Human leukemia inhibitory factor inhibits development of experimental atherosclerosis

Abstract

The effect of human leukemia inhibitory factor (hLIF) on the development of atherosclerosis was investigated in an experimental animal model. Two conditions were examined: one in which lesions could arise because of the influence of both "injury" (cuffed vessel) and diet and one in which only the effect of diet could be significant in other areas of the vasculature (aorta). At time zero, the right carotid artery of rabbits (n=32) was ensheathed in a soft Silastic cuff, and an osmotic minipump (2-mL capacity; 2.5 μL/h; 28 days) containing either hLIF or saline was inserted into the peritoneal cavity. Rabbits were divided into four groups (n=8): group 1 received normal diet/saline; group 2, normal diet/LIF (30 μg·kg-1·d-1); group 3, 1% cholesterol diet/saline; and group 4, 1% cholesterol diet/LIF (30 μg·kg-1·d-1). After 28 days, the cholesterol diet (group 3) resulted in a sixfold increase in plasma cholesterol level compared with group 1 rabbits on a normal diet (3.80±0.50 versus 0.55±0.01 mmol/L). This was significantly lower (P=.01) with hLIF treatment in group 4 rabbits (2.80±0.44 mmol/L). Group 2 rabbits had higher aortic tissue cholesterol levels (1.40±0.35 mg/g) compared with group 1 rabbits on a normal diet (0.10±0.06 mg/g) (P=.01), whereas hLIF treatment decreased tissue cholesterol levels by 60% in group 4 rabbits (0.60±0.05 mg/g) (P=.01). Group 3 rabbits developed lipid-filled lesions covering 63.25±17.66% of the thoracic aorta surface, whereas lesions were significantly reduced (9.88±8.79%) (P=.01) with LIF treatment (group 4). No thoracic aorta lesions were present in groups 1 and 2. Cuff placement around the right carotid artery in groups 1 and 3 resulted in neointimal formation (as percent total wall crosssectional area) of 15.06±8.83% and 19.30±5.36%, respectively. In LIF-treated animals (groups 2 and 4), neointima composed only 2.12±5.45% and 2.09±4.32%, respectively, which represented a significant reduction (P=.01). Immunocytochemical analysis revealed no subendothelial macrophages in rabbits fed a normolipidic diet with or without hLIF treatment, but 153.60±15.60 and 165.30±22.40 (P=.677) macrophages per ring segment were found in animals fed a cholesterol diet in the presence and absence of hLIF, respectively. Together these results show that hLIF lowers plasma cholesterol and inhibits the development of aortic fatty streaks in cholesterol-fed rabbits. hLIF also inhibits the development of cuff-induced carotid neointimal thickening in rabbits fed either a normal or cholesterol diet but does not affect mono-cyte invasion of the vessel wall.