Unexpected role of the copper transporter ATP7A in PDGF-induced vascular smooth muscle cell migration

Abstract

Rationale: Copper, an essential nutrient, has been implicated in vascular remodeling and atherosclerosis with unknown mechanism. Bioavailability of intracellular copper is regulated not only by the copper importer CTR1 (copper transporter 1) but also by the copper exporter ATP7A (Menkes ATPase), whose function is achieved through copper-dependent translocation from trans-Golgi network (TGN). Platelet-derived growth factor (PDGF) promotes vascular smooth muscle cell (VSMC) migration, a key component of neointimal formation.

Objective: To determine the role of copper transporter ATP7A in PDGF-induced VSMC migration.

Methods and results: Depletion of ATP7A inhibited VSMC migration in response to PDGF or wound scratch in a CTR1/copper-dependent manner. PDGF stimulation promoted ATP7A translocation from the TGN to lipid rafts, which localized at the leading edge, where it colocalized with PDGF receptor and Rac1, in migrating VSMCs. Mechanistically, ATP7A small interfering RNA or CTR small interfering RNA prevented PDGF-induced Rac1 translocation to the leading edge, thereby inhibiting lamellipodia formation. In addition, ATP7A depletion prevented a PDGF-induced decrease in copper level and secretory copper enzyme precursor prolysyl oxidase (Pro-LOX) in lipid raft fraction, as well as PDGF-induced increase in LOX activity. In vivo, ATP7A expression was markedly increased and copper accumulation was observed by synchrotron-based x-ray fluorescence microscopy at neointimal VSMCs in wire injury model.

Conclusions: These findings suggest that ATP7A plays an important role in copper-dependent PDGF-stimulated VSMC migration via recruiting Rac1 to lipid rafts at the leading edge, as well as regulating LOX activity. This may contribute to neointimal formation after vascular injury. Our findings provide insight into ATP7A as a novel therapeutic target for vascular remodeling and atherosclerosis.

Figure 1. ATP7A is involved in VSMC migration stimulated by PDGF or wound scratch in a copper-dependent manner

A, RT-PCR analysis of ATP7A and CTR1 mRNA expression in various VSMCs (rat, mouse, and human aortic smooth muscle cells (RASM HASM and MASM)). B, Western blot analysis of ATP7A protein expression in various VSMCs. C and D, RASMs were transfected with ATP7A, CTR1, or control siRNA for 48 hours, or treated with the copper chelator BCS (cell impermeable) (200 μM for 72 h) or the copper chelator TTM (cell permeable) (10 nM for 24 h). Cell migration was assessed by the modified Boyden chamber assay after stimulation with or without 50 ng/ml of PDGF for 8 hours. E, Wound scratch assay was performed in confluent monolayers of RASMs transfected with siRNA or treated with copper chelator in the presence of PDGF (50 ng/ml), as described in the online supplement. Images were captured immediately after rinsing at 0 h and at 24 h after the wounding in the cells. Bar graph represents averaged data, expressed as cell number per field. *p<0.05 vs. control siRNA-treated (C and E), or untreated (D and E) cells. Values are the mean ± S.D. for 3 independent experiments.

Figure 2. ATP7A is translocated to the leading edge in VSMCs stimulated by wound scratch or PDGF in a copper-dependent manner

A and B, Confluent monolayer of RASMs before (Top) and after wound scratch in the presence of 50 ng/ml PDGF for 18 hours (Bottom) were stained with anti-ATP7A (green) and Alexa Fluor 568-phalloidin (red) antibodies. Small white arrowheads point to the leading edge and large arrows point to direction of migration. C and D, Growth-arrested RASMs were stimulated with or without 50 ng/ml PDGF for 5 min. Cells were co-stained for ATP7A and phalloidin. In some experiments, RASMs transfected with CTR1, or control siRNA, or treated with copper chelators BCS or TTM were stimulated by wound scratch (B) or PDGF (D). Results for A–D are representative of 3 independent replicates of immunofluorescence images.

Figure 3. ATP7A is involved in PDGF-stimulated lamellipodia formation and Rac1 translocation in a CTR1-dependent manner in VSMCs

A and B, RASMs transfected with control siRNA or CTR1 siRNA or ATP7A siRNA were stimulated with wound scratch (A) or 50 ng/ml PDGF (B) as described in Figure 2, and cells were stained for phalloidin to visualize lamellipodia formation. Cells with lamellipodia formation were expressed as % of cell number of wound edge (A) or total cell number (B) (mean ± S.D, n=3). In B, cells in boxes are magnified in insets. Small white arrowheads point to the leading edge and large arrows point to direction of migration. *p<0.05 vs. control siRNA-treated cells. C and D, RASMs stimulated with or without PDGF as described above, were co-stained with anti-ATP7A antibody (red) and anti-Rac1 antibody (green). In D, RASMs were transfected with control siRNA or CTR1 siRNA or ATP7A siRNA as described. All fluorescence images were taken at 5 different fields/well, and the cell images are representative of more than 3 different experiments.

