Low-dose irradiation and HL ischemia induce progenitor mobilization. (A–C) MMP-9^+/+ and MMP-9^−/− mice were irradiated with 2 Gy. (A) BM cell number per femur following low-dose IR and (B) white blood cell recovery (n = 8) was determined (*P < 0.05 comparing MMP-9^+/+ with MMP-9^−/− mice). (C) BM cells were collected after IR and cultured overnight in serum-free conditions. (a) Supernatants were assayed via zymography for active MMP-9 and densitometric quantification (small blot; n = 3). (b) Immunohistochemical staining for MMP-9 in BM sections of irradiated animals. Arrows indicate staining for MMP-9. (D, E) Plasma was assayed for KitL (D) and VEGF (E) via ELISA (n = 2; *P < 0.05 comparing MMP-9^+/+ group vs. MMP-9^+/+ + 2 Gy, MMP-9^−/−, or MMP-9^−/− + 2 Gy). (F–J) HL ischemia was induced in MMP-9^+/+ and MMP-9^−/− mice followed by low-dose 2 Gy IR or no IR. (F) Muscle tissue of mice irradiated with 2 Gy was stained for MMP-9 in WT animals on day 0 and day 10. Vessels stained strongly positive for MMP-9 after IR. Arrows indicate positively staining vessels. (G, H) PBMCs were analyzed for the presence of CFU-Cs (G) and CFU-ECs (H) (n = 6/group). (I) Plasma VEGF was measured (n = 2; 3–5 mice/group). (J) PBMCs isolated 7 d after IR and HL ischemia was induced were stained for VEGFR-2/Cy2 and c-kit/PE and analyzed via FACS (left). Single- and double-positive VEGFR-2/c-kit PBMCs isolated with a MoFlo cell sorter were stained for Toluidine blue (right). Error bars indicate mean ± SEM.

MMP-9–dependent release of VEGF by mast cells and of KitL by stromal cells promotes mast cell migration, a necessary requirement for ischemic regeneration. (A, B) HL ischemia was induced in MMP-9^+/+ and MMP-9^−/− mice, and mice were irradiated with 2 Gy or left unirradiated. Sections of ischemic muscle (A) and skin (B) were stained with Toluidine blue to detect mast cells (a–m). Arrows indicate Toluidine positive mast cells in the ischemic muscle tissue. Adjacent sections of muscle tissue were used for VEGF in situ hybridization (e–h) and immunostaining for VEGFR-2 (i–l). (A, i and l) Arrows indicate the same mast cells stained on adjacent sections for VEGFR2 in ischemic muscle tissue. (A, k) Arrow indicates a VEGFR2 positively stained vessel in ischemic muscle tissue. (B) Arrows indicate Toluidine positively stained mast cells in the muscle tissue. *P < 0.05 (three independent experiments, 3 mice/group; original magnification ×100). (C) BM-derived mast cells from MMP-9^+/+ and MMP-9^−/− mice were plated into transwells. The chemoattractants KitL and VEGF were added to the lower chamber. Data are shown as a percentage of migrated cells (two independent experiments; *P < 0.05 for migration of MMP-9^+/+ vs. MMP-9^−/− cells). (D) Mast cells from MMP-9 WT mice, a mast cell line PT18, and stromal cells (MS-5) were cultured in serum-free medium overnight. Supernatants were assayed for VEGF via ELISA. *P < 0.05 comparing Media with 2 Gy or 2 Gy + KitL groups or 2 Gy group with 2 Gy + KitL group. (E) BM stromal cells and BM-derived mast cells from MMP-9^+/+ mice were irradiated with 2 Gy in the presence and absence of MPI CGS 27023A. After 24 h in culture, supernatants were assayed for KitL (two independent experiments). *P < 0.05 comparing Media versus MPI or 2 Gy and 2 Gy versus 2 Gy + MPI. (F–G) HL ischemia was induced in Sl/Sl^d and WBB6F1^+/+ mice. Groups of mice were irradiated with 2 Gy or were left unirradiated (n = 7). (F) Mice were analyzed via laser Doppler for blood perfusion in the ligated limb (left and right column) or were evaluated macroscopically (middle column). (G) Quantification of blood flow (Flux) after vessel ligation. Insert: percentage of limb rescue. *P < 0.05 comparing WT versus Sl/Sl^d mice and HL versus HL + 2 Gy group. Error bars indicate mean ± SEM.

Irradiation modulates neo-angiogenesis within the ischemic microenvironment through VEGF release from mast cells. IR induces up-regulation of MMP-9, resulting in the release of KitL. KitL not only promotes mobilization of progenitors into circulation, which ultimately can be incorporated in ischemic tissue, but also acts as a potent proliferation and migration factor for mast cells. In a partly MMP-9–dependent manner, IR enhances VEGF release from mast cells, which further promote local accumulation of mast cells. VEGF directly can locally support angiogenesis and can further up-regulate MMP-9.

