OUP user menu

Leptin promotes the mobilization of vascular progenitor cells and neovascularization by NOX2-mediated activation of MMP9

Marco R. Schroeter, Susanne Stein, Nana-Maria Heida, Maren Leifheit-Nestler, I-Fen Cheng, Rajinikanth Gogiraju, Hans Christiansen, Lars S. Maier, Ajay M. Shah, Gerd Hasenfuss, Stavros Konstantinides, Katrin Schäfer
DOI: http://dx.doi.org/10.1093/cvr/cvr275 170-180 First published online: 7 November 2011


Aims Bone marrow (BM) progenitors participate in new vessel formation and endothelial repair. The leptin receptor (ObR) is expressed on hematopoietic cells; however, the effects of leptin on BM progenitor cells and their angiogenic potential are unknown.

Methods and results In the present study, we show that the short-term administration of leptin (over five consecutive days) into wild-type mice increased the number of circulating, BM-derived sca-1+, flk-1+ vascular progenitors, 95 ± 1.7% of which also expressed ObR. Ex vivo stimulation of BM cells with leptin enhanced the expression of NADPH oxidase isoform 2 (NOX2), and the leptin-induced increase in reactive oxygen species production, matrix metalloproteinase-9 (MMP9) expression and circulating soluble KitL levels was absent in mice lacking NOX2. Furthermore, intraperitoneal injections of leptin improved perfusion and increased the number of BM-derived, CD31-positive endothelial cells in ischaemic hindlimbs after femoral artery ligation. The effects of leptin on the mobilization of sca-1+, flk-1+ cells and neovascularization were abolished in mice transplanted with BM from ObR-deficient and in NOX2−/− mice.

Conclusion Our findings suggest that the angiogenic effects of leptin involve sca-1+, flk-1+ vascular progenitor cells mobilized from the BM in response to ObR-mediated activation of NOX2, increased MMP9 expression, and sKitL release.

  • Angiogenesis
  • Foetal liver kinase-1
  • Leptin
  • NADPH oxidase
  • Progenitor cell mobilization

1. Introduction

Angiogenesis, the formation of new capillaries from pre-existing vessels, is a complex process involving the proliferation, migration, and capillary tube formation of endothelial cells (EC) in response to growth factors and cytokines. In addition, progenitor cells released from the bone marrow (BM) in response to ischaemia or mediators signalling tissue injury have been shown to play an important role. For example, cytokines or growth factors such as stem cell factor (SCF or soluble kit ligand, sKitL), stromal cell-derived factor-1α, or vascular endothelial growth factor (VEGF) have been shown to create a chemotactic gradient, thus facilitating the evasion of progenitor cells from the BM into the circulation.1,2 Importantly, the cytokine-induced mobilization of progenitor cells was found to enhance the repair of injured arteries and to promote myocardial or peripheral neovascularization.3,4

The adipose tissue-derived cytokine leptin, a 16 kDa hormone identified and cloned in 1994,5 has been shown to exert angiogenic effects on EC,6,7 and to stimulate corneal neovascularization in rats and leptin-deficient ob/ob mice.8 Also, prevention of leptin binding to its receptor inhibited angiogenesis,9 and leptin receptor (ObR)-deficient db/db mice were characterized by reduced basal capillary density and defective neovascularization of ischaemic muscles.10 Despite the consistency of these findings, the cellular targets and molecular mechanisms underlying the angiogenic effects of leptin are largely unknown. Recently, we showed that ex vivo pre-stimulation of cultured human circulating angiogenic cells (also referred to as early outgrowth endothelial progenitor cells or EPC11) improved their functional properties, including integrin-dependent adhesion, re-endothelialization, and angiogenesis.12,13 It is further known that ObR is expressed on a number of different cell types, including haematopoietic cells.14 Therefore, in the present study, we sought to dissect the effects of leptin on BM-derived progenitor cells and particularly on their mobilization and contribution to ischaemia-induced angiogenic processes.

2. Methods

A detailed description of the Methods is available in a separate Supplementary material online.

