The beta secretase BACE1 regulates the expression of insulin receptor in the liver

Abstract

Insulin receptor (IR) plays a key role in the control of glucose homeostasis; however, the regulation of its cellular expression remains poorly understood. Here we show that the amount of biologically active IR is regulated by the cleavage of its ectodomain, by the β-site amyloid precursor protein cleaving enzyme 1 (BACE1), in a glucose concentration-dependent manner. In vivo studies demonstrate that BACE1 regulates the amount of IR and insulin signaling in the liver. During diabetes, BACE1-dependent cleavage of IR is increased and the amount of IR in the liver is reduced, whereas infusion of a BACE1 inhibitor partially restores liver IR. We suggest the potential use of BACE1 inhibitors to enhance insulin signaling during diabetes. Additionally, we show that plasma levels of cleaved IR reflect IR isoform A expression levels in liver tumors, which prompts us to propose that the measurement of circulating cleaved IR may assist hepatic cancer detection and management.

Fig. 1

Detection and characterization of IR fragments. a Schematic representation of IR showing the position of the peptides used as antigens to produce antibodies C-19, C-4, H-78, and that of the epitope recognized by antibody 18-44 (blue rectangles). N and C-terminal extremities and α and β subunits are localized. The membrane is represented by the light gray rectangles. b HEK 293 cells were transfected with IR expression vector and treated overnight with the γ-secretase inhibitor (DAPT; 5 μM). IR was detected by immunoblot, using the indicated IR β-subunit (IRβ) specific antibodies (LE indicates a long exposure of the immunoblot). Positions of IR precursor (proIR), IRβ, and C-terminal fragment of IR (IRctf) are indicated. c Immunoblot analysis using the H-78 antibody of 24 h serum-free conditioned media (CM) from cells transfected with an empty plasmid (−) or with a plasmid encoding human IR (+). d Lysates (L) and CM from IR overexpressing cells were collected and where indicated in vitro, deglycosylated by PGNaseF. IR was detected by immunoblot, using H-78 antibody (upper panel) or C-19 antibody (lower panel). e CM and cell lysates from IR overexpressing cells (+) or cells transfected with an empty plasmid (−) were collected and subjected to immunoprecipitation (IP), using the indicated antibodies. IR was detected by immunoblot using H-78 antibody. IRα indicates the position of IR α-subunit

Fig. 2

BACE1 is involved in IR cleavage. IRsol was measured by ELISA in 24 h conditioned media of a HEK 293 cells transfected with an empty vector (EV) or overexpressing IR. The mean IRsol values from IR overexpressing cells were arbitrarily set at 1. b Cells overexpressing IR and untreated (Cont) or treated with the indicated protease inhibitors (10 μM, 24 h). c Cells overexpressing IR together with BACE1 specific shRNA (shBACE1a, shBACE1b) or a control shRNA (shcont). d Cells transfected with the IR expression vector together with an empty plasmid (Cont), BACE1, inactive BACE1 (BACE1i), BACE2, ADAM10, or ADAM17. Expression of BACE1 and BACE2 were confirmed by immunoblot, using HA antibody (lower panel). e Cells transfected with combinations of IR and BACE1 expression vectors were treated or not with DAPT, then IR and BACE1 were detected by immunoblot. LE indicates a long exposure of the immunoblot. The graph shows the ratios IR/proIR obtained from the quantification of six independent experiments; data from the same experiment are connected by a line. f Cells overexpressing IR and BACE1 or inactive BACE1 (BACE1i) were treated with the BACE1 inhibitor (C3) then IR and BACE1 were detected by immunoblot. g Cells were treated as f, then BACE1 was immunoprecipitated (IP) with anti HA antibody and IR, and BACE1 were detected by immunoblot in cell lysates and immunoprecipitated fractions. h Cells were transfected with the indicated combinations of IR, IRFlag, and BACE1i expression vectors, then overexpressed proteins were immunoprecipitated using C-19, FlagM2 (Flag) and HA antibodies, respectively, and detected by immunoblots. i BACE1 overexpressing cells treated with proprotein convertase inhibitor dec-RVKR-cmk (RVKR; 24 h; upper panel) or cells overexpressing BACE1 without (−) of with furin (fur; lower panel) were lyzed then BACE1 was immunoprecipitated and detected by immunoblot. Positions of the heavy chain of the precipitating antibody (Ig), immature (imm) and mature (mat) BACE1 are indicated. Data are means ± s.d. Statistical analyses were made using unpaired t-test (a, e) or ANOVA followed by Dunnett’s test (b–d); **p < 0.01; ***p < 0.001