Irradiation-induced BM regeneration and progenitor mobilization are MMP-9–dependent and coincide with an increase in angiogenic factors. MMP-9^+/+ and MMP-9^−/− mice were irradiated with 6.5 Gy. (A) BM cell number (n = 8/group), (B) CFU-Cs (n = 3/group), and (C) CFU-ECs (n = 6/group) were assessed in BM cells following IR (two separate experiments). (D) White blood cell counts (n = 30/group). (E, top) BM cell lysates were assayed via Western blot analysis. (bottom) BM cells obtained after IR were cultured overnight in serum-free medium. BM supernatants were assayed for active MMP-9 via zymography (n = 2; representative experiment shown). Zymographic blot and densitometric quantification are shown. (F) Immunohistochemistry of BM sections before (a) and after IR (b–d) for MMP-9 (brown staining, arrows) in MMP-9^+/+ and MMP-9^−/− mice (original magnification, ×100; insert, ×200). Number of (G) CFU-Cs and (H) CFU-ECs was determined in peripheral blood of irradiated mice. (I, J) Plasma of MMP-9^−/− and MMP-9^+/+ mice was assayed for (I) KitL (n = 5/group) and (J) VEGF (n =5 /group) levels via ELISA. *P < 0.05 comparing the MMP-9^+/+ and MMP-9^−/− group. Error bars indicate mean ± SEM.

Low-dose IR-mediated revascularization is delayed in MMP-9^−/− mice. (A–E) HL ischemia was induced in MMP-9^+/+ and MMP-9^−/− mice. Mice then received a single low-dose TBI (2 Gy) or no IR. (A) Muscle sections from the ischemic limb were stained with hematoxylin-eosin for vWF via immunohistochemistry and with the endothelial marker CD31 (red fluorescence) and SMA (green fluorescence). Mice were analyzed for mean body temperature via thermography 10 d (n = 6/group) and laser Doppler for blood perfusion in the ligated limb 14 d after vessel ligation. (B) Necrosis area of hematoxylin-eosin–stained muscle sections was quantified (n = 3/group). (C) vWF⁺ vessels were counted in muscle sections 7 d following ligation and concomitant IR (n = 3/group). (D) Quantification of the mean temperature 10 d after ligation of the artery and vein (n = 6/group). (E) Blood flux was still low in MMP-9^−/− mice on day 14, whereas blood flux had returned to levels seen in nonligated limbs in irradiated MMP-9^+/+ mice. *P < 0.001 comparing MMP-9^+/+ mice with HL ischemia without and with IR. **P < 0.05 comparing MMP-9^−/− mice with HL ischemia without and with IR or comparing MMP-9^+/+ and MMP-9^−/− with HL ischemia. Error bars indicate mean ± SEM.

Irradiation-induced angiogenesis is dictated by MMP-9–mediated factor release from mast and stroma cells and incorporation of BM-derived cells. (A–C) HL ischemia was induced in Sl/Sl^d and WBB6F1^+/+ mice, followed by 2 Gy IR or no IR (n = 7). PBMCs were analyzed for the presence of CFU-Cs (A) and CFU-ECs (B). (C) Plasma VEGF was determined (n = 7/group). (D) HL ischemia was induced in MMP-9^+/+ mice followed by 2 Gy IR. Mice were subdivided into groups receiving neutralizing mAbs against VEGFR-1, VEGFR-2 or a combination of both mAbs. (a) Blood flow was determined. Insert: Percentage of mice showing limb rescue after ligation. (b) Macroscopic pictures taken of ischemic and contralateral limb 7 d after ligation (*P < 0.05 comparing WT^+/+ + HL + 2 Gy against VEGFR treatment groups). (E) Administration of rVEGF intraperitoneally augmented MMP-9 staining (arrows indicate MMP-9 staining, brown) in the BM (a) and increased levels for KitL in the plasma of VEGF-treated mice (b) (*P < 0.05 comparing VEGF with PBS control). (E, F) MMP-9^+/+ mice received vessel ligation (red bar) followed by local IR (white arrow) with 2 Gy of the ischemic or the contralateral, nonischemic limb. Arrows indicate that the specified limb received local irradiation, whereas the contralateral limb did not. Macroscopic tissue recovery (on day 16) and blood flow as determined via Doppler analysis on day 10 are shown. (G) From day 1–14, PBMCs from irradiated and unirradiated donor mice were isolated and injected daily into C57BL/6 recipients after HL ischemia surgery. Blood flow was determined on day 7. (H) Bilateral vessel ligation (both limbs) was performed in C57BL/6 mice. Groups of mice received 2 Gy IR unilateral (left, lt). The nonirradiated limb (right, rt) served as a control. Blood flow was quantified via laser Doppler analysis (*P < 0.05 comparing bilateral HL vs. bilateral HL + local 2 Gy). (I) C57BL/6 mice were transplanted with GFP⁺ BMNCs. After cell engraftment, mice underwent vessel ligation with or without 2 Gy IR. (left). Ischemic muscle tissue was stained with SMA-Cy Red. (right). The percentage of GFP⁺ vessels per total vessels is shown. (A–C) *P < 0.05 comparing WT^+/+ + HL against WT^+/+ + HL + 2 Gy, Sl/Sl^d + HL, or Sl/Sl^d + HL + 2 Gy. Error bars indicate mean ± SEM.