2.1 Experimental animals

Male C57Bl6 wild-type (WT) mice or mice deficient in the catalytic subunit of NADPH oxidase-2 (NOX2−/−)15 were intraperitoneally injected with either recombinant murine leptin (R&D Systems; 0.6 µg/g body weight in 100 µL normal saline),16 the granulocyte colony-stimulating factor (G-CSF) analog filgrastim (Amgen; 0.25 µg/g body weight),17 or an equal volume of vehicle alone (control). Mice were injected once daily for a total of five consecutive days. A subgroup of mice was lethally irradiated, followed by transplantation of BM from Green Fluorescent Protein (GFP) transgenic or ObR-deficient db/db mice. Blood, spleen, or BM was harvested at the time of sacrifice (on day 5 of treatment; at least 3 h after the last injection) from anaesthetized (via 2% isoflurane inhalation) mice. All animal care and experimental procedures were approved by the State Board of Animal Welfare in accordance with national (i.e. Directive 2010/63/EU of the European Parliament) and institutional guidelines for the care and use of laboratory animals.

2.2 Flow cytometry

Peripheral blood (PB) and BM cells were immunolabelled by incubating cell suspensions (1 × 106 cells) with fluorescence-conjugated antibodies against murine c-kit (BD Biosciences), flk-1 (eBioscience), ObR (R&D Systems), or sca-1 (BD Biosciences), followed by flow cytometry. The generation of reactive oxygen species (ROS) was measured by loading cells with dichlorodihydrofluorescein diacetate (DCF-DA; Invitrogen).

2.3 Isolation, cultivation, and analysis of spleen-derived murine progenitor cells

The angiogenic properties of murine progenitor cells in vitro were examined after cultivation of mononuclear, spleen-derived cells for 7 days under EC-specific conditions, using the matrigel or the spheroid angiogenesis assays.12,18

2.4 Murine hindlimb ischaemia model

On day 3 of leptin or vehicle administration, neovascularization in vivo was examined using the unilateral hindlimb ischaemia mouse model, as described.12,18 Briefly, mice were anaesthetized by intraperitoneal injection of 2% xylazin (6 mg/kg body weight)/10% ketamine hydrochloride (100 mg/kg BW), and placed on a 37°C heating plate. Anaesthesia was monitored by keeping the breath rate at 60–100 min−1 and the absence of reactions to painful stimuli. The right femoral artery and the distal portion of the saphenous artery were ligated, followed by excision of the segment in between both ligations to minimize collateral artery formation. Blood perfusion was determined using laser Doppler perfusion imaging (PeriScan PIM III, Perimed). Capillary density was assessed after immunostaining with antibodies against murine CD31 (Santa Cruz Biotechnology).

2.5 Immunohistochemistry

Paraformaldehyde-fixed, EDTA-decalcified, paraffin-embedded longitudinal sections through the tibia were analysed using antibodies against murine matrix metalloproteinase-9 (MMP9) or ObR (both SantaCruz).

2.6 RNA preparation and polymerase chain reaction

Total RNA from BM was extracted using TRI Reagent (Ambion), and reverse transcription (RT) or quantitative real-time polymerase chain reaction (PCR) was carried out. The primer sequences, conditions and predicted sizes of the amplified product are listed in the Supplementary material online, Table.

2.7 Determination of leptin and soluble c-kit ligand plasma levels

Plasma was obtained from venous blood after cardiac puncture at the time of sacrifice. Plasma leptin or sKitL levels were determined using enzyme-linked immunoassays (R&D Systems).

2.8 Western blot analysis

Equal amounts of bone marrow cell (BMC) protein were fractionated by electrophoresis, transferred to nitrocellulose membranes, and incubated with antibodies against MMP9 (SantaCruz) as well as protein kinase B (AKT) and phospho-AKT, signal transducer and activator of transcription-3 (STAT3) and phospho-STAT3, phospho-protein kinase C (PKC, pan; all Cell Signaling Technology) or GAPDH (HyTest). Visualization of protein bands was achieved by chemiluminescent detection and autoradiography.

2.9 Statistical analysis

All data are presented as mean ± SEM. Differences between leptin- and vehicle-treatment were tested by Student's t-test for unpaired means. If more than two groups were compared, the one way analysis of variance test was performed, followed by the Bonferroni's multiple comparison test. Plasma leptin level data sets did not pass the normality test and are presented as median values. They were analysed using the Kruskal–Wallis test followed by Dunn's multiple comparison test. Statistical analyses were performed using GraphPad PRISM software (version 4.01).

