Photomicrographs of polymer control-stented arteries (A and C) and everolimus-stented arteries (B and D) stained for RAM-11 (A and B; brown = macrophages [M]) and -smooth muscle cell (SMC) actin (C and D; brown = SMC). Stent struts (S) were separated from the media (M) by plaque tissue (P). Polymer control-stented plaques contained abundant M, whereas everolimus-stented plaques showed a marked reduction of the macrophage content, with preservation of the SMC content. Scale bar = 20 µm. (E) Quantification of M and SMC around the stent struts. The strut circumference surrounded by M was significantly decreased in everolimus-stented plaques (***p < 0.001 vs. polymer control), whereas the strut circumference surrounded by SMC was unaffected (p = 0.64). (F) The RAM-11–positive area in the plaque was lowest in everolimus-stented plaques (*p < 0.05 vs. polymer control).

(A) Western blot analysis of everolimus-binding protein FKBP12 and phosphorylated (Ser2448) and total mTOR. ß-Actin served as a loading control. (B) Western blot analysis of phosphorylated and total p70 S6 kinase in cells treated with everolimus (10 µmol/l) for 0 to 12 h showed dephosphorylation of the downstream mTOR target p70 S6 kinase at site Thr389 and to a lesser extent at sites Thr421/Ser424. Everolimus treatment also resulted in an increased expression of the downstream mTOR target 4E-BP1 without changing its phosphorylation at site Thr37/46. This corresponds to a relative reduction of 4E-BP1 phosphorylation. In contrast, phosphorylation of initiation factor eIF2 and elongation factor eEF2 increased. All of these effects were similar between M and SMC. All protein bands correspond to their molecular weight (size markers not shown). (C) Administration of everolimus (10 µmol/l) resulted in a significant reduction of de novo protein synthesis in both M and SMC. Cycloheximide (10 µg/ml; CHX) was used as a positive control. Versus control: **p < 0.01; ***p < 0.001. 4E-BP1 = 4E-binding protein 1; eEF2 = eukaryotic elongation factor 2; eIF2 = eukaryotic initiation factor 2; FKBP12 = FK506-binding protein 12; mTOR = mammalian target of rapamycin; other abbreviations as in Figure 1.

(A) Cells were exposed to 10 µmol/l everolimus (0 to 24 h). (B) Cells were exposed to 1 to 10 µmol/l everolimus for 24 h. Neutral red uptake showed that viability was preserved exclusively to SMC. In contrast, macrophages underwent cell death. (C) Cells were exposed to 10 µmol/l tacrolimus (0 to 24 h). (D) Cells were exposed to 1 to 10 µmol/l tacrolimus for 24 h. Neutral red uptake showed that tacrolimus did not induce cell death in M and SMC. (E, F) Similar to everolimus, cycloheximide (10 µg/ml) also induced rapid cell death of macrophages, whereas the effect on viability of SMC was limited. Versus control: *p < 0.05; **p < 0.01; ***p < 0.001. Open bars = M; shaded bars = SMC. Abbreviations as in Figure 1.

(A) Downregulation of mTOR gene expression at the mRNA level in both macrophages (M) and smooth muscle cells (SMC) 24 h after transfection with mTOR-specific siRNA but not with siControl nontargeting siRNA. (B) Downregulation of mTOR gene expression at the protein level in both M and SMC after transfection with mTOR-specific siRNA but not with siControl nontargeting siRNA. (C) mTOR gene silencing induces macrophage cell death of M but not of SMC. Versus control, unpaired Student t test: **p < 0.01; ***p < 0.001. mTOR = mammalian target of rapamycin.

(A) Macrophages and smooth muscle cells (SMC) were treated with cycloheximide (10µg/ml) or everolimus (10 µmol/l) for 0 to 24 h. To characterize the type of cell death induced by both compounds, cleavage of procaspase-3 (procasp-3) and internucleosomal DNA fragmentation (both apoptosis markers) were analyzed using Western blotting (upper panel) and agarose gel electrophoresis (lower panel), respectively. SMC treated with the combination of tumor necrosis factor (TNF)-alpha (30 ng/ml) and cycloheximide (CHX, 20 µg/ml) for 12 h served as a positive control. (B) Assaying degradation of long-lived proteins as a biochemical marker for autophagy was studied by treating cells with everolimus (10 µmol/l) or Earle's balanced salt solution (EBSS) for 12 h. Versus control: **p < 0.01; ***p < 0.001. (C) Western blot analysis of microtubule-associated protein light chain 3 (LC3) processing was performed in both cell types as an alternative method to detect autophagy. Cells were treated with everolimus (10 µmol/l; 0 to 12 h) or EBSS (8 h). All results are representative of 3 independent experiments.

(A) Ultrastructural features of a normal mouse macrophage as an untreated control with normal cell morphology. (B to D) Treatment of macrophages with everolimus (10 µmol/l) showing different stages of autophagic cell death, which was characterized by cell shrinkage, extensive vacuolization (*), depletion of organelles, and presence of an intact, nonpyknotic nucleus (N). (E) Ultrastructural features of a normal mouse smooth muscle cells (SMC) as an untreated control. (F) Autophagy was not induced in everolimus-treated (10 µmol/l, 6 h) SMC. Scale bar = 3 µm.

(A) Ultrastructural features of a macrophage in an atherosclerotic plaque of rabbit carotid arteries. (B) In vitro treatment of these atherosclerotic plaques with everolimus (10 µmol/l) for 3 days resulted in autophagic cell death and was characterized by cell shrinkage, depletion of organelles, and presence of large autophagosomes containing membranous whorls and remnants of cytoplasmatic material. (C) Ultrastructural features of a smooth muscle cell (SMC) in atherosclerotic plaques of rabbit carotid arteries. (D) Autophagy was not induced in SMC in atherosclerotic plaques treated with everolimus (10 µmol/l) for 3 days. Scale bar=3 µm. Arrowheads = autophagy vesicles; arrows = myeline figure; L = lipid droplet; N = nucleus.