Fig. 3

BACE1 regulates the amount of cell surface IR. a HEK 293 cells, stably expressing BACE1, were co-transfected with IR expression vector and control shRNA (shcont) or BACE1 specific shRNA (shBACE1) then IR and BACE1 were analyzed by immunoblot. LE indicates a long exposure of the immunoblot. The graph shows the ratio IR/proIR obtained from the quantification of six independent experiments; data from the same experiment are connected by a line. b IR expression was measured by flow cytometry at the surface of cells, stably expressing BACE1 and co-transfected with the IR expression vector associated with control shRNA (shcont) or BACE1 specific shRNA (shBACE1). The median fluorescence intensity (MFI) corrected for the value obtained with cells transfected with the empty vector is shown. c Cells expressing BACE1 were transfected with IR expression vector and treated overnight with the BACE1 inhibitor (C3; 20 μM) in the absence of serum, then stimulated with insulin (5 nM, 5 min) and IR, and phospho IR (pIRβ) were analyzed by immunoblot. The graphs show the ratios IR/proIR and pIR/IR obtained from the quantification of seven and six independent experiments, respectively. Native HEK 293 cells (d) or HEK 293 cells expressing BACE1 (e) were transfected with IR coding vector (+) or with empty vector (−), treated with BACE1 inhibitor (C3; 20 μM; gray columns), serum deprived for 20 h and stimulated with insulin (5 nM, 45 min; +). Levels of c-Fos and Egr1 mRNA were measured by RT-PCR and expressed as fold over the situation without IR overexpression nor insulin stimulation. Data are means ± s.d. Statistical analyses were made using unpaired t-test (a–c) and ANOVA followed by Bonferroni’s test (e); *p < 0.05; **p < 0.01

Fig. 4

Identification of the IR cleavage region. a Schematic representation of IR positioning the N-glycosylation sites in the β-subunit. b HEK 293 cells were transfected and treated as described in Fig. 1b. Where indicated cell lysates were in vitro deglycosylated by PGNaseF. c CHO cells were transfected with an empty plasmid (Cont), an IR expression vector (WT) or with a plasmid expressing a mutated form of IR that cannot be glycosylated on its β-subunit (4NA), then treated with DAPT. d HEK 293 cells were transfected with an IR expression vector (WT) or with a plasmid expressing a mutated form of IR that cannot be glycosylated on the position 933 (N₉₃₃A), then treated with DAPT. HEK 293 cells expressing BACE1 were transfected with wild-type IR or the indicated mutated forms of IR, then the IRsol accumulated in the conditioned media was measured by ELISA, and cellular expression of IR was analyzed by immunoblot (e), cell surface expression of IR was measured by flow cytometry (f). g Cells were treated as in Fig. 3c then IR and phospho IR (pIRβ) were detected by immunoblot. LE indicates a long exposure of the immunoblot. The graph shows the ratios IR/proIR obtained from the quantification of six independent experiments. Data are means ± s.d. Statistical analyses were made using ANOVA, followed by Bonferonni’s test (e, f) or unpaired t-test (g). *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 5

Effect of glucose on IR cleavage and BACE1 expression. HEK 293 cells expressing IR and BACE1 were incubated for 24 h in media containing the indicated glucose concentrations, in the absence (−) or presence of deoxynorleucine (DON; 5 mM) or glucosamine (GlcN; 1 mM) then: a IRsol accumulated in the media was measured by ELISA; b IR and BACE1 were analyzed by immunoblot in total cell lysate or after BACE1 immunoprecipitation, respectively. Positions of the heavy chain of the precipitating antibody (Ig), immature (imm), and mature (mat) BACE1 are indicated. The graph shows the ratios BACE1/proBACE1 obtained from the quantification of seven different experiments; data from the same experiment are connected by a line. Data are means ± s.d. Statistical analyses were made using ANOVA, followed by Bonferonni’s test (a) and unpaired t-test (b). *p < 0.05; ***p < 0.001