3. Results

3.1 Leptin mobilizes vascular progenitor cells from the BM

To examine the role of leptin in progenitor cell mobilization, WT mice were intraperitoneally injected with recombinant murine leptin (0.6 µg/g body weight) or an equal volume (100 µL) of vehicle (0.9% NaCl) alone. As positive control for progenitor cell mobilization, a subgroup of mice was injected with G-CSF and analysed in parallel. On day 5 of daily administration of either leptin, G-SCF, or vehicle, the number of sca-1+, c-kit+, and sca-1+, flk-1+ circulating progenitor cells was quantified. Of note, plasma leptin levels were significantly elevated in mice injected with leptin (median, 67 ng/mL) compared with those receiving vehicle (1.3 ng/mL; P< 0.001). Flow cytometry analysis revealed that G-CSF significantly increased the number of sca-1+, c-kit+ progenitor cells mobilized into the PB (P< 0.05 vs. controls), whereas only moderate, non-significant changes were detected after leptin administration (Figure 1A). On the other hand, leptin significantly increased the number of circulating sca-1+, flk-1+ vascular progenitor cells (P< 0.05 vs. controls; Figure 1B, representative dot blots shown in panel C). Their BM origin was confirmed by repeating experiments in WT mice after transplantation of BM from GFP-transgenic donors. Again, the number of circulating GFP+, flk-1+ cells was significantly increased after leptin administration (P< 0.01 vs. control; Figure 1D). Of note, leptin did not alter the number of total leucocytes (3.4 ± 0.5 vs. 3.7 ± 0.5 × 103 cells/µL in controls; P= ns), whereas white blood counts were markedly increased in mice receiving G-CSF (6.9 ± 1.0 × 103 cells/µL; P< 0.01 vs. controls and P< 0.05 vs. leptin).

Figure 1

The effects of leptin or G-CSF on the number of c-kit+, sca-1+ (A) and sca-1+, flk-1+ (B) progenitor cells within the PB were examined. *P< 0.05 vs. control-treated mice. Summarized findings in 14–17 mice per group. (C) Representative dot blots after incubation of PB cells from control-, G-CSF-, or leptin-treated mice with isotype control or specific antibodies, respectively, and flow cytometry. (D) The number of circulating flk-1+ cells was examined in mice transplanted with BM from GFP transgenic mice. Summarized findings in eight mice per group. **P< 0.01 vs. controls.

3.2 The effects of leptin on the mobilization of vascular progenitor cells require expression of the leptin receptor

RT–PCR analysis of whole BM homogenates isolated from the femurs and tibias of WT mice confirmed mRNA expression of the short and the long isoform of ObR, whereas leptin gene transcripts could not be detected (Figure 2A). Expression of ObR protein on cells within the BM was confirmed using immunohistochemistry (Figure 2B). Flow cytometry revealed that 20 ± 2.5% of BMC were positive for ObR (Figure 2C). Of those, 59 ± 0.7% also expressed sca-1 and c-kit, and 95 ± 1.7% were positive for sca-1 and flk-1 (Supplementary material online, Figure S1). Importantly, leptin administration to WT mice, transplanted with BM from ObR-deficient db/db mice, failed to significantly increase the number of circulating sca-1+, flk-1+ vascular progenitor cells, in contrast to WT mice transplanted with BM from WT mice (P< 0.05; Figure 2D).

Figure 2

(A) Messenger RNA expression of the short (ObRshort) and long (ObRlong) ObR isoform or leptin in BMC. Total RNA from brain (Br) or adipose tissue (AT) was analysed as positive control. Messenger RNA expression of β-actin served as internal standard. (B) Immunohistochemistry of tibia sections from WT mice confirmed expression of ObR (red signal; left panel) on BMC. Results after omission of the first antibody (negative control, NC; right panel) are also shown. Size bars, 100 µm. (C) Flow cytometry confirmed that BM cells express ObR. Representative histograms (n= 3 experiments) using FITC-labelled antibodies against ObR (black curve) or isotype control (IgG; grey filled curve) are shown. (D) The effects of leptin on the number of sca-1+, flk-1+ progenitor cells were examined in WT mice transplanted with BM from WT (WTBM→WT) or db/db (db/dbBM→WT) mice. Summarized findings in four mice per group. *P< 0.05 vs. control-treated mice.

