Identification of lysosomal lipolysis as an essential noncanonical mediator of adipocyte fasting and cold-induced lipolysis

Abstract

Adipose tissue lipolysis is the process by which triglycerides in lipid stores are hydrolyzed into free fatty acids (FFAs), serving as fuel during fasting or cold-induced thermogenesis. Although cytosolic lipases are considered the predominant mechanism of liberating FFAs, lipolysis also occurs in lysosomes via lysosomal acid lipase (LIPA), albeit with unclear roles in lipid storage and whole-body metabolism. We found that adipocyte LIPA expression increased in adipose tissue of mice when lipolysis was stimulated during fasting, cold exposure, or β-adrenergic agonism. This was functionally important, as inhibition of LIPA genetically or pharmacologically resulted in lower plasma FFAs under lipolytic conditions. Furthermore, adipocyte LIPA deficiency impaired thermogenesis and oxygen consumption and rendered mice susceptible to diet-induced obesity. Importantly, lysosomal lipolysis was independent of adipose triglyceride lipase, the rate-limiting enzyme of cytosolic lipolysis. Our data suggest a significant role for LIPA and lysosomal lipolysis in adipocyte lipid metabolism beyond classical cytosolic lipolysis.

Keywords: Adipose tissue; Endocrinology; Lysosomes; Metabolism; Obesity; Therapeutics.

Figure 1. LIPA in adipocytes is stimulated by fasting, CE, and β-agonists.

(A) Schematic illustration of cytosolic versus lysosomal lipolysis and key enzymes involved. (B) Experimental information and log₂ fold changes of lipase gene expression in datasets (–63) characterizing eWAT during fasting. (C) Gene expression of Lipa and cytosolic lipases and (D) protein expression of LIPA and ATGL (PNPLA2) in eWAT from mice fasted 16 hours (n = 3–4). (E) log₂ Fold changes of Lipa and ATGL (Pnpla2) gene expression in datasets (–68) examining CE or CL treatment. (F) Gene expression of Lipa and ATGL (Pnpla2) (n = 8) and (G) protein expression of LIPA and ATGL (PNPLA2) (n = 3–4) in iWAT from mice housed at 4ºC (left panel) or treated with CL (right panel) for 1 or 3 days. (H) Gene expression of Lipa and ATGL (Pnpla2) (n = 5–6) and (I) protein expression of LIPA and ATGL (PNPLA2) in differentiated primary adipocytes derived from iWAT SV cells treated with isoproterenol for indicted durations. (J) Gene expression of Lipa and ATGL (Pnpla2) (n = 4) and (K) protein expression of LIPA and ATGL (PNPLA2) in C3H1T1/2 adipocytes treated with isoproterenol for indicated durations. All mice were male and fed an ND. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test (C, D, H, and J) or 1-way ANOVA with a post hoc Tukey’s HSD (F and G) for comparisons with the indicated groups. *P < 0.05; **P < 0.01; ***P < 0.001. See also associated Supplemental Figures 1–4.

Figure 2. LIPA disruption suppresses fasting-, cold-, and β-agonist-induced lipolysis in mouse models.

(A) Experimental strategy outlining Adipoq-Cre–driven knockout of LIPA to generate A-Lipa KO mice. (B) Characterization of LIPA depletion at the mRNA (n = 7 in iWAT; n = 3 in BAT, eWAT and liver) and (C) protein levels in tissues from control (Ctrl) versus A-Lipa KO mice. (D) Lipa gene expression in homogenized eWAT separated by centrifugation into floating adipocyte and pelleted SV fractions from control (n = 6) and A-Lipa KO (n = 7) mice. (E) Schematic illustration of lipolysis assay by using genetic knockout mice model, A-Lipa KO mice, or pharmacological inhibition via lalistat-2 treatment. (F) Plasma FFA levels were measured at indicated time points in fasted A-Lipa KO and control mice (n = 12) (left) or in C57BL/6J mice fasted for 16 hours and injected intraperitoneally with 30 mg/kg lalistat-2 (n = 11) or vehicle (n = 10), followed by an additional 4 hours fasting. (G) Plasma FFA measured in A-Lipa KO and control mice (n = 4 each) fasted and individually housed at 4°C for indicated time points (left panel) or in C57BL/6J mice (n = 9 or 11) injected with 30 mg/kg lalistat-2 or vehicle 1 hour prior to fasting and individual housing at 4°C (right panel). (H) Plasma FFA (n = 4) measured at indicated time points in A-Lipa KO and control mice fasted for 16 hours then intraperitoneally injected with 1 mg/kg CL (left panel) or in C57BL/6J mice fasted for 16 hours, injected with 30 mg/kg lalistat-2 (n = 9) or vehicle (n = 8), and fasted for an additional 90 minutes prior to administration of 1 mg/kg CL without refeeding for indicated durations (right panel). All mice were male and fed an ND. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test compared with their control group. *P < 0.05; **P < 0.01; ***P < 0.001. See also Supplemental Figures 4–10.

Figure 3. LIPA disruption suppresses β-agonist-induced adipose lipolysis in cultured adipocytes and WAT explant.

