Apolipoprotein A-I enhances insulin-dependent and insulin-independent glucose uptake by skeletal muscle

Therapeutic interventions that increase plasma high density lipoprotein (HDL) and apolipoprotein (apo) A-I levels have been reported to reduce plasma glucose levels and attenuate insulin resistance. The present study asks if this is a direct effect of increased glucose uptake by skeletal muscle. Incubation of primary human skeletal muscle cells (HSKMCs) with apoA-I increased insulin-dependent and insulin–independent glucose uptake in a time- and concentration-dependent manner. The increased glucose uptake was accompanied by enhanced phosphorylation of the insulin receptor (IR), insulin receptor substrate-1 (IRS-1), the serine/threonine kinase Akt and Akt substrate of 160 kDa (AS160). Cell surface levels of the glucose transporter type 4, GLUT4, were also increased. The apoA-I-mediated increase in glucose uptake by HSKMCs was dependent on phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K)/Akt, the ATP binding cassette transporter A1 (ABCA1) and scavenger receptor class B type I (SR-B1). Taken together, these results establish that apoA-I increases glucose disposal in skeletal muscle by activating the IR/IRS-1/PI3K/Akt/AS160 signal transduction pathway. The findings suggest that therapeutic agents that increase apoA-I levels may improve glycemic control in people with type 2 diabetes.

cell surface, and increases glucose uptake 17 . The present study establishes that apoA-I enhances this signalling pathway and the translocation of GLUT4 to the cell surface.

ApoA-I increases insulin-stimulated glucose uptake into HSKMCs in an IRS
To determine if activation of PI3K/Akt was required for the apoA-I-mediated uptake of glucose into HSKMCs, incubations were also carried out with the PI3K/Akt inhibitor, wortmannin. Akt (Fig. 3C), and AS160 phosphorylation ( Fig. 3D) as well as total AS160 levels ( Fig. 3E) were all decreased when HSKMCs were incubated with wortmannin ( Fig. 3C-E, closed bars), irrespective of whether or not apoA-I and insulin were present. Incubation of HSKMCs with insulin or apoA-I in the absence of wortmannin increased glucose uptake from 1.0 nmol/ mg/h to 1.21 ± 0.01 nmol/mg/h and 1.25 ± 0.08 nmol/mg/h, respectively (Fig. 3F, open bars) (p < 0.001 for both). Incubation with apoA-I plus insulin in the absence of wortmannin further increased glucose uptake to 1.45 ± 0.07 nmol/mg/h (p < 0.001 vs. control).
Glucose uptake was decreased in the wortmannin-treated HSKMCs, irrespective of whether or not insulin was present in the incubation (Fig. 3F, closed bars). Incubation of wortmannin-treated HSKMCs with ApoA-I increases GLUT4 translocation to the plasma membrane in HSKMCs. The normalised plasma membrane GLUT4 level was 100 ± 18.7 mean fluorescence intensity units (MFI) in untreated HSKMCs (Fig. 4A, open bar). Inclusion of insulin or apoA-I in the incubation increased the plasma membrane GLUT4 level to 264.7 ± 8.547 (p < 0.0001 vs. control) and 181.8 ± 33 MFI, respectively (p < 0.01 vs. control) (Fig. 4A, closed bars). Incubation with insulin plus apoA-I further increased the HSKMC plasma membrane GLUT4 level to 345.1 ± 10.92 MFI (p < 0.001 vs. control; p < 0.01 vs. cells incubated with insulin alone). Incubation of HSKMCs with insulin or apoA-I did not affect total cell GLUT4 levels (Fig. 4B), but increased GLUT4 mRNA levels by 25.2 ± 7.0% and 32.3 ± 5.3%, respectively (Fig. 4C, p < 0.05 for both). Glycosylation of GLUT4 can cause heterogeneous bands in western blots. However, GLUT4 western blots can also be well focussed, as is the case in Fig. 4B. This variation is dependent on the amount of protein loaded onto the gel, whether or not a stacking gel was used and the duration of electrophoresis. High sample loading, use of a stacking gel and a low running voltage all have the capacity to produce poorly focused GLUT4 bands in western blots [18][19][20][21] . Incubation of HSKMCs with apoA-I plus insulin increased cellular GLUT4 mRNA levels by 45.3 ± 21% (p < 0.01 vs. control). Incubation of HSKMCs with apoA-I did not alter IR, IRS-1, PI3K and Akt mRNA levels (Supplemental Fig. II A-D).
To determine if the effects of ABCA1 and SR-B1 knockdown on glucose uptake could be explained by an increase in HSKMC cholesterol content, cholesterol levels in the transfected cells were quantified by HPLC. The cholesterol content of the scrambled siRNA-transfected cells was 73.6 ± 4.4 nmol/mg protein (Fig. 5E, open bar), compared to 69.9 ± 2.5 nmol/mg protein and 68.6 ± 4.5 nmol/mg protein for the ABCA1 siRNA-and SR-B1 siRNA-transfected cells, respectively (Fig. 5E, closed bars). This indicates that ABCA1 and SR-B1 regulate apoA-I-mediated glucose uptake by HSKMCs by a mechanism that is independent of their role in maintaining cellular cholesterol homeostasis.