3.3 Leptin enhances the expression of NOX2 and MMP9 within the BM

In order to investigate potential mechanisms mediating the effects of leptin on sca-1+, flk-1+ progenitor cell mobilization, BM homogenates from mice treated with leptin, G-CSF, or vehicle alone were analysed for the expression of factors known to be involved in progenitor cell release, focusing on leptin-induced NADPH oxidase activation and potential downstream targets. Preliminary experiments confirmed that NOX2 is the NADPH isoform predominantly expressed in murine BMC (Figure 3A).19 Leptin significantly increased the mRNA expression of NOX2 (P< 0.001; Figure 3B) as well as of the NADPH oxidase subunits p22phox (P< 0.05) and p47phox (P< 0.05), whereas p67phox mRNA expression levels were unchanged (P = ns; data not shown). Consistently, ex vivo stimulation of WT BMC with leptin (100 ng/mL) increased the production of ROS (e.g. DCF-positive BMC increased from 17.8 to 45.4% of total cells), whereas this effect was completely absent in BMC isolated from NOX2−/− mice (Figure 3C). Furthermore, increased levels of MMP9 mRNA (Figure 3D) and protein (Figure 3E) expression were detected in the BM of leptin- compared with vehicle-treated animals, and ex vivo stimulation of BMC from WT mice with leptin confirmed a time-dependent increase in MMP9 protein (data not shown). Findings that leptin failed to augment MMP9 mRNA (Supplementary material online, Figure S2A) and protein (data not shown) expression in db/db BM-transplanted WT mice, but also in NOX2−/− mice suggested that activation of NOX2 downstream of the leptin receptor may be involved in the leptin-induced MMP9 production. In this regard, western blot analyses revealed a ObR-, but also NOX2-dependent phosphorylation of protein kinase B (AKT), PKC, and STAT3 in response to ex vivo stimulation of BMC with leptin (Supplementary material online, Figure S2B). Proteolytic cleavage of c-kit via MMP9 has been shown to mediate the release of soluble c-kit ligand (sKitL), thus permitting the transfer of endothelial and haematopoietic stem cells from the quiescent to the proliferative niche.20 Thus, plasma sKitL levels were determined and found to be markedly elevated in leptin-compared with vehicle-treated mice (P< 0.0001), whereas leptin did not alter sKitL levels in db/db BM-chimeric WT (P = ns vs. control-treated mice) or NOX2−/− mice (P = ns; Figure 4A). Similar findings were observed after analysis of BM supernatants (53.4 ± 6.4 vs. 71.2 ± 4.1 pg/mL and P< 0.05 for the difference between control- and leptin-treated mice; n= 12 per group). Consistently, administration of leptin into NOX2−/− mice did not alter the number of circulating flk-1+ cells (P = ns vs. control-treated and P< 0.05 vs. leptin-treated mice; Figures 4B and C).

Figure 3

(A) PCR analysis of NOX isoform 1, 2, and 4 mRNA expression in BMC from WT and NOX2−/− mice. WT mouse spleen homogenisates were used as positive control. (B) Systemic administration of leptin increased NOX2 mRNA expression within BMC. ***P< 0.001 vs. control-treated cells. (C) Flow cytometry revealed increased ROS production (visualized by the ROS indicator DCF) in BMC isolated from WT (left histogram), but not from NOX2−/− (right histogram) mice, after ex vivo stimulation with leptin (100 ng/mL; black line) compared with PBS (control; red line). BMC incubated with 0.01% H2O2 (green line) served as positive control. Representative histograms from three independent experiments are shown. (D) Normalized (to β-actin) MMP9 mRNA expression was significantly increased in BM from G-CSF- (positive control; n= 11) and leptin-treated mice (n= 14). *P< 0.05 and **P< 0.01 vs. controls (n= 16). (E) Immunohistochemistry revealed increased MMP9 expression (brown signal) within the BM (representative tibia sections) of leptin- (right panel) compared with control-treated (left panel) mice. Size bars, 100 µm.

Figure 4

(A) Elevated plasma sKitL levels were detected in leptin- (n= 51) compared with vehicle-treated mice (n= 46), whereas this finding was absent in db/db BM-transplanted WT mice (db/dbBM→WT; n= 8), or NOX2−/− mice (n= 6). ***P< 0.001 vs. control-treated mice. (B and C) Flow cytometry confirmed a significant increase in flk-1+ cells in the PB of WT mice, whereas this finding was absent in NOX2−/− mice. Representative histograms (B) after analysis of BM from control- (grey curve) or leptin-treated mice (black curve) and the summary of quantitative analyses in three to four mice per group are shown (C). *P< 0.05 vs. control; #P< 0.05 vs. leptin-treated WT mice.

