Mesenchymal stem cells can participate in ischemic neovascularization

Abstract

Background: Cells from the bone marrow contribute to ischemic neovascularization, but the identity of these cells remains unclear. The authors identify mesenchymal stem cells as a bone marrow-derived progenitor population that is able to engraft into peripheral tissue in response to ischemia.

Methods: A murine model of skin ischemia was used. Bone marrow, blood, and skin were harvested at different time points and subjected to flow cytometric analysis for mesenchymal and hematopoietic markers (n = 3 to 7 per time point). Using a parabiotic model pairing donor green fluorescent protein (GFP)-positive with recipient wild-type mice, progenitor cell engraftment was examined in ischemic tissue by fluorescence microscopy, and engrafted cells were analyzed by flow cytometry for endothelial and mesenchymal markers. In vitro, the ability of both bone marrow- and adipose-derived mesenchymal stem cells to adopt endothelial characteristics was examined by analyzing (1) the ability of mesenchymal stem cells to take up DiI-acetylated low-density lipoprotein and Alexa Fluor lectin, and (2) phenotypic changes of mesenchymal stem cells co-cultured with GFP-labeled endothelial cells or under hypoxic/vascular endothelial growth factor stimulation.

Results: In vivo, the bone marrow mesenchymal stem cell population decreased significantly immediately after surgery, with subsequent engraftment of these cells in ischemic tissue. Engrafted cells lacked the panhematopoietic antigen CD45, consistent with a mesenchymal origin. In vitro, bone marrow- and adipose-derived mesenchymal stem cells took up DiI-acetylated low-density lipoprotein and Alexa Fluor lectin, and expressed endothelial markers under hypoxic conditions.

Conclusions: The authors' data suggest that mesenchymal precursor cells can give rise to endothelial progenitors. Consequently, cell-based therapies augmenting the mesenchymal stem cell population could represent powerful alternatives to current therapies for ischemic vascular disease.

Fig. 1

Surface markers for mesenchymal and hematopoietic stem cells.

Fig. 2

Changes in mesenchymal stem cell prevalence within the total cell population of (above, left) bone marrow, (above, right) circulation, and (below, left) the ischemic skin flap at days 0 to 20 after the ischemic insult were analyzed using flow cytometry. Bone marrow mesenchymal stem cells were identified as lin−/Sca-1 +/CD45−. Similar time-course studies were performed for (below, right) bone marrow–derived hematopoietic stem cells identified as lin−/Sca-1 +/c-kit+/CD45+. The x axis units are expressed in terms of days postoperation (with the exception of a 12-hour time point that is provided for the mesenchymal and hematopoietic stem cell bone marrow analysis) and y axis units are expressed in terms of percentage of total cells for the studied tissue compartment.

Fig. 3

We used a GFP+/wild-type parabiotic model, where circulation is shared between paired mice permitting circulating and bone marrow–derived cells from the GFP+ donor to migrate and engraft within ischemic skin in the wild-type recipient. The ischemic flap was harvested at postoperative day 14.

Fig. 4

Sections of the flap were analyzed using fluorescence microscopy for the presence of GFP+ cells. GFP+ cells in ischemic recipient skin on day 14. Cells from the digested flap were also analyzed further by flow cytometry/fluorescence-activated cell sorting.

Fig. 5

Cells harvested from ischemic skin were initially gated for lineage negativity and CD31 positivity (left). This population was further subgated for bone marrow–derived (GFP+) cells (center). Of the bone marrow–derived endothelial cells present within ischemic skin, the majority were CD45− (i.e., of nonhematopoietic origin) (right). For all flow cytometric plots, x axes are expressed in terms of arbitrary fluorescence units (log scale), whereas y axes are expressed in terms of arbitrary logarithmic side-scatter units.

Fig. 6

Morphology of unlabeled bone marrow–derived mesenchymal stem cells after 5 days of co-culture with GFP+ bEnd.3 cells (left). An image of GFP+ bEnd.3 endothelial cells after 5 days of culture alone (right) is provided for comparison.

Fig. 7

After 1 and 5 days of mesenchymal stem cell/GFP+ bEnd.3 cell co-culture, cells were harvested and analyzed by flow cytometry to evaluate the fraction of cells expressing the endothelial marker CD31 that were derived from GFP− cells (i.e., mesenchymal stem cells or mesenchymal stem cell derivatives).

Fig. 8

Cultured bone marrow–derived mesenchymal stem cells were visualized by fluorescence microscopy after nuclear staining with DAPI (left) and assayed for DiI-acLDL uptake (right).

Fig. 9

Adipose-derived mesenchymal stem cells were visualized by fluorescence microscopy after nuclear staining with DAPI (above), assayed for DiI-acLDL uptake (center), and evaluated for Alexa Fluor lectin binding (below).

Fig. 10

After 5 days of culture in hypoxic conditions with VEGF, expression of the endothelial markers Flk-1 and CD31 by adipose-derived mesenchymal stem cells was evaluated using flow cytometry. Values are expressed as means ± SEM.