Altered amyloid precursor protein processing regulates glucose uptake and oxidation in cultured rodent myotubes
- 1.6k Downloads
Impaired glucose uptake in skeletal muscle is an important contributor to glucose intolerance in type 2 diabetes. The aspartate protease, beta-site APP-cleaving enzyme 1 (BACE1), a critical regulator of amyloid precursor protein (APP) processing, modulates in vivo glucose disposal and insulin sensitivity in mice. Insulin-independent pathways to stimulate glucose uptake and GLUT4 translocation may offer alternative therapeutic avenues for the treatment of diabetes. We therefore addressed whether BACE1 activity, via APP processing, in skeletal muscle modifies glucose uptake and oxidation independently of insulin.
Skeletal muscle cell lines were used to investigate the effects of BACE1 and α-secretase inhibition and BACE1 and APP overexpression on glucose uptake, GLUT4 cell surface translocation, glucose oxidation and cellular respiration.
In the absence of insulin, reduction of BACE1 activity increased glucose uptake and oxidation, GLUT4myc cell surface translocation, and basal rate of oxygen consumption. In contrast, overexpressing BACE1 in C2C12 myotubes decreased glucose uptake, glucose oxidation and oxygen consumption rate. APP overexpression increased and α-secretase inhibition decreased glucose uptake in C2C12 myotubes. The increase in glucose uptake elicited by BACE1 inhibition is dependent on phosphoinositide 3-kinase (PI3K) and mimicked by soluble APPα (sAPPα).
Inhibition of muscle BACE1 activity increases insulin-independent, PI3K-dependent glucose uptake and cell surface translocation of GLUT4. As APP overexpression raises basal glucose uptake, and direct application of sAPPα increases PI3K–protein kinase B signalling and glucose uptake in myotubes, we suggest that α-secretase-dependent shedding of sAPPα regulates insulin-independent glucose uptake in skeletal muscle.
KeywordsAmyloid BACE1 Glucose uptake Glut4 Insulin PI3K Skeletal muscle Type 2 diabetes
Amyloid precursor protein
Beta-site APP-cleaving enzyme 1
Minimal essential media
Oxygen consumption rate
Protein kinase B
A fundamental characteristic of type 2 diabetes is the occurrence of insulin resistance in liver, adipose tissue and skeletal muscle. However, the first detectable defect in individuals predisposed to type 2 diabetes is altered responsiveness of skeletal muscle to insulin [1, 2], the principal site for insulin-stimulated glucose disposal . Glucose transport is the rate-limiting step for glucose metabolism in healthy and diabetic individuals [4, 5], with decreased insulin-stimulated glucose uptake into skeletal muscle being a major contributory factor to insulin resistance in patients with type 2 diabetes [6, 7, 8]. Insulin promotes glucose transport in skeletal muscle by stimulating the translocation of the insulin-responsive glucose transporter, GLUT4, from intracellular storage vesicles to the plasma membrane . In insulin-resistant humans, normal total expression of GLUT4 is reported , but with decreased cell surface levels [11, 12] due to reduced translocation of GLUT4-containing vesicles . Furthermore, muscle-specific deletion of GLUT4 causes glucose intolerance and diabetes in mice  with reduced insulin-stimulated glucose uptake . Conversely, muscle-specific overexpression of GLUT4 improves insulin action and decreases plasma glucose levels in diabetic mice .
Some studies have linked type 2 diabetes to increased risk of Alzheimer’s disease (AD), with AD patients being more prone to impaired glucose metabolism, hyperinsulinaemia and insulin resistance [17, 18]. A key feature of AD development is enhanced proteolytic cleavage of the amyloid precursor protein (APP) by the aspartyl protease, beta-site APP-cleaving enzyme 1 (BACE1), which, with γ-secretase, raises levels of β-amyloid peptides leading to amyloid aggregation and plaque formation . However, cleavage of APP by α-secretases (e.g. a disintegrin and metalloprotease [ADAM] family member) predominates under basal conditions. BACE1 and α-secretase are sheddases, cleaving the extracellular portions of APP and releasing their ectodomains (soluble [s]APPβ and sAPPα, respectively) from the cell surface. These soluble ectodomains may have physiological functions, with sAPPα being neuroprotective and able to modulate cognitive performance and neuronal plasticity .