Discussion
The present study establishes that apoA-I, the predominant HDL apolipoprotein, significantly increases insulin-dependent and insulin-independent glucose disposal in skeletal muscle by a mechanism that involves phosphorylation of IRβ and IRS-1, activation of the PI3K/Akt/AS160 signal transduction pathway and increased translocation of GLUT4 to the cell surface. The ability of insulin to activate the IRβ/IRS-1/PI3K/Akt/AS160 pathway and regulate GLUT4 translocation to the cell membrane is well known 12,22 , but the capacity of apoA-I to enhance these events directly has not been reported previously. The outcome of the present study also provides a One of the most interesting observations to emerge from the present study is that apoA-I can act alone, as well as synergistically with insulin, to increase IR, IRS-1, Akt and AS160 phosphorylation. This is consistent with previous reports showing that apoA-I increases glucose uptake into skeletal muscle cells in the absence of insulin, and that apoA-I and insulin have a synergistic effect on AMPK-dependent glucose uptake in C2C12 myotubes 6,8 . It is also consistent with our previous report showing that apoA-I increases insulin-dependent glucose uptake in skeletal muscle from diabetic db/db mice by increasing insulin sensitivity in the muscle 10 . Previous studies have shown that apoA-I knockout mice have reduced glucose tolerance 6 , whereas overexpression of apoA-I in transgenic mice increases insulin sensitivity 7 . The results in the present study indicate that apoA-I may mediate these effects directly by increasing glucose disposal in skeletal muscle, although the possible involvement of an apoA-I-mediated improvement in pancreatic beta cell function, as we have reported previously 5,24 , cannot be excluded.
We also found that knockdown of IRS-1 and inhibition of Akt and AS160 phosphorylation with wortmannin attenuated apoA-I-mediated insulin-dependent glucose uptake, but did not affect the ability of apoA-I to promote insulin-independent glucose uptake in HSKMCs (Fig. 3). This may reflect the differential regulation of insulinand AMPK-induced glucose transport in skeletal muscle cells, with reports showing that AS160 phosphorylation is mediated both directly and indirectly by insulin and AMPK 22,25,26 . It does not, however, explain the synergistic phosphorylation of IRS-1 and Akt that occurred when HSKMCs were incubated with apoA-I plus insulin. Hence, in addition to the AMPK pathway, this observation raises the possibility that apoA-I may enhance glucose uptake by additional, yet-to-be identified insulin-independent pathways.
The current study also establishes that apoA-I increases GLUT4 translocation to the cell surface in the presence and absence of insulin. This is consistent with a previous report showing that HDL can increase GLUT4 translocation to the adipocyte surface 27 . As activation of IRS-1/PI3K/Akt/AS160 insulin signal transduction has been reported to promote GLUT4 translocation to the cell surface in skeletal muscle 28 , it is likely that activation of this pathway by apoA-I was responsible for the increase in cell membrane GLUT4 content in Fig. 4. As apoA-I also activates the AMPK pathway in skeletal muscle 6,8 , this pathway may have further contributed to the observed increase in GLUT4 translocation 29 .
The absence of change in total GLUT4 protein levels in HKSMCs that were incubated with apoA-I was unexpected. This is possibly because GLUT4 is continuously recycling between the cell surface and intracellular compartments 30 , with the increase in cell surface GLUT4 levels reflecting either increased GLUT4 translocation from intracellular compartments to the cell membrane, decreased recycling of cell surface GLUT4 back to intracellular compartments, or an increase in GLUT4 synthesis. It is also important to note that although insulin primarily promotes GLUT4 translocation to the cell surface, GLUT4 continues to recycle and undergo degradation in insulin-stimulated cells 31 . This may explain why GLUT4 mRNA levels were increased by apoA-I, but GLUT4 protein expression in the cells was unaltered. The present results also establish that the ability of apoA-I to increase insulin-dependent and insulinindependent glucose uptake by skeletal muscle is regulated by ABCA1 and SR-B1. This is in agreement with a previous report showing that ABCA1 is involved in apoA-I-mediated, insulin-independent glucose uptake in skeletal muscle cells 8 . The present study extends this finding by establishing that the ability of apoA-I to increase insulin-dependent and insulin-independent glucose uptake by skeletal muscle also requires SR-B1. This is of particular interest as SR-B1 is known to activate several signal transduction pathways 3 . The result showing that ABCA1 and SR-B1 knockdown did not affect cellular cholesterol content indicates that regulation of apoA-I-mediated glucose uptake by ABCA1 and SR-B1 is independent of apoA-I acting as an acceptor of cellular cholesterol.
In conclusion, this study provides compelling evidence that apoA-I increases glucose disposal in skeletal muscle, thus supporting a role for HDL in reducing insulin resistance and improving glycaemic control in people with type 2 diabetes. This study further strengthens recent observations showing that HDL have anti-diabetic properties. It also suggests that HDL-raising interventions may be beneficial in the management of people with pre-diabetes as well as individuals with type 2 diabetes.

