New evidence suggests that the artery wall is a recipient and source of MSCs. The long-recognized formation of ectopic mesenchymal tissue in the artery wall was a clue that MSCs are present in the adult artery wall. These stem cells, which were first identified in the marrow stroma, can differentiate along multiple lineages and give rise to cartilage, bone, fat, muscle, and vascular tissue in vitro and in vivo. Similar cells with multilineage and self-renewal capacity have also been harvested and cultured from muscle, blood, fat, and now adult vascular tissue. PBMCs also may serve as progenitors for endothelium and possibly smooth muscle. Marrow cells appear to enter the circulation and engraft vascular and other connective tissues, especially at sites of injury and in tissue transplant grafts. Conversely, provocative evidence suggests that cells from vascular tissue may enter the circulation and engraft remote organs. Yet, the possibility remains that MSCs from the bone marrow or the artery wall may permit human tissue regeneration and repair, an exciting future prospect for cardiovascular disease.

In pioneering work, Prockop et al. showed that transplanted marrow cells engraft connective tissues such as spleen and liver. Liechty et al. showed that human marrow cells engraft multiple tissues when injected in utero in sheep. The concept of a continuous replacement of connective tissue with marrow cells parallels the known continuous replacement of blood by bone marrow hematopoietic cells. For hematopoietic stem cells, lineage acquisition is regulated by well-defined colony-stimulating factors. Little is known about the corresponding regulatory mechanisms for MSCs. Hematopoietic stem cells were first found in the marrow. They are now harvested from peripheral blood for clinical regeneration of marrow. MSCs were also first identified in the marrow, and they are now harvested from a variety of tissue sources. It has been suggested that MSCs serve as universal repair cells in adult tissues as they undergo regeneration or remodeling.

MSCs in Bone Marrow

Marrow stromal cells, a subpopulation of nonhematopoietic cells in the bone marrow, are the prototypical MSCs. They represent a small percentage of marrow cells, and they can be partially distinguished from hematopoeitic cells by their ability to adhere to tissue culture dishes. In culture, they may be guided to differentiate into bone, fat, cartilage, or muscle cells using specific media. When cultured in ordinary media containing 10% to 20% fetal serum, they form multiple clusters of different lineages within a single dish. They are guided into specific, single-lineage differentiation by culture in serum-free �induction media� containing growth factors and other treatments such as insulin, dexamethasone, and indomethacin. Recent evidence indicates that even marrow MSCs that have already fully differentiated into 1 lineage are capable of transdifferentiation into another lineage in response to induction media. Marrow MSCs also maintain multipotential capacity in vivo, tending to differentiate into the cell type of the tissue they engraft. This tendency to adopt local identity may be directed by local cytokines and matrix factors from host cells. Thus, adequate contact with host cells may be required to ensure that engrafted cells assume host tissue identity rather than differentiating along other mesenchymal lineages. Importantly, the evidence that marrow MSCs transdifferentiate to adopt local tissue identity may be confounded by the phenomenon of cell fusion. In this process, a labeled donor cell fuses with a differentiated host cell, resulting in a cell positive for label and differentiation marker, falsely indicating transdifferentiation. To date, artifact attributable to cell fusion has been an issue in studies of nonvascular tissues, but in principal, it may occur in any cell type. Marrow-derived MSCs also have vascular differentiation potential. In culture, they differentiate into smooth muscle cells (SMCs). In vivo, bone marrow�derived cells that were seeded on a synthetic vascular graft produced smooth muscle and endothelial layers. Although it is attractive to consider that growth of MSCs in injured vascular tissue may simply regenerate normal vascular tissue, it is also possible for these cells to produce ectopic tissues, such as those seen in advanced atherosclerotic calcification. In addition, although marrow cell�induced changes may result directly from assimilation of marrow cells, the benefit may also be an indirect, paracrine effect of cytokines released by the marrow-derived cells. The factors governing these possibilities are not clear, but it appears that metabolic diseases, such as atherosclerosis or diabetes, favor the formation of ectopic tissue.

MSCs in Arteries

The tunica media has been widely perceived as a homogeneous, terminally differentiated layer of SMCs. However, many investigators have shown that cultured SMCs are heterogeneous and that they undergo dedifferentiation in culture. Interestingly, vascular SMCs were described as �multifunctional mesenchymal cells� 36 years ago, foreshadowing the finding of multipotential mesenchymal cells as subpopulations of SMCs within the artery wall

Engraftment by Marrow MSCs

In human gender-mismatched bone marrow transplant recipients, donor marrow cells engraft all layers of atherosclerotic plaques but not disease-free segments, suggesting that engraftment may require inflammation or injury. Similarly, in mice with atherosclerosis, labeled donor marrow cells engraft plaque. Interestingly, recipients have less atherosclerosis when donor marrow-derived progenitor cells are from young versus older mice. In transplanted hearts, a neointima forms throughout the coronaries of the transplanted heart, leading to diffuse transplant coronary arteriopathy and often ischemia. The neointima was originally believed to derive locally from the donor tunica media. However, studies in mouse cardiovascular transplant models have shown that much of the neointima derived from the host. When labeled veins are transplanted into the carotids of unlabeled recipient mice, the label disappears over time, suggesting engraftment by circulating host cells; conversely, unlabeled donor veins transplanted into labeled recipients gain label, again suggesting engraftment by host cells. Although these results could be explained by local migration of vascular cells, the changes in labeling occurred uniformly throughout the graft rather than predominantly at the edges, supporting a circulatory origin for the engrafting cells. Because circulating cells include hematopoeitic as well as mesenchymal cells, some instances of host cell �engraftment� may be attributable to ordinary diapedesis by host leukocytes, especially in areas of inflammation or injury. To exclude this possibility, label and phenotypic markers should be assessed in studies of mesenchymal cell engraftment. In some experimental conditions, even injured vascular tissue is not engrafted by marrow cells. Location and type of injury may determine whether marrow cells engraft. External injury by cuff or ligation in mice stimulates little neointimal and rare SMC layer engraftment, whereas internal injury by femoral wire denudation induces substantial neointimal (40%) and SMC (25%) engraftment. Thus, technical differences may affect results of engraftment studies and need to be taken into consideration in future clinical applications.

Keywords: Marrow stromal cells; hematopoietic cells; engraft; smooth muscle cells; artery wall