Reduction or loss of BACE1 by genetic manipulation in mice improves glucose disposal and insulin sensitivity on regular and high-fat diets, and pharmacological inhibition of BACE1 in mouse C2C12 myotubes increases insulin sensitivity . In addition, BACE1 inhibition or overexpression in C2C12 myotubes increases or decreases, respectively, insulin-stimulated glucose uptake. This suggests that skeletal muscle BACE1 activity is important in the regulation of insulin-driven glucose disposal, but the mechanism by which this occurs is unknown. Skeletal muscle glucose uptake has insulin-dependent and insulin-independent components [22, 23]. Furthermore, in skeletal muscle from patients with type 2 diabetes, the insulin resistance associated with diminished glucose uptake is probably due to post-receptor defect(s), with impaired GLUT4 translocation to the plasma membrane [11, 23, 24]. As current therapies aimed at alleviating insulin resistance and improving glucose uptake are insufficient and often incompletely effective, there is an urgent need for information on novel pathways that modulate GLUT4 translocation and glucose uptake which are amenable to pharmacological intervention.
Consequently, we examined whether BACE1 activity and APP processing in skeletal muscle modulates glucose uptake primarily through altered insulin sensitivity or is capable of engaging an insulin-independent pathway.
L6-GLUT4myc rat myoblasts (kindly provided by Amira Klip, Toronto, Canada) were maintained in α-minimum essential media (MEM) with 10% FBS. C2C12 and L6-GLUT4myc myoblasts were differentiated to myotubes as described previously . C2C12 cells were transfected with 12 μg DNA of empty vector (EV; pcDNA3.1) or pcDNA3.1 containing full-length BACE1, BACE1 active site mutant (mBACE1 ) or APP using Lipofectamine 2000 (Invitrogen Life Technologies, Paisley, UK). Cells were selected with 1 mg/ml G418 (Sigma–Aldrich, Gillingham, UK) and differentiated. A minimum of two independently generated stable cell lines with concurrently produced EV controls was used. After 4 days of differentiation, cells were treated with TAPI-1 (Invitrogen), Merck-3 (β-secretase inhibitor IV [M-3]), BACE1 inhibitor II, batimastat, palmitate or d-erythro-sphingosine, N-acetyl-C2-ceramide (all Calbiochem, Nottingham, UK), or appropriate vehicle, overnight (∼20 h).
Full-length myc-his human BACE1 in pcDNA3.1 was obtained from GlaxoSmithKline (Harlow, UK). mBACE1 (a gift from Professor Wolfe [Brigham and Women’s Hospital, Boston, MA, USA]) was subcloned into pcDNA3.1. Full-length human APP (Genbank AAH65529.1) was amplified with primers APP Fwd: 5'-AAAGCTAGCATGCTGCCCGGTTTG-3'; APP Rev: 5'-TTTAAGCTTCTAGTTCTGCATCTGCTCAAA-3' and cloned into pcDNA3.1.
Immunoblotting and gene expression
Protein isolation and immunoblotting procedures were as described previously . For quantification of sAPP fragments, differentiated cells were treated in 10 ml Optimem (Gibco Life Technologies, Paisley, UK), and media were concentrated (30 kDa Amicon Ultra 15 ml filter) by centrifugation (4,000 g) and subjected to SDS-PAGE with amounts presented relative to total protein. Primary antibodies used were: anti-sAPPβ (Covance, Alnwick, UK; 1:1,000), anti-sAPPα (IBL International, Hamburg, Germany; 1:50), anti-APP (Ab54, GlaxoSmithKline; 1:4,000), anti-phospho-PKB (protein kinase B; Ser473), anti-PKB and anti-HKII (hexokinase II) (Cell Signaling, Hitchin, UK; 1:1,000), anti-GLUT1 (Millipore, Nottingham, UK; 1:1,000) and anti-GLUT4 (Abcam, Cambridge, UK; 1:1,000). Glut1 (also known as Slc2a1), Glut4 (also known as Slc2a4) and HkII (also known as Hk2) mRNA was determined by TaqMan RT-PCR (Applied Biosystems, Paisley, UK; Prism Model 7700) using commercial primers and probe sets.
