Recruitment of the inflammatory subset of monocytes to sites of ischemia induces angiogenesis in a monocyte chemoattractant protein-1-dependent fashion

Abstract

There is accumulating evidence that delivery of bone marrow cells to sites of ischemia by direct local injection or mobilization into the blood can stimulate angiogenesis. This has stimulated tremendous interest in the translational potential of angiogenic cell population(s) in the bone marrow to mediate therapeutic angiogenesis. However, the mechanisms by which these cells stimulate angiogenesis are unclear. Herein, we show that the inflammatory subset of monocytes is selectively mobilized into blood after surgical induction of hindlimb ischemia in mice and is selectively recruited to ischemic muscle. Adoptive-transfer studies show that delivery of a small number of inflammatory monocytes early (within 48 h) of induction of ischemia results in a marked increase in the local production of MCP-1, which in turn, is associated with a secondary, more robust wave of monocyte recruitment. Studies of mice genetically deficient in MCP-1 or CCR2 indicate that although not required for the early recruitment of monocytes, the secondary wave of monocyte recruitment and subsequent stimulation of angiogenesis are dependent on CCR2 signaling. Collectively, these data suggest a novel role for MCP-1 in the inflammatory, angiogenic response to ischemia.

Fig. 1.

Recruitment of adoptively transferred bone marrow mononuclear cells to the ischemic hindlimb. CFSE-labeled bone marrow mononuclear cells (1×10⁶) were infused i.v. into C57BL/6 recipient mice 24 h after the induction of unilateral hindlimb ischemia. The adductor muscle from the ischemic and nonischemic hindlimb was isolated at the indicated times after adoptive transfer and was digested with collagenase to generate a single-cell suspension for flow cytometry. (A) Number of CFSE⁺ F4/80⁺ monocytes recruited to the ischemic and nonischemic adductor muscle over time (n=3–6 per time-point). (B) Number of CFSE⁺ Gr-1⁺ neutrophils recruited to the ischemic and nonischemic adductor muscle (n=3 per time-point). (C) Number of endogenous (non-CFSE-labeled) F4/80⁺ monocytes recruited to the ischemic adductor muscle (n=3–6 per time-point; *, P<0.001, compared with no adoptive-transfer control at the same time-point). (D) Number of endogenous (non-CFSE-labeled), Gr-1⁺ neutrophils recruited to the ischemic adductor muscle (n=3–6 per time-point). Data represent the mean ± sem.

Fig. 2.

Histological assessment of monocyte recruitment to ischemic muscle. Representative photomicrographs showing monocytes (F4/80⁺, red) and endothelial cells (CD31⁺, green) in ischemic adductor muscle 24 h after adoptive transfer of bone marrow mononuclear cells (A) or saline alone (B). Sections were counterstained with 4′,6-diamidino-2-phenylindole to identify nuclei. (C). Isotype control for F4/80. Data are representative of four independent experiments.

Fig. 3.

Contribution of the inflammatory subset of monocytes to reperfusion. Hindlimb ischemia was surgically induced in CX3CR1^GFP/+ mice, and bone marrow, peripheral blood, and adductor muscle were harvested at baseline, 24 h, and 48 h after surgery. Inflammatory and resident monocytes were quantified in each tissue by flow cytometry. (A) Peripheral blood; (B) bone marrow; (C) ischemic adductor muscle. (D) 2.5 × 10⁵ CX3CR1^hi Gr-1⁻ resident or CX3CR1^lo Gr-1⁺ inflammatory monocytes were adoptively transferred into wild-type recipient mice 24 h after induction of hindlimb ischemia. Blood flow was measured by laser Doppler perfusion imaging. *, P < 0.05, relative to Time 0 for the same genotype; **, P < 0.01, relative to Time 0 for the same genotype. Data represent the mean ± sem.

Fig. 4.

Local production of MCP-1, VEGFA, and SDF-1 in ischemic adductor muscle. Ischemic adductor muscle was harvested at the indicated time, and the tissue was digested with collagenase to generate a single-cell suspension. Cells and debris were removed by centrifugation, and the protein levels of MCP-1 (A), SDF-1 (B), and VEGFA (C) in the tissue supernatant were measured by ELISA (n=3–6 per time-point). *, P < 0.05; **, P < 0.001, compared with no adoptive transfer). Data represent the mean ± sem. WT, Wild-type.

Fig. 5.

Reperfusion in MCP-1^−/− mice. (A) Restoration of blood flow in wild-type (C57BL/6) or MCP-1^−/− mice following induction of hindlimb ischemia. (B) Restoration of blood flow after adoptive transfer of 1 × 10⁶ wild-type bone marrow mononuclear cells (BM-MNC; or saline) into wild-type or MCP-1^−/− mice 24 h after induction of hindlimb ischemia. (C and D) CFSE-labeled bone marrow mononuclear cells (1×10⁶) were infused i.v. into wild-type or MCP-1^−/− recipient mice 24 h after the induction of hindlimb ischemia. (C) Number of CFSE⁺, F4/80⁺ adoptively transferred monocytes recruited to the ischemic adductor muscle 4 h after adoptive transfer (n=6 for each genotype). NS, Not significant. (D) Number of endogenous (non-CFSE-labeled) F4/80⁺ monocytes recruited to the ischemic adductor muscle 28, 48, and 72 h after the induction of hindlimb ischemia (n=3–6 per time-point; **, P<0.001). Data represent the mean ± sem.

Fig. 6.

MCP-1, SDF-1, and VEGFA levels in ischemic muscle from MCP-1^−/− mice. The level of MCP-1 (A), SDF-1 (B), and VEGFA (C) in muscle supernatant was measured, as described in Figure 4. *, P < 0.05; **, P < 0.01; ***, P < 0.001, relative to no adoptive transfer. Data represent the mean ± sem.

Fig. 7.

Reperfusion in CCR2^−/− chimeric mice. CCR2^−/− bone marrow chimeras were established by transplanting CCR2^−/− bone marrow cells into irradiated wild-type mice. Following hematopoietic reconstitution (5–6 weeks), hindlimb ischemia was induced in these chimeras (or wild-type C57BL/6 mice), and 24 h later, CFSE-labeled, wild-type bone marrow cells were infused i.v. (A) Number of CFSE⁺, F4/80⁺ monocytes recruited to the ischemic adductor muscle 4 h after adoptive transfer (n=3–6 for each genotype). (B) Number of endogenous (non-CFSE-labeled) F4/80⁺ monocytes recruited to the ischemic adductor muscle 48 h after induction of hindlimb ischemia (n=3–6 for each genotype; ***, P<0.0001). (C) Restoration of blood flow in wild-type mice (C57BL/6) or CCR2^−/− chimeras after adoptive transfer of 1 × 10⁶ wild-type bone marrow cells or saline alone (PBS). Data represent the mean ± sem.