Angiopoietin-like protein 8 differentially regulates ANGPTL3 and ANGPTL4 during postprandial partitioning of fatty acids

Abstract

Angiopoietin-like protein (ANGPTL)8 has been implicated in metabolic syndrome and reported to regulate adipose FA uptake through unknown mechanisms. Here, we studied how complex formation of ANGPTL8 with ANGPTL3 or ANGPTL4 varies with feeding to regulate LPL. In human serum, ANGPTL3/8 and ANGPTL4/8 complexes both increased postprandially, correlated negatively with HDL, and correlated positively with all other metabolic syndrome markers. ANGPTL3/8 also correlated positively with LDL-C and blocked LPL-facilitated hepatocyte VLDL-C uptake. LPL-inhibitory activity of ANGPTL3/8 was >100-fold more potent than that of ANGPTL3, and LPL-inhibitory activity of ANGPTL4/8 was >100-fold less potent than that of ANGPTL4. Quantitative analyses of inhibitory activities and competition experiments among the complexes suggested a model in which localized ANGPTL4/8 blocks the LPL-inhibitory activity of both circulating ANGPTL3/8 and localized ANGPTL4, allowing lipid sequestration into fat rather than muscle during the fed state. Supporting this model, insulin increased ANGPTL3/8 secretion from hepatocytes and ANGPTL4/8 secretion from adipocytes. These results suggest that low ANGPTL8 levels during fasting enable ANGPTL4-mediated LPL inhibition in fat tissue to minimize adipose FA uptake. During feeding, increased ANGPTL8 increases ANGPTL3 inhibition of LPL in muscle via circulating ANGPTL3/8, while decreasing ANGPTL4 inhibition of LPL in adipose tissue through localized ANGPTL4/8, thereby increasing FA uptake into adipose tissue. Excessive caloric intake may shift this system toward the latter conditions, possibly predisposing to metabolic syndrome.

Keywords: adipose tissue; angiopoietin-like protein 3; angiopoietin-like protein 4; lipid metabolism; lipoprotein lipase; metabolic syndrome; muscle; obesity; postprandial condition; triglycerides.

Fig. 1.

ANGPTL8 circulates in ANGPTL3/8 and ANGPTL4/8 complexes. A: Anti-ANGPTL8, anti-ANGPTL4, and anti-ANGPTL3 antibodies covalently coupled to beads, with heavy and light chains further cross-linked, were used to immunoprecipitate (IP) human serum. Proteins were separated on a 12% Bis-Tris gel and transferred to PVDF. Co-immunoprecipitating proteins were visualized via Western blotting. Results are representative of two independent experiments. B: ANGPTL8, ANGPTL4, and ANGPTL3 were immunoprecipitated from human serum. Beads were washed using PBS, and bound proteins were reduced with DTT and alkylated. Following digestion, digests were acidified, and co-immunoprecipitating proteins were quantified using a mass spectrometry LC-MRM method. Results are shown as the mean ± SEM (n = 3) from one experiment representative of two independent experiments.

Fig. 2.

ANGPTL3/8 and ANGPTL4/8 complexes increase with feeding. A: Recombinant human ANGPTL proteins and complexes used for immunoassays were characterized via electrophoresis. One microgram of each recombinant protein or complex was analyzed using gradient gel electrophoresis with a 4–20% Tris-glycine gel, followed by Coomassie Blue staining. B: Active ANGPTL4 (defined as full-length ANGPTL4 or the N-terminal fragment of ANGPTL4), CTDC ANGPTL4, ANGPTL3, ANGPTL8, ANGPTL3/8, and ANGPTL4/8 were measured in 50 normal donors using dedicated sandwich immunoassays. C: ANGPTL3/8, ANGPTL4/8, ANGPTL3, and ANGPTL8 were measured using dedicated sandwich immunoassays in 10 normal donors during fasting conditions and 1 and 2 h following a mixed meal challenge. Results are shown as the mean ± SEM. Significance for the feeding effect on ANGPTL proteins and complexes was assessed using a paired t-test.

Fig. 3.

ANGPTL3/8 blocks LPL-facilitated hepatocyte VLDL-C uptake. Cholesterol uptake in Huh7 hepatocytes was measured in the absence or presence of LPL pre-incubated with vehicle, ANGPTL3/8 complex, or ANGPTL4/8 complex for 1 h before mixing with fluorescent-labeled VLDL, followed by addition to the Huh7 hepatocytes for 30 min. The media was then replaced with fixative. Cells were fixed for 20 min, washed twice with PBS, and covered with PBS. Fluorescence at 495/525 nm was measured, with VLDL uptake calculated as relative fluorescent units at 525 nm. Results are shown as the mean ± SEM (n = 3).

Fig. 4.

ANGPTL3/8 and ANGPTL4/8 manifest different binding patterns to LPL. A: The ability of ANGPTL3 and ANGPTL3/8 to bind LPL was assessed with bio-layer interferometry. Avidin-tagged LPL was immobilized on streptavidin biosensors and incubated with ANGPTL3 or ANGPTL3/8 and transferred to buffer-only wells to monitor dissociation. The left side of the graph shows the association of ANGPTL3 and ANGPTL3/8 with LPL. The right side shows their respective dissociations. Results are representative of three independent experiments. B: The ability of ANGPTL4 and ANGPTL4/8 to bind LPL was assessed with bio-layer interferometry. Avidin-tagged LPL was immobilized on streptavidin biosensors and incubated with ANGPTL4 or ANGPTL4/8 and transferred to buffer-only wells to monitor dissociation. The left side of the graph shows the association of ANGPTL4 and ANGPTL4/8 with LPL. The right side shows their respective dissociations. Results are representative of three independent experiments.