Cell surface GLUT4myc detection
For details, see Wang et al . Briefly, L6-GLUT4myc myoblasts or myotubes were serum-starved (4 h), treated with M-3 (250 nmol/l), stimulated with insulin (20 or 100 nmol/l) or vehicle at 37°C for 30 min, and fixed in 3% paraformaldehyde. Anti-c-myc (A14; Santa Cruz, Heidelberg, Germany; 1:100) primary antibody was applied (1 h), and peroxidase-conjugated rabbit anti-mouse secondary antibody (1:1,000) in 3% goat serum was added, followed by 1 ml o-phenylenediamine dihydrochloride (OPD reagent; Sigma), and incubated (in the dark) for 30 min. The reaction was stopped with 250 μl 3 mol/l HCl, and absorbance of the supernatant fraction was measured.
C2C12 myotubes, stably expressing BACE1, APP or mBACE1, or EV controls were exposed to M-3 (250 nmol/l), BACE1 inhibitor II (0.7 μmol/l), palmitate (750 μmol/l), ceramide (50 μmol/l) or batimastat (5 μmol/l). Cells were serum-starved (2 h) and exposed to insulin (100 nmol/l), sAPPα (0.3, 3 or 10 nmol/l), sAPPβ (0.3, 3 or 10 nmol/l) or vehicle control. For phosphoinositide 3-kinase (PI3K)-dependence, myotubes were pretreated overnight with M-3 or vehicle, then with wortmannin (100 nmol/l) for 1 h before insulin. Myotubes were incubated (12 min) with 10 μmol/l 2-deoxy-d-[3H]glucose (2DG; 24.4 kBq/ml; PerkinElmer, Cambridge, UK) at 20°C. Non-specific uptake was determined using 10 μmol/l cytochalasin B (Sigma–Aldrich). After lysis, cell-associated radioactivity was measured (Beckman, High Wycombe, UK; LS 6000IC scintillation counter), and protein was quantified using the Bradford reagent.
C2C12 myotubes pre-exposed to M-3 (250 nmol/l), BACE1 inhibitor II (0.7 μmol/l) or batimastat (5 μmol/l) were incubated with 5 mmol/l glucose and 74 kBq/ml d-[U-14C]glucose (PerkinElmer) for 2 h at 37°C (glucose oxidation assay) or 5 mmol/l glucose and 750 μmol/l palmitate (conjugated to BSA)/[14C]palmitate (74 kBq/ml) for 3 h (palmitate oxidation assay). Medium was transferred to 15 ml centrifuge tubes, 14CO2 released using 60% (v/v) perchloric acid was trapped by a Whatman (GF/B) filter paper disc presoaked with 1 mol/l KOH, and radioactivity was quantified by liquid-scintillation counting.
Immunostaining and imaging
C2C12 myotubes were fixed in 4% paraformaldehyde, permeabilised in PBST (PBS + 0.1% Triton X-100), blocked in 10% donkey serum (Sigma–Aldrich) and then incubated with primary antibodies: anti-APP (Ab54, 1:100) or anti-BACE1 (Sigma–Aldrich, 1:250). Secondary antibodies were Cy3 (Jackson ImmunoResearch, Newmarket, UK; 1:250) and Alexa Fluor 488 (Invitrogen, 1:250). Images were acquired with a confocal laser-scanning microscope (Leica, Milton Keynes, UK; TCS SP5 II).
Myoblasts were seeded in XF 24-well culture microplates (Seahorse Bioscience, Copenhagen, Denmark), differentiated for 4 days and changed to Opti-MEM-reduced serum medium incubated at 37°C (24 h). Cells were washed in Krebs–Henseleit buffer (in mmol/l: 111 NaCl, 4.7 KCl, 2 MgSO4.7H2O and 1.2 Na2HPO4) containing 5 mmol/l glucose and 0.5 mmol/l l-carnitine, and baseline recordings were taken to ensure a steady respiratory rate before injection of palmitate (750 μmol/l).
Comparisons between groups were performed using an unpaired two-tailed Student’s t test, one-sample Student’s t test or ANOVA with repeated measures and Tukey’s multiple comparison test, as appropriate, using GraphPad (Prism 5) software (GraphPad Software, La Jolla, CA, USA). p values ≤ 0.05 were considered significant.
