CXCR4 blockade augments bone marrow progenitor cell recruitment to the neovasculature and reduces mortality after myocardial infarction

Abstract

We hypothesized that a small molecule CXCR4 antagonist, AMD3100 (AMD), could augment the mobilization of bone marrow (BM)-derived endothelial progenitor cells (EPCs), thereby enhancing neovascularization and functional recovery after myocardial infarction. Single-dose AMD injection administered after the onset of myocardial infarction increased circulating EPC counts and myocardial vascularity, reduced fibrosis, and improved cardiac function and survival. In mice transplanted with traceable BM cells, AMD increased BM-derived cell incorporation in the ischemic border zone. In contrast, continuous infusion of AMD, although increasing EPCs in the circulation, worsened outcome by blocking EPC incorporation. In addition to its effects as a CXCR4 antagonist, AMD also up-regulated VEGF and matrix metalloproteinase 9 (MMP-9) expression, and the benefits of AMD were not observed in the absence of MMP-9 expression in the BM. These findings suggest that AMD3100 preserves cardiac function after myocardial infarction by enhancing BM-EPC-mediated neovascularization, and that these benefits require MMP-9 expression in the BM, but not in the ischemic region. Our results indicate that AMD3100 could be a potentially useful therapy for the treatment of myocardial infarction.

Fig. 1.

A single injection of AMD3100 after MI mobilizes EPCs, improves survival, reduces adverse cardiac remodeling, and increases capillary density. Mice received single s.c. injections of AMD3100 (AMD) or saline 1 h after MI. (A) Peripheral blood (PB) EPC counts were evaluated via the EPC culture assay and costaining for acLDL uptake and BS-1 lectin (n = 5). (B–D) To ensure suitable mortality for statistical analyses, MI was induced at the most proximal site of the coronary artery. (B) Survival rate of mice treated with AMD (solid line, n = 48) or saline (broken line, n = 23); subsequent assessments were performed in mice that survived through day 28. (C) Fibrosis was quantified as the ratio of the length of fibrosis to the LV circumference (AMD: n = 12; saline: n = 5); fibrotic tissue appears blue-green. (D) Capillary density was evaluated by staining the vasculature via in vivo BS-1 lectin perfusion before killing, and then quantifying the number of BS-1 lectin–positive (brown) vessels (AMD: n = 12; saline: n = 5). HPF, high-power field. *, P < 0.05; **, P < 0.03.

Fig. 2.

A single injection of AMD3100 after MI increases BM EPC incorporation into ischemic tissue. WT mice transplanted with Tie2-LacZ BM received single s.c. injections of AMD3100 (AMD) or saline 1 h after MI. (A) BM EPCs were identified in the LV by X-gal staining (blue-green) for LacZ expression, and (B) the number of BM EPCs in the ischemic border zone was quantified (n = 4). (C) BM EPC incorporation in the ischemic border zone on day 28 was evaluated by staining for β-gal (green) and CD31 (red) expression and quantifying the number of double-positive cells (yellow) (n = 3). HPF, high-power field. *, P < 0.05; ***, P < 0.01. (Scale bar: 100 μm.)

Fig. 3.

Continuous AMD3100 infusion increases adverse cardiac remodeling, decreases capillary density, and impairs BM EPC incorporation. (A and B) WT mice or (C) WT mice transplanted with Tie2-LacZ BM received single injections of AMD3100 (AMD) or were implanted with osmotic pumps that continuously infused AMD3100 or saline for 2 weeks after MI. (A) Fibrosis was quantified as the ratio of the length of fibrosis to the LV circumference (n = 5); fibrotic tissue appears blue. (B) Capillary density was evaluated by staining the vasculature via in vivo BS-1 lectin perfusion before killing, and then quantifying the number of BS-1-lectin–positive vessels (green) (n = 5). (C) BM EPC incorporation in the ischemic border zone on day 14 after MI was quantified as the number of β-gal-positive cells (red) (n = 3). HPF, high-power field; i, ischemic tissue; n, nonischemic tissue. *, P < 0.05; ***, P < 0.01.