3.4 The angiogenic effects of leptin involve cells from the BM

BM-derived progenitor cells may contribute to new blood vessel formation,21,22 and previous studies have shown that leptin promotes angiogenesis.6,7,12 To examine the importance of the leptin-induced vascular progenitor mobilization for angiogenic processes, unilateral hindlimb ischaemia was induced in WT mice after 3 daily i.p. injections of recombinant leptin. Serial laser Doppler perfusion imaging revealed an enhanced blood flow recovery in leptin-compared with vehicle-treated mice which became significant on days 21 and 28 after induction of hindlimb ischaemia (P< 0.05 for both time points; Figure 5A). Histological analysis of tissue sections through the ischaemic muscle 3 weeks after vascular ligation confirmed significantly increased numbers of CD31+ capillary EC per muscle fibre in mice injected with leptin (P< 0.001; Figure 5B), and similar findings were obtained in GFP BM chimeric WT mice confirming their origin from the BM (P< 0.01; Figure 5C). Furthermore, systemic administration of leptin increased the number of acLDL+, lectin+ MNC cultivated from murine spleen homogenates under endothelial conditions (P< 0.001; Supplementary material online, Figure S3A). Also, higher numbers of progenitor cells isolated from the spleen of leptin-treated mice were found to adhere to EC networks (P< 0.001) or sprouts (P< 0.001), as shown in the matrigel (Supplementary material online, Figure S3B) and the spheroid (Supplementary material online, Figure S3C) angiogenesis assay, and similar findings were obtained after analysis of progenitor cells from GFP BM chimeric mice (Supplementary material online, Figure S4A and B). Importantly, leptin failed to augment neovascularization after induction of hindlimb ischaemia in db/db BM-transplanted WT mice, in contrast to its effects in WT mice transplanted with WT BM, as demonstrated by quantification of CD31-positive EC (P< 0.01; Figure 5D) and laser Doppler measurements of hindlimb blood perfusion (P< 0.05; Figure 5E and F). Similarly, hindlimb perfusion of NOX2−/− mice injected with leptin did not differ from vehicle-treated WT mice (Figures 5E and G).

Figure 5

(A) Summary of findings after serial laser Doppler imaging of hindlimb blood perfusion before and up to 28 days after hindlimb ischaemia in WT mice treated with leptin or vehicle (n= 15 mice per group; *P< 0.05 vs. controls). (B) Leptin (i.p. injection once daily for five consecutive days) enhanced the number of CD31+ EC within ischaemic muscles. Representative images of CD31-immunopositive EC (red) in gastrocnemius muscle sections from control- or leptin-treated WT mice. Cell nuclei were counterstained with DAPI (blue). Size bars, 100 µm. The number of CD31+ cells per muscle fibre was manually counted in 9–10 mice per group. ***P< 0.001 vs. controls. (C) Analysis of WT mice transplanted with BM from GFP transgenic mice revealed that leptin enhanced the number of GFP+ cells co-localizing with CD31+ EC. Size bars, 100 µm (n= 6 mice per group). **P< 0.01 vs. controls. (D) Summarized findings in n= 8–12 mice per group after quantification of CD31+ EC (red) in gastrocnemius muscle sections from control- and leptin-treated WT mice, transplanted with BM from either WT (WTBM→WT) or db/db (db/dbBM→WT) mice. *P< 0.05 and **P< 0.01 vs. control-treated mice; ##P< 0.01 vs. leptin-treated mice. (E) Representative laser Doppler images immediately before or 1 h and 3 weeks after induction of hindlimb ischaemia in vehicle- or leptin-injected WT mice, transplanted with BM from WT (WTBM) or db/db (db/dbBM) mice, and in NOX2−/− mice are shown (n= 3–5 mice per group). The summary of the quantitative analysis of WTBM and db/dbBM mice is given in (F) and of NOX2−/− mice in (G). *P< 0.05 for control- vs. leptin-treated WTBM mice; ##P< 0.01 for leptin-treated db/dbBM vs. WTBM mice.