Glucose uptake and GLUT4 translocation in myotubes are modulated by BACE1 activity in an insulin-independent manner
Palmitate and ceramide increase BACE1 and inhibit glucose uptake
BACE1 activity alters substrate oxidation of C2C12 myotubes
APP cleavage may mediate changes in glucose uptake by BACE1 activity
M-3-dependent glucose uptake is PI3K-dependent and mimicked by sAPPα
Inhibition of BACE1 in rodent myotubes increases glucose uptake and GLUT4 translocation independently of, but additive to, insulin-stimulated glucose uptake. Accordingly, reduced BACE1 activity increases insulin-stimulated glucose uptake in agreement with our previous study . Impairment of insulin-stimulated glucose uptake by skeletal muscle is recognised as an early defect in the pathogenesis of type 2 diabetes, although the ability of alternative stimuli such as exercise/contraction to increase glucose uptake is unaffected . However, non-insulin-dependent (basal) glucose uptake is also impaired in patients with type 2 diabetes. In the fasted state, when plasma insulin levels are low and skeletal muscle glucose uptake is insulin-independent, muscle glucose uptake is decreased in insulin-resistant individuals [34, 35]. Consequently, a novel pharmacological approach to increase glucose disposal by targeting basal skeletal muscle glucose uptake may have therapeutic utility.
Skeletal muscle glucose uptake occurs predominantly via the insulin-sensitive transporter, GLUT4, although GLUT1 and GLUT12 are also expressed with GLUT1 and thought to contribute to basal glucose uptake . BACE1 inhibition did not alter glucose transporter mRNA or HkII mRNA or protein levels, although a small increase in GLUT1 and GLUT4 protein levels was observed. Nevertheless, the major effect of BACE1 inhibition was increased cell surface GLUT4myc in the absence of insulin. The increased basal and insulin-stimulated glucose uptake and GLUT4 translocation elicited by BACE1 inhibition was prevented by wortmannin, indicating a PI3K-regulated mechanism. Consequently, it is likely that the increase in basal glucose uptake observed following BACE1 inhibition is predominantly due to enhanced translocation of GLUT4, through activation of the canonical class 1A PI3K pathway .
Although we have not completely delineated the mechanism by which BACE1 modulates glucose uptake, our results indicate a key role for APP-cleavage products. APP membrane processing occurs predominantly by α-secretases (the ‘non-amyloidogenic’ pathway), most likely ADAM10 . This ectodomain-shedding process liberates a soluble truncated form of APP, sAPPα. In contrast, BACE1 (the ‘amyloidogenic’ pathway) cleaves APP at a different site and releases a shorter soluble APP isoform, sAPPβ. Thus α- and β-secretases compete for APP cleavage, with the α-secretase pathway prevailing. Nevertheless, events such as chronic stress raise BACE1 activity and increase sAPPβ levels with a compensatory decline in sAPPα. Our results strongly indicate a role for the α-secretase pathway in modulating glucose uptake because: (1) α-secretase inhibition reduces glucose uptake in conjunction with diminished sAPPα in the medium; (2) APP overexpression (thus increased sAPPα) enhances basal and insulin-dependent glucose uptake; and (3) direct application of sAPPα increases basal and insulin-dependent glucose uptake and PI3K signalling, whereas sAPPβ does not affect either process.
Therefore we suggest that constitutive α-secretase activity maintains basal glucose uptake in muscle cells, whereas increased BACE1 activity inhibits this process by diverting APP down the amyloidogenic pathway. This reduces sAPPα in the medium and diminishes PI3K-driven GLUT4 translocation. Indeed, sAPPα exhibits neuroprotective properties and increases PI3K–PKB signalling and glucose uptake in neurons, effects not replicated by sAPPβ [38, 39]. These actions have been attributed to the C-terminal part of sAPPα, which differs from sAPPβ by the presence of an additional 16 amino acids. Thus it is likely that this region is also responsible for the increased glucose uptake and PI3K signalling in skeletal muscle. The receptor by which sAPPα may mediate these effects is at present unclear. However, sAPPα has structural similarities to cysteine-rich growth factors  and its neuroprotective function has been linked with activation of IGF-1 receptor/insulin receptor PI3K–PKB signalling in neurons .