Methods
Isolation of apoA-I. HDLs were isolated from pooled samples of autologously donated human plasma (Healthscope Pathology, Adelaide, South Australia, AUS) by sequential ultracentrifugation in the 1.063< d <1.21 g/mL density range and delipidated as described 32 . ApoA-I was isolated from the resulting apoHDL by chromatography on a Q-Sepharose Fast-Flow column 33  . The cells were then rinsed with ice-cold Dulbecco's PBS, lysed with SDS (0.05%, w/v) and subjected to liquid scintillation counting. Non-carrier-mediated glucose uptake was determined by incubating the cells for 15 min with cytochalasin B (10 µmol/L, #C2743, Sigma) prior to [ 3 H]2-deoxy-glucose uptake 34 . The results were corrected for cell protein and non-carrier-mediated glucose uptake. The values for the control cells that were incubated in the absence of insulin and apoA-I were normalised to 1 nmol/mg protein/h. All experiments were carried out in triplicate. The results are representative of three or more independent experiments.
For incubation with wortmannin, fully differentiated HSKMCs were incubated for 1 h in serum-free MEM-α with wortmannin (0.1 µmol/L final concentration, #W3144, Sigma), then incubated for 16 h with wortmannin in the presence or absence of apoA-I (1 mg/mL final concentration). The apoA-I was removed, and the cells were incubated for 1 h with wortmannin with or without insulin (0.1 µmol/L final concentration). [ 3 H]2-deoxy-glucose uptake was determined as described above.

IRS-1, ABCA1 and SR-B1 knockdown.
Near-confluent HSKMCs (1.7 × 10 4 cells/mL) were plated in 12-well collagen-coated plates. Skeletal muscle differentiation medium was added to the cells the following day. After 3 days (~48 h prior to full differentiation), the cells were transfected for 16 h with scrambled siRNA, IRS-1 siRNA, ABCA1 siRNA or SR-B1 siRNA (0.1 µmol/mL final concentration for all; ON-TARGET SMART pool siRNA, GE Healthcare, San Francisco, CA, USA) using DharmaFECT 4 siRNA transfection reagent (#T-2004-03, GE Healthcare). GLUT4 translocation. GLUT4 translocation was measured as described previously 35 . Briefly, HSKMCs were incubated at 37 °C for 16 h with or without apoA-I (1 mg/mL). Primary anti-GLUT4 (#sc-1608, Santa Cruz Biotechnology, Dallas, TX, USA) and secondary DyLight488 donkey anti-goat (#ab96931, Abcam, Cambridge, MA, UK) antibodies were combined and maintained at room temperature for 10 min. The cells were washed with PBS and incubated for 30 min with the antibody complex in the presence and absence of insulin (0.1 μmol/L final concentration). After washing and fixing the cells with 2.5% (v/v) neutral buffered formalin for 20 min, GLUT4 translocation to the cell surface was analyzed using a FACSverse flow cytometer (BD Biosciences, San Jose, CA, USA).
Western blotting. Fully differentiated HSKMCs were incubated at 37 °C for 16 h in serum-free MEM-α with or without apoA-I (1 mg/mL final concentration). The apoA-I was removed, and the cells were incubated