4. Discussion

The findings of the present study show that systemic administration of leptin resulting in the acute elevation of circulating leptin levels promotes neovascularization in response to ischaemia and that cells derived from the BM participate in this process. Moreover, we demonstrate that leptin enhances the mobilization of sca-1+, flk-1+ BM progenitors and that this phenomenon requires the expression of the ObR on BM cells. Mechanistically, an ObR-dependent increase in BM NOX2 expression and ROS production was observed in response to leptin stimulation, resulting in elevated expression of MMP9 as well as circulating levels of the potent stem-cell mobilizing factor sKitL.

Although angiogenic effects of the adipokine leptin have been described previously,6,23,24 the molecular mechanisms underlying this pleiotropic action of the adipokine have remained largely unknown. New vessel formation does not only involve the proliferation and migration of local EC, but also appears to require BM-derived, circulating progenitor cells.2,4 We recently showed that leptin promotes re-endothelialization and ischaemia-induced neovascularization by modulating the function of human-circulating angiogenic cells (also referred to as early outgrowth endothelial progenitor cells or EPC).12,13 In these previous studies, EPC (which were isolated from the mononuclear cell fraction of human PB and cultivated for 7 days) were stimulated with recombinant human leptin ex vivo, i.e. before being intravenously injected into immunodeficient mice. To examine the effects of circulating leptin on ischaemia-induced angiogenesis, and in particular to determine whether they involve BM-derived (progenitor) cells, recombinant murine leptin was administered into mice once daily via intraperitoneal injection. Similar to previous studies, injections were performed on five consecutive days25 and hindlimb ischaemia induced on day 3. Although our study was designed to examine the short-term effects of elevated leptin levels, but not those that may occur in states of chronic hyperleptinaemia in obesity, preliminary analysis of WT-mice fed a high-fat diet for 3 weeks in order to induce obesity and hyperleptinaemia indicated that leptin administration retains its ability to increase the number of circulating sca-1+, flk-1+ progenitor cells also in states of chronically elevated leptin levels (Supplementary material online, Figure S5). Interestingly, a short-term increase in circulating leptin levels, peaking on day 2, has been reported in patients after acute myocardial infarction,26,27 pointing to a possible physiological relevance of our findings. Our analyses not only confirm that leptin enhanced neovascularization in response to ischaemia, but also suggest a role for BM-derived cells in this process, in accordance with previous reports demonstrating that circulating BM-derived cells participate in adult new vessel formation.2,4,28

The mobilization of vascular progenitors from the BM in response to ischaemia or tissue injury is a dynamic process controlled by haematopoietic growth factors and cytokines.1 ObR is known to be expressed on haematopoietic cells,14 and our analyses confirm the expression of its two major isoforms on cells within the BM. In addition, we demonstrate that administration of leptin into mice was associated with elevated numbers of circulating sca-1+, flk-1+ endothelial progenitor cells, the majority of which also expressed the leptin receptor. The requirement of ObR for mediating the effects of the adipokine on the mobilization of vascular progenitors from the BM, but also for ischaemia-induced neovascularization, was demonstrated by the absence of findings in WT mice transplanted with BM from ObR-deficient mice. Of note, progenitor cell mobilization in response to leptin was not the result of a non-specific inflammatory reaction, as observed after administration of G-CSF, known to target a broad range of haematopoietic and non-haematopoietic cells. Also, administration of G-CSF increased the number of sca-1+, c-kit+ cells, a more primitive murine haematopoietic stem cell fraction, while only moderately affecting the number of flk-1+ vascular progenitors. Similar findings, i.e. a specific mobilization of endothelial-committed progenitor cells, have been reported for other factors with angiogenic properties, including VEGF,29 erythropoeitin,30 or drugs such as statins.31

Proinflammatory cytokines may promote the mobilization of progenitor cells by enhancing the secretion of proteolytic enzymes from haematopoietic and stromal cells within the BM. For example, daily administration of G-CSF was found to mobilize haematopoietic progenitor cells by activating MMP9 expression.25 With regard to leptin, previous studies in human endothelial and arterial smooth muscle cells have shown that the adipokine may upregulate MMP2, MMP9, and TIMP1 mRNA expression.23 In the present study, administration of leptin into mice, as well as ex vivo stimulation of freshly isolated BM cells were found to enhance the MMP9 gene and protein expression, whereas the expression of MMP2 and TIMP1 was unaltered or downregulated, respectively (data not shown). Recently, BM MMP9 expression was shown to be essential for ischaemia-induced neovascularization and involved in the mobilization of sca-1+, flk-1+ progenitors.32 The angiogenic effects of leptin were absent in db/db BM transplanted WT and in NOX2−/− mice, both of which also were characterized by the absence of leptin-induced MMP9 expression, thus indirectly supporting a role of the protease in mediating the effects of leptin.