Fig. 5.

ANGPTL8 markedly increases ANGPTL3 inhibition of LPL but dramatically decreases ANGPTL4 inhibition of LPL. A: The ability of ANGPTL3 or ANGPTL3/8 to inhibit LPL was assessed using LPL-stable expression cells incubated with ANGPTL3 or ANGPTL3/8 prior to the addition of lipase substrate. Fluorescence was monitored at 1 and 30 min to correct for background. ANGPTL3/8 showed a 186-fold increase in LPL inhibition compared to ANGPTL3 alone (IC₅₀ values of 0.14 nM versus 26 nM, respectively). Results are shown as the mean ± SEM (n = 5). B: The ability of ANGPTL4 or ANGPTL4/8 to inhibit LPL was similarly assessed. ANGPTL4/8 showed a 128-fold decrease in LPL inhibition compared to ANGPTL4 alone (IC₅₀ values of 37 nM versus 0.29 nM, respectively). Results are shown as the mean ± SEM (n = 6). C: ANGPTL4 or ANGPTL4/8 inhibition of LPL was assessed using VLDL as a substrate. The assay was similar to that used in A and B, except that lipase substrate was replaced with VLDL and free FAs were measured. ANGPTL4/8 showed a 239-fold decrease in LPL inhibition compared to ANGPTL4 alone (IC₅₀ values of 105 nM versus 0.44 nM, respectively). Results are shown as the mean ± SEM (n = 3).

Fig. 6.

ANGPTL4/8 blocks ANGPTL3/8- and ANGPTL4-mediated inhibition of LPL. A: To study the ability of ANGPTL4/8 to protect LPL from ANGPTL3/8 inhibition, various concentrations of ANGPTL4/8 were pre-incubated with LPL-stable expression cells for 1 h. Afterward, 1 nM of ANGPTL3/8 was added for a further 1 h incubation, prior to the addition of lipase substrate. Fluorescence was monitored as in Fig. 5A. Results are shown as the mean ± SEM (n = 4). B: To study the ability of ANGPTL4/8 to protect LPL from ANGPTL4 inhibition, various concentrations of ANGPTL4/8 were pre-incubated with LPL-stable expression cells for 1 h. Afterward, 1 nM of ANGPTL4 was added for a further 1 h incubation, prior to the addition of lipase substrate. Fluorescence was monitored as in Fig. 5A. Results are shown as the mean ± SEM (n = 6).

Fig. 7.

Insulin stimulates human hepatocyte secretion of ANGPTL3/8. A: Insulin-naïve patients (n = 279) were administered the hepatic-preferential insulin BIL, and serum samples were obtained under morning fasting conditions over the course of 1 year of BIL treatment. ANGPTL3/8 and ANGPTL4/8 levels were measured at baseline and after 12, 26, and 52 weeks of BIL administration. Results are shown as the mean ± SEM (*P < 0.0001 versus week 0). B: Human primary hepatocytes obtained in the HepatoPac platform were washed in serum-free application media and pre-incubated in application media in the absence of insulin. Following aspiration, cells were incubated with application media in the absence or presence of 1 nM of insulin. ANGPTL3/8 and ANGPTL4/8 levels in the media were measured using sandwich immunoassays, with the results shown as the mean ± SEM (n = 8).

Fig. 8.

Insulin stimulates ANGPTL4/8 secretion from human adipocytes. A: Human adipocytes were incubated in the absence or presence of insulin, and 1 μg of total RNA was reverse transcribed. ANGPTL8 transcript levels were quantitated. Insulin treatment resulted in an approximate 8-fold increase in ANGPTL8 mRNA levels. Results are shown as the mean ± SEM (n = 3). B: ANGPTL4 transcript levels were quantitated in the human adipocytes in A. Results are shown as the mean ± SEM (n = 3). C: Flag-tagged ANGPTL4 and HIS-tagged ANGPTL8 constructs were transfected into HEK293 cells. Afterward, dextran sulfate was added, media were harvested, and equal volumes from each condition were immunoblotted with anti-Flag or anti-HIS antibody. D: Human adipocytes were treated in heparin-containing media supplemented with 0–100 nM insulin in the absence and presence of 10 nM GIP. Media were collected and analyzed for ANGPTL4/8. Results are shown as the mean ± SEM (n = 6, *P < 0.0001 versus control).

Fig. 9.

A possible model for how ANGPTL8 shifts FA toward adipose tissue after feeding. A: While fasting, ANGPTL8 levels are low. Localized ANGPTL4 inhibits adipose tissue LPL to minimize FA uptake into the fat for storage, and FAs are mainly taken up into skeletal muscle for use as energy. B: During feeding, ANGPTL8 forms a circulating complex with ANGPTL3 that increases its ability to inhibit LPL, thus minimizing FA uptake into skeletal muscle. ANGPTL8 also forms a mostly localized complex with ANGPTL4 in adipose tissue that decreases the ability of ANGPTL4 to inhibit LPL. The ANGPTL4/8 complex also protects LPL in the fat from inhibition by circulating ANGPTL3/8 and localized ANGPTL4 (denoted by red Xs), thereby preserving adipose tissue LPL activity to promote FA uptake into the fat for storage as TG.