We read with interest the recent article by Hu et al1 who investigated the origin of smooth muscle cells (SMCs) in transplant atherosclerotic lesions. The authors proposed that SMCs in graft vasculopathy derive from non-bone marrow cells of host origin. 1 On the other hand, recent evidence suggests that bone marrow cells also differentiate into the SMCs that substantially contribute to vascular diseases, including graft vasculopathy. 2-4

We are concerned that such opposite conclusions are drawn using different experimental systems to test the hypothesis. Hu et al1 used a novel mouse model of arterial transplantation. The authors grafted a donor aorta to a recipient carotid artery using the cuff technique in the absence of immunosuppressants. The pathophysiology of vasculopathy in the authors’ model may differ from that of arteriosclerosis observed in immunosuppressed cardiac allografts, as many studies have suggested that immunosuppressants potentially affect the pathogenesis of transplant arteriosclerosis. By inducing 3 distinct types of mechanical vascular injuries to a single mouse, we found that recruitment of bone marrow-derived cells to vascular lesions depends largely on the type of model used (K. Tanaka and M. Sata, unpublished observation, 2002). We detected a robust contribution of bone marrow-derived cells after mechanical endovascular injury, whereas there were only a few marker-positive cells in the neointima after ligation of the common carotid artery. We seldom found bone marrow-derived cells in the lesion induced by perivascular cuff replacement. These results suggest that in certain types of models, we may underestimate the potential contribution of bone marrow cells in vascular remodeling. Furthermore, Hu et al1 used a transgenic mouse that expresses LacZ in vascular SMCs under the transcriptional control of the SM22 promoter. Human vascular SMCs are quite heterogeneous in the expression of differentiation markers, although α-actin is abundantly expressed in all SMCs. 5 Thus, some α-actin–positive cells may not express LacZ in the transgenic mouse. Consequently, although the transgenic mice are useful to identify SMCs, they may not be adequate to identify “SMC-like cells” in the neointima. Finally, we are aware that irradiation and bone marrow transplantation have tremendous effects on many tissues, even when the blood system appears to have been reconstituted. Therefore, Hu et al1 might have underestimated the potential contribution of bone marrow cells to graft vasculopathy. We also believe that there are other sources of smooth muscle progenitors. We suggest caution before extrapolating the conclusions provided by Hu et al to the pathogenesis of graft vasculopathy.