Activation of MMP9 mediates the proteolytic cleavage and release of sKitL.20 Consistent with the observation that stimulation of BM cells with leptin induced MMP9 expression, plasma sKitL levels were markedly elevated in mice having received recombinant leptin. Soluble KitL binds to the receptor tyrosine kinase c-kit, highly expressed on adult haematopoietic stem cells,33 and has been shown to be critical for the mobilization of BM-derived progenitor cells in response to different agents (e.g. estradiol, VEGF, PlGF).34 Thus, our findings suggest that the leptin-induced, MMP9-mediated elevation of sKitL may underlie the vascular progenitor cell-mobilizing effects of leptin in vivo. Interestingly, flow cytometry analysis indicated that approximately 59% of the ObR+ cells within the BM also express c-kit.

Previous studies indicate that MMP9 expression could be regulated by ROS.35 In this regard, NADPH oxidase is a major source of ROS in many cell types, including haematopoietic and progenitor cells,19 and can be activated by growth factors and cytokines.36 For example, stimulation of EC with VEGF was found to result in a NOX2-dependent increase in ROS levels,37,38 and BM NOX2 expression was recently shown to be essential for EPC mobilization and neovascularization.19,39 Also, previous findings in vascular cells revealed that leptin may stimulate ROS formation through NADPH oxidase-dependent pathways.24,40,41 In the present study, we show that leptin increased the expression of NOX2 as well as the NADPH oxidase subunits p22phox and p47phox in BM cells. Moreover, the leptin-induced mobilization of flk-1+ cells was abolished in NOX2-deficient mice, and findings that the increase in BM MMP9, but also NOX2 expression in response to leptin was absent in mice transplanted with ObR-deficient BM suggest a signalling cascade involving NOX2-induced MMP9 expression downstream of the leptin receptor. Also, increased phosphorylation of signalling proteins (STAT3, AKT, PKC) known to be activated after binding of leptin to its receptor were detected after stimulation of BMC with leptin, and this finding was absent not only in ObR-deficient BM-chimeric mice, but also in mice lacking NOX2, underscoring the importance of NOX2-generated ROS as second messengers in the leptin-induced signal transduction, as previously shown for EC.24,41

Besides the mobilization of sca-1+, flk-1+ vascular progenitor cells, potential mechanisms of leptin's contribution to the improved neovascularization after systemic administration may also involve the enhanced recruitment of mobilized progenitors to sites of ischaemia and the stimulation of specific angiogenic processes such as adhesion, proliferation, and migration. In this regard, we have previously shown that leptin promotes the integrin-mediated adhesion of circulating EPC and enhanced re-endothelialization of vascular lesions.13 Consistently, in the present study, the effects of leptin were not limited to increasing the number of circulating sca-1+, flk-1+ progenitor cells, but also enhanced their recruitment to sites of injury and improved their functional capacities in vitro, as suggested by findings in spleen-derived endothelial progenitors isolated from leptin- or vehicle-treated mice.

Taken together, a (short-term) increase in circulating leptin levels in mice was found to augment neovascularization after ischaemia in vivo. We showed that this effect involved the interaction of leptin with ObR-positive cells within the BM and activation of specific signal transduction pathways (in particular, NOX2, MMP9, and sKitL) leading to an enhanced mobilization of differentiated and functionally active vascular progenitors from the BM. The schematic diagram in Supplementary material online, Figure S6 depicts possible mechanisms involved in mediating the effects of leptin on flk-1+, sca-1+ vascular progenitor cell mobilization.


This work was supported in part by research grants from the German Research Foundation (Deutsche Forschungsgemeinschaft) and the ‘Novartis Stiftung für Therapeutische Forschung’ (both to K.S.).


The authors gratefully acknowledge the technical expertise of Sarah Barke and Anika Hunold.

Conflict of interest: none declared.


  • Present address. Department of Cardiology, Democritus University of Thrace, Alexandroupolis, Greece.


View Abstract