The increased fat supply associated with obesity is a primary drive for the induction of skeletal muscle insulin resistance. Fatty acids are an important fuel source for muscle, and excess long-chain fatty acids, particularly saturated ones, in skeletal muscle raise levels of the lipid intermediates, diacylglycerol and ceramide, which are strongly implicated in the pathogenesis of insulin resistance [41, 42]. The skeletal muscle accumulation of diacylglycerol and ceramide is associated with impaired insulin signalling predominantly via the PI3K–PKB pathway [42, 43]. Exposure of C2C12 myotubes to palmitate or ceramide depressed insulin-dependent and -independent glucose uptake in the presence of BACE1 inhibitor. Thus the molecular mechanism underlying BACE1 action, which requires PI3K activity, is also sensitive to ceramide. Indeed, skeletal muscle BACE1 activity is increased by high-fat diet , and the finding that ceramide, concomitant with high BACE1 activity, strongly suppresses basal and insulin-dependent glucose uptake fits with the idea that lowering of sAPPα reduces non-insulin-dependent glucose uptake in muscle. Thus the balance between α- and β-secretase activities may be an important checkpoint in the control of skeletal muscle metabolism. Accordingly, pharmacological reduction of BACE1, or increased α-secretase activity, may represent a novel and reasonable therapeutic target to improve glucose uptake in tissues exposed to excess lipid.
BACE1 inhibition is also associated with increased glucose oxidation in C2C12 myotubes, with an increased proportion (over palmitate) of glucose oxidised for the same amount of oxygen consumed. Such an action in vivo would raise postprandial glucose uptake and oxidation and increase overall glucose disposal. Conversely, overexpression of BACE1 inhibited glucose oxidation and oxygen consumption with no reduction in palmitate oxidation. Furthermore, inhibition of α-secretase with batimastat reduced the increase in OCR elicited by palmitate in the presence of glucose, indicating diminished metabolic flexibility. Impairment of resting and insulin-stimulated mitochondrial oxidative phosphorylation has been reported in skeletal muscle of patients with type 2 diabetes [44, 45], and reduced glucose oxidation, with unchanged lipid oxidation, has been found to be associated with loss of metabolic flexibility in obese individuals . At present, the molecular mechanism whereby BACE1 reduces glucose oxidation is unknown, with sAPPβ an unlikely mediator from our results. A plausible candidate is one or more of the β-amyloid peptides resulting from the sequential cleavage of APP by BACE1 and γ-secretase, as these aggravate the diabetic phenotype of rodents [47, 48].
Taken together, our results indicate that preventing skeletal muscle BACE1 activity from increasing excessively in times of chronic nutrient stress may have important implications, and indeed offer new therapeutic avenues, for the maintenance of glucose uptake and oxidation and the preservation of metabolic flexibility.
We gratefully acknowledge assistance from Dundee Cell Products with cloning.
This study was funded by grants from Diabetes UK (0003681), the Medical Research Council (K003291/1), Alzheimer’s Research UK (ART-PhD2010-2) and the Wellcome Trust (086989).
Access to research materials
The research materials supporting this publication can be accessed by contacting MLJA.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
DLH, JAF, GM, PJM, DB and SMJ contributed to the acquisition of data. DLH, JAF, GM, PJM and MLJA made substantial contributions to the analysis and interpretation of data. DLH, JAF, GM, PJM and MLJA contributed to the conception and design of the study. DLH, JAF, GM, PJM, DB, SMJ and MLJA wrote or critically revised the manuscript. All authors approved the final version. MLJA is the guarantor of this work.
- 38.Jimenez S, Torres M, Vizuete M et al (2011) Age-dependent accumulation of soluble amyloid β (Aβ) oligomers reverses the neuroprotective effect of soluble amyloid precursor protein-α (sAPPα) by modulating phosphatidylinositol 3-kinase (PI3K)/Akt-GSK-3β pathway in Alzheimer mouse model. J Biol Chem 286:18414–18425PubMedCentralPubMedCrossRefGoogle Scholar
- 39.Mattson MP, Guo ZH, Geiger JD (1999) Secreted form of amyloid precursor protein enhances basal glucose and glutamate transport and protects against oxidative impairment of glucose and glutamate transport in synaptosomes by a cyclic GMP-mediated mechanism. J Neurochem 73:532–537PubMedCrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.