Fig. 6

In vivo cleavage of IR by BACE1. IRsol was measured in the plasma from a liver-specific IR knockout mice (LIRKO; fed n = 4; fasted n = 5)and their floxed control (IRflox; fed n = 4; fasted n = 5), b BACE1^−/− mice (n = 8) and their control littermates (n = 8). c IRsol was measured in the conditioned media from control and BACE1^−/− liver explants treated or not with the BACE1 inhibitor (C3). d Levels of IR, IRA, and IRB mRNA were measured by RT-PCR in control and BACE1^−/− livers. e Levels of IR, PTEN, and GAPDH (loading control) were analyzed by immunoblot in control and BACE1^−/− livers, a densitometric analysis of the proIR and IRβ bands normalized to GAPDH was performed (below the blot), values are expressed as fold over the means of the wild-type mice. f Primary hepatocytes isolated form control (n = 4) and BACE1^−/− (n = 3) mice were stimulated for 10 min with the indicated concentration of insulin. A representative immunoblot of insulin-stimulated phosphorylation of PKB at Ser⁴⁷³ is shown. The curves on the right of the immunoblot are the normalized means ± s.e.m of the immunoblots (both curves differed significantly, p < 0.0001, F-test). g mRNA levels of PDK4 were measured by RT-PCR in control and BACE1^−/− primary hepatocytes stimulated for 6 h with the indicated concentrations of insulin (both curves differed significantly, p < 0.01, F-test). h Human primary hepatocytes (two different preparations) were incubated for 40 h in the presence of the indicated concentrations of BACE1 inhibitor (C3), then IR and GAPDH or actin (taken as loading controls) were analyzed by immunoblot. Data are means ± s.d. Statistical analyses was made using Mann–Withney (a), unpaired t-test (b, d, e), ANOVA followed by Dunnett’s test (c) or F-test (f, g) *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 7

Regulation of glucokinase expression by IR amount. a Levels of GCK mRNA were measured by RT-PCR in control and BACE1^−/− primary hepatocytes. Native HEK 293 cells (b) or HEK 293 cells overexpressing BACE1 (c) were transfected with empty vector (EV) or with IR coding vector together with human GCK promoter-driven Firefly luciferase and SV40-driven Renilla luciferase coding vectors where indicated, serum-deprived cells were treated with the BACE1 inhibitor (C3; 20 μM) and then stimulated with insulin (2 nM, 7 h). Firefly and Renilla luciferase were measured in cell lysates, and GCK promoter activity was calculated as the ratio of Firefly/Renilla luciferase and expressed as fold over the situation without IR overexpression nor insulin stimulation. Data are means ± s.d. Statistical analyses were made using unpaired t-test (a) or ANOVA, followed by Bonferonni’s test (b, c). *p < 0.05; **p < 0.01

Fig. 8

IR cleavage in a mouse model of diabetes. a IRsol was measured in the plasma from db/db mice (n = 10) and from their control littermate (db/+; n = 10). b mRNA levels of IR and BACE1 were measured by RT-PCR in the liver. c Liver expression of IR, BACE1, and β-actin (loading control) were analyzed by immunoblot; a densitometric analysis of the proIR, IRα, IRβ, and BACE1 bands was performed (right), values are expressed as fold over the means of the wild-type mice. d db/db mice were treated with saline or with BACE1 inhibitor (C3, 4 weeks), then liver expression of IR and GAPDH (loading control) were analyzed by immunoblot; a densitometric analysis of the proIR and IRβ bands was performed (right). e mRNA levels of IR were measured by RT-PCR in the liver of db/db mice treated with saline or with the BACE1 inhibitor (C3). Data are means ± s.d. Statistical analyses were made using unpaired t-test. **p < 0.01; ***p < 0.001

Fig. 9

BACE1-dependent cleavage of IRA and IRB. a HEK 293 cells were transfected with an empty vector (Cont) or with IRA or IRB expression vector, then IR and endogenous GAPDH expression were detected by immunoblot. b IRsol was measured in conditioned media of cell expressing IRA or IRB by ELISA (left) and by immunobloting of the conditioned media untreated or treated with PGNaseF. c Cells were transfected with the indicated combinations of IRA, IRB and BACE1, and inactive BACE1 (BACE1i) expression vectors; then IRsol was measured by ELISA in conditioned media. d BACE1 overexpressing cells were transfected with IRA or IRB expression vector and IR was detected by immunoblot (left) or at the cell surface by flow cytometry (right). LE indicates a long exposure of the immunoblot. e Inactive BACE1 was co-immunoprecipitated with IRA or IRB as described in Fig. 2g. Immature (imm) and mature (mat) BACE1 are indicated. Data are means ± s.d. Statistical analyses were made using unpaired t-test. *p < 0.05; ***p < 0.001

Fig. 10

IR cleavage in tumors. a Pearson’s correlation of the changes in IRA mRNA levels in hepatocellular carcinoma vs. normal tissue with those of BACE1 (values are expressed in Log₂). Pearson’s correlation coefficient (r) and p-value are reported. b IR mRNA levels were measured by RT-PCR in tumors from native Huh7 cells and from Huh7 cells overexpressing IRA (Huh7 IRA). c IRsol was measured in the plasma of mice-bearing tumors from native Huh7 cells and from Huh7 cells overexpressing IRA (Huh7 IRA). Data are means ± s.d. **p < 0.01; ***p < 0.001. d Correlation of IRA (circles) and IRB (squares) mRNA levels in tumors from native Huh7 cells with IRsol plasma levels (IRsol values are expressed as fold over the mean). Pearson’s correlation coefficient (r) and p-value are reported (NS, no significant correlation)