Fig. 4.

The beneficial effects associated with a single injection of AMD3100 are dependent on MMP-9 expression. (A) WT mice, (B) MMP-9 knockout (KO) mice, or (C and D) WT and MMP-9 KO mice received single s.c. injections of AMD3100 (AMD) or saline 1 h after MI. (A) MMP-9 expression in PB or BM MNCs was evaluated via quantitative RT-PCR; measurements were normalized to 18S rRNA and expressed as the fold-change from levels before MI (n = 4). (B) PB EPC counts were determined in MMP-9 KO mice via flow-cytometry analyses for coexpression of Sca1 and Flk1 (n = 4). (C) Capillary density on day 28 after MI was evaluated by staining the vasculature via in vivo BS-1 lectin perfusion before killing, and then quantifying the number of BS-1 lectin–positive (brown) vessels (n = 5). (D) Fibrosis was quantified as the ratio of the length of fibrosis to the LV circumference; fibrotic tissue appears blue-gray (n = 5). *, P < 0.05; NS, not significant.

Fig. 5.

The beneficial effects associated with a single injection of AMD3100 are dependent on BM MMP-9 expression, but not on MMP-9 expression in the ischemic tissue. MMP-9 KO mice transplanted with WT BM and WT mice transplanted with MMP-9 KO BM received single s.c. injections of AMD3100 (AMD) or saline 1 h after MI. (A and B) PB EPC counts were evaluated via flow-cytometry analysis for coexpression of Sca1 and Flk1. (A) n = 3 at all time points. (B) Before MI, n = 9; day 1, n = 3 for AMD and 4 for saline; day 3, n = 3; day 7, n = 4 for AMD and 3 for saline; day 28, n = 4. (C and D) LV fractional shortening was evaluated echocardiographically. (C) Before MI, n = 3; day 7, n = 5; day 14, n = 10; day 28, n = 11 for AMD and 9 for saline. (D) Before MI, n = 5; day 7, n = 4; day 14, n = 10; day 28, n = 10 for AMD and 11 for saline. (E) Fibrosis on day 28 after MI was quantified as the ratio of the length of fibrosis to the LV circumference. MMP-9 KO/WT_BM, n = 7 for saline and 10 for AMD; WT/MMP-9 KO_BM, n = 11 for saline and n = 6 for AMD. (F) Capillary density on day 28 was evaluated by staining the vasculature via in-vivo BS-1 lectin perfusion before killing, and then quantifying the number of BS-1 lectin–positive vessels. MMP-9 KO/WT_BM, n = 6 for saline and 7 for AMD; WT/MMP-9 KO_BM, n = 14 for saline and 7 for AMD. HPF, high-power field. *, P < 0.05; **, P < 0.03; ***, P < 0.01; NS, not significant.

Fig. 6.

AMD3100 administration immediately after MI increases VEGF expression in PB and BM MNCs 3 to 7 d later. VEGF activity is required for AMD3100-induced MMP-9 expression. (A and B) Mice received single s.c. injections of AMD3100 (AMD) or saline 1 h after MI and then VEGF mRNA expression in (A) PB and (B) BM MNCs was measured via quantitative RT-PCR, normalized to endogenous 18S rRNA expression, and expressed as the fold-change from levels before MI. *, P < 0.05; n = 5. (C and D) AMD3100- and saline-treated mice received i.p. injections of a VEGF-neutralizing antibody (antiVEGF) or an IgG antibody after surgically induced MI. Five days later, MMP-9 mRNA expression was measured in (C) PB and (D) BM MNCs. mRNA measurements were normalized to endogenous 18S rRNA expression and expressed as the fold-change from measurements obtained in uninjured, untreated mice. *, P < 0.05; ***, P < 0.01; n = 5.