(A) Schematic illustration of strategy to generate LIPA knockout primary adipocytes through differentiating the SV cells isolated from the iWAT of A-Lipa KO and control mice. (B) Confirmation of LIPA depletion at mRNA and (C) protein expression levels in Lipa KO adipocytes (n = 6). (D) Isoproterenol-induced lipolysis measured as glycerol and FFA release in supernatants from Lipa KO adipocytes (n = 6) or (E) lalistat-2–treated primary adipocytes (n = 3). (F) Glycerol and FFA levels of iWAT explant supernatant from A-Lipa KO versus control mice or (G) from lalistat-2– or vehicle-injected C57BL/6J mice treated with or without 1 μM isoproterenol at indicated time points (n = 4). All mice were male and fed an ND. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test (B and C) or 2-way ANOVA with a post-hoc Tukey’s HSD test (D–G) for comparisons with the indicated groups (baseline groups: **P < 0.01; ***P < 0.001, or iso-treated groups: ^#P < 0.05; ^##P < 0.01; ^###P < 0.001). See also Supplemental Figure 11–15.

Figure 4. Adipocyte LIPA deficiency disrupts the regulation of body temperature and oxygen consumption upon CE and CL treatment.

(A) Strategy to phenotype the role of LIPA in modulation of body temperature and oxygen consumption using genetic knockout or pharmacological inhibitor. (B) Core body temperatures monitored in A-Lipa KO and control mice (left panel) (n = 8 and 7) or in C57BL/6J mice (right panel) (n = 9 or 11) injected with 30 mg/kg lalistat-2 or vehicle 1 hour prior to indicated duration of individual housing at 4°C without food. (C) Body temperature (n = 10) and (D) oxygen consumption (n = 6) measured at indicated time points in A-Lipa KO and control mice intraperitoneally injected with 1 mg/kg CL or in C57BL/6J mice injected with 30 mg/kg lalistat-2 or vehicle 2 hours prior to CL treatment. All mice were male and fed an ND. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test (B) or 2-way ANOVA (C and D) compared with baseline groups: *P < 0.05; **P < 0.01. See also Supplemental Figure 16.

Figure 5. Adipocyte LIPA is associated with obesity and affects adipose tissue morphology.

(A) Study information and log₂ fold changes of lipase gene expression in indicated study datasets (, –72) profiling eWAT from models of obesity including mice or rats fed high-fat high sugar diet (HFHSD) or HFD or those which were genetically predisposed (ob/ob model). (B) Gene expression of lipases in eWAT from ND- or HFD-fed mice (n = 10 or 8). (C) Lipa gene expression in adipocyte fractions isolated from eWAT of ND- or HFD-fed mice (n = 5 or 6). (D) Correlations between Lipa expression in eWAT and body weight (left panel), fat mass (middle panel), or eWAT weight (right panel) among ND and HFD-fed mice. (E) Experimental strategy outlining characterization of adipose tissue in A-Lipa KO and control mice. (F) Adipose tissue weight (n = 10 and 11) and (G) H&E-stained tissue sections of iWAT (top panel), eWAT (middle panel), and BAT (bottom panel) with (H) quantification of adipocyte size (n = 6) from A-Lipa KO and control mice at 16 weeks of age. Scale bars: 20 μm. All mice were male. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test compared with control or ND groups: *P < 0.05; **P < 0.01; ***P < 0.001. The correlation (r square) and P value were calculated by Pearson’s r test. See also Supplemental Figures 17–19.

Figure 6. Adipocyte LIPA deficiency promotes the development of diet-induced obesity.

(A) Experimental strategy outlining characterization of A-Lipa KO and control mice in response to diet-induced obesity with metabolic assays performed. (B) Body weight, (C) body composition, (D) adipose depot tissue weights, and (E) H&E-stained tissue sections of iWAT (left panel), eWAT (middle panel), and BAT (right panel) from A-Lipa KO and control mice after 16 weeks of HFD treatment (n = 15). Scale bars: 100 μm. (F) Oxygen consumption and (G) RER in A-Lipa KO and control mice 10 weeks into the HFD study (n = 5). (H) Glucose tolerance test and calculated area under the curve in A-Lipa KO and control mice assessed following 12 weeks of HFD (n = 10). (I) Insulin tolerance test and calculated area under the curve in A-Lipa KO and control mice assessed following 10 weeks of HFD (n = 9). All mice were male. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test compared with ND groups: *P < 0.05; **P < 0.01; ***P < 0.001. See also Supplemental Figure 20.

Figure 7. Lipolysis mediated by LIPA is independent of ATGL.

(A) Expression of mRNA (n = 3–4) and (B) protein (n = 6–8) in lipases, ATGL (PNPLA2), and HSL (LIPE), and ATGL-related cofactors G0S2, HILPDA, and CGI-58 (ABHD5) in eWAT from A-Lipa KO and control mice at 12 weeks old. (C) Expression of mRNA and (D) protein in lipases, ATGL (PNPLA2), and HSL (LIPE), and ATGL-related cofactors G0S2, HILPDA, and CGI-58 (ABHD5) in Lipa KO adipocytes (n = 6). (E) Protein expression of lipases, ATGL (PNPLA2), and HSL (LIPE), and ATGL-related cofactors G0S2, HILPDA, and CGI-58 (ABHD5) in lipid droplets isolated from eWAT of A-Lipa and control mice at 12 weeks old. (F) Glycerol and FFA release into the supernatant of isoproterenol-stimulated iWAT explants from A-Lipa KO versus control mice with or without atglistatin cotreatment (n = 6). (G) Glycerol and FFA release into the supernatant of isoproterenol-stimulated Lipa KO versus control adipocytes with or without atglistatin cotreatment (n = 4). All mice were male. Values are presented as mean ± SEM. Significant differences were determined by Student’s t test (A–D) or by 2-way ANOVA with a post hoc pairwise t test (F and G) for comparisons with the indicated groups: *P < 0.05; **P < 0.01; ***P < 0.001. See also Supplemental Figures 21 and 22.