Both experimental and clinical studies have recently suggested that bone marrow (BM)-derived cells can contribute to the repair of an infarcted heart, although the precise mechanism underlying this clinically promising effect remains to be determined. Orlic et al. were the first to report that BM stem cells could be mobilized by a combination of granulocyte colony-stimulating factor (G-CSF) and stem cell factor, and differentiated into cardiomyocytes in the infarcted myocardium, improving cardiac remodeling (

Orlic D et al. Proc Natl Acad Sci USA. 2001;98:10344
). A number of subsequent studies have shown that G-CSF alone can also improve cardiac function and survival rate (
Harada M et al. Nat Med 2005;11:305
,
Ince HPM et al. Circulation 2004;110(suppl III):352
). We have analyzed the precise lineages of cells in the infarcted myocardium, to determine the contribution of subpopulations of BM-derived cells to G-CSF-associated cardiac remodeling. In the previous experiments in which the hematopoietic or mesenchymal compartment of irradiated mice was repopulated with hematopoietic stem cells (HSCs) or mesenchymal stem cells (MSCs) labeled with green fluorescence protein (GFP) followed by G-CSF treatment, we showed that GFP+ cardiomyocytes were generated by MSCs, but not by HSCs (
Kawada H et al. Blood 2004;104:3581
). However, the number of regenerated cardiomyocytes derived from MSCs was so small that they could not solely explain the beneficial effect of G-CSF on cardiac remodeling. On the other hand, we also found many HSC-derived non-myocytes concentrated in the infarcted area. Therefore, we next focused on the contribution of HSC-derived cells to the healing process after MI. Single CD34c-kit+Sca-1+lineage side population cells from GFP-transgenic mice were used to reconstitute hematopoiesis in the recipients. MI was then induced. Our results first confirmed that G-CSF administration after MI significantly improved the survival rate, cardiac remodeling and function. At the acute phase (7th day), the infarcted area contained a number of GFP+, elongated cells, which expressed vimentin and alpha-smooth muscle actin (SMA), indicating that they were myofibroblasts. G-CSF augmented the recruitment of these cells to the infarcted area. At the chronic phase (60th day), most of the GFP+ cells in the infarcted area were elongated cells and still positive for vimentin, but negative for SMA, suggesting that myofibroblasts had differentiated into mature fibroblasts. The number of GFP+ fibroblasts in the G-CSF-treated mice was significantly higher than in the control mice. GFP+ cardiomyocytes were not observed in the infarcted myocardium of G-CSF-treated mice both at the acute and chronic phases. Taken together, our results strongly suggested that HSC-derived cardiac myofibroblast numbers can be increased by G-CSF administration, and these cells are responsible for improved cardiac function and survival by restoring structural integrity to the infarcted myocardium.

Author notes

Corresponding author

Sign in via your Institution