, H) and quantification (G, I) of basal and CL-induced PKA-activation in differentiated primary iWAT (F, G) and eWAT (H, I) adipocytes from VasTg and WT mice. J Gene set enrichment analysis (GSEA, using KEGG pathways) of genes correlating with SERPINA12 in subcutaneous adipose tissue (SAT). The top 10 enriched pathways for positively and negatively correlating genes are shown sorted by adj. p and set size (please see “Methods” section). K Overlay images of immunofluorescence microscopy images of human mature SAT adipocytes from one patient stained for LRP1 (green) and treated with TAMRA-labeled vaspin (red). L Brightfield images of differentiated SAT SVF-derived human adipocytes from four different patients. M Basal and ISO-induced free fatty acid release in vaspin treated (0.5 µg/ml) differentiated SAT SVF-derived human adipocytes from four patients (n = 9-10 per condition and patient, please see source data file). N, O Western blot analysis (I) and quantification (J) of basal and ISO-induced PKA-activation of differentiated SAT SVF-derived human adipocytes. Data are presented as mean ± SEM of at least two (independent experiments (G, I) or from one the four patients analyzed (O). Statistical significance was evaluated by two-way ANOVA with Šídák’s (A–D) or Tukey’s (G, I) post-hoc test or uncorrected Fischer’s LSD (E, M), or one-way ANOVA with uncorrected Fischer’s LSD (O). *p value < 0.05, **p value < 0.01, ***p value < 0.001. Scale bar: 50 µm.

Fig. 5. Proposed mechanisms and interplay of thermogenic brakes via the vaspin-LRP1 and kallikrein-kinin pathways.

The vaspin-LRP1 thermogenic brake is reported here, together with possible cross talk with the kallikrein-kinin thermogenic brake reported by Peyrou and colleagues [53]. ADCY adenylate cyclase, ADRB adrenergic receptor, BK bradikinin, B1/B2 bradykinin receptors B1/B2, Ga G-alpha protein, KLK kallikrein, LRP1 low-density lipoprotein receptor 1, NE norepinephrine, PDE phosphodiesterase, PKA protein kinase A. Created in BioRender, https://BioRender.com/sxodmnd.