Figure 4. ATP7A colocalizes with PDGFR at the leading edge in a CTR1-dependent manner in PDGF-stimulated VSMCs

A–C, Growth-arrested RASMs were stimulated with 50 ng/mL of PDGF for 5 min. All fluorescence images were taken at 5 different fields/well, and the cell images are representative of more than 3 different experiments. A, Effect of PDGF on subcellular localization of ATP7A and PDGFR in VSMCs. RASMs were stained with anti-ATP7A antibody (green) and anti-PDGFR antibody (red). B, Effect of ATP7A or CTR1 siRNA on subcellular localization of ATP7A and PDGFR in PDGF-treated cells. RASMs transfected with CTR1, ATP7A, or control siRNA were double-stained with anti-ATP7A antibody (green) and anti-PDGFR antibody (red). C, PDGF stimulation promoted PDGFR association with ATP7A in a CTR1-dependent manner. RASMs were transfected with control siRNA or CTR1 siRNA. Growth-arrested confluent monolayer of RASMs was stimulated with 50 ng/mL of PDGF for indicated times (min). Lysates were IP with anti-PDGFR antibody, followed by IB with ATP7A, phospho-PDGFR and PDGFR antibody.

Figure 5. ATP7A is translocated to the lipid rafts localized at the leading edge, in a CTR1-dependent manner, in migrating VSMCs stimulated with PDGF

A and B, Sucrose gradient centrifugation was performed for RASMs in basal state (A) or RASMs transfected with CTR1 or control siRNA, followed by stimulated with or without PDGF for 5 min (B). Fractions from the top (fraction 1) to the bottom (fraction 13) were immunoblotted with antibodies as indicated (A). In B, equal amounts of caveolin-enriched lipid rafts fractions (caveolae/lipid rafts, fraction 4–5) were immunoblotted with antibodies as indicated. Right panel shows averaged data for ATP7A protein, expressed as % of unstimulated, control siRNA-treated VSMCs (means ± S.D, n=3). *p<0.05 vs. control siRNA-treated cells. C and D, Growth-arrested RASMs were treated with (C) or without (D) 10 mM of methyl-β-cyclodextrin for 2 h, and stimulated with 50 ng/ml of PDGF for 5 min. Cells were co-stained with anti-ATP7A antibody (green) and Alexa 555-cholera toxin subunit B (CTxB)(red). All fluorescence images were taken at 5 different fields/well, and are representative of more than 3 different experiments.

Figure 6. PDGF stimulation reduces copper content and Pro-LOX in lipid rafts fractions in VSMCs

A–C, Copper contents were measured by inductively coupled plasma mass spectrometry (ICP-MS) in whole cells (A), caveolae/lipid rafts or non-caveolae/lipid rafts (B and C) in RASMs with or without PDGF stimulation for 5 min. Equal amounts of proteins in caveolae/lipid rafts (fraction 4–5) or non-caveolae/lipid rafts (fraction 9–13) were obtained by sucrose gradient fractionation as described in Figure 5. (Mean ± S.D. n=3) *p<0.05 vs. non-lipid rafts or unstimulated cells. D, Identification of Pro-LOX, but not LOX, in caveolae/lipid rafts in VSMCs. RASMs were fractionated by sucrose gradient centrifugation, followed by immunoblotted with anti-LOX-PP which detects both Pro-Lox and LOX-PP, anti-LOX, or anti-caveolin-1 antibodies. E, Effect of PDGF treatment on Pro-LOX level in caveolae/lipid rafts fractions in VSMCs. Equal amounts of proteins in caveolae/lipid rafts (fraction 4–5) were immunoblotted with anti-LOX-PP, LOX, or caveolin-1 antibodies in RASMs with or without 50 ng/ml of PDGF for 5 min.

Figure 7. ATP7A is involved in neointimal formation in response to vascular injury in vivo

A and B, ATP7A is highly expressed in neointimal VSMCs of wire-injured carotid arteries of ApoE deficient atherosclerotic mice. Immunohistochemial (A) or immunofluorescence (B) analysis for uninjured (control) or injured (4 weeks after) carotid artery stained with anti-ATP7A antibody (A); or co-stained with anti-ATP7A (green) and α-smooth muscle actin (red) antibodies (B). C, XFM scans of the neointima of wire-injured carotid arteries. Areas of neointimal lesions were identified by H & E staining (left). XFM scans (1–2 sec per pixel) were performed in paraffin-embedded tissue (right). Map of Cu shows areas of the lowest to the highest content scaled to a rainbow color (bottom). Copper accumulation in neointimal lesions is shown in white arrow. The minimal and maximal Cu content displayed in micrograms per square centimeter is shown in the image.

Figure 8. Proposed model for role of ATP7A in PDGF-induced VSMC migration

PDGF promotes ATP7A translocation from the transGolgi network (TGN) to the lipid rafts which localize at the leading edge, thereby stimulating lamellipodia formation via recruiting Rac1, which in turn promotes directional VSMC migration involved in neointimal formation. This is associated with a decrease in cellular copper level and secretory copper enzyme pro-lysyl oxidase (Pro-LOX) at the lipid rafts, which is extracellularly processed and activated by proteolysis to a mature lysyl oxidase (LOX) and a propeptide (LOX-PP), which may promote extracellular matrix (ECM) remodeling and VSMC migration. Secreted copper may also contribute to PDGF induced cell migration.