• Open Access

Adipose stromal cell and sarpogrelate orchestrate the recovery of inflammation-induced angiogenesis in aged hindlimb ischemic mice

Authors

  • Weiwei Fan,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    2. Department of Cardiology and Geriatrics, Southeast Hospital Affiliated to Xiamen University, Zhangzhou, Fujian, China
    Search for more papers by this author
  • Chengxiang Li,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Xing Qin,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Shenxu Wang,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Hu Da,

    1. Institute of Orthopaedics and Traumatology, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Kang Cheng,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Ri Zhou,

    1. Institute of Orthopaedics and Traumatology, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Chao Tong,

    1. State Key Laboratory of Cancer Biology, Institute of Digestive Diseases, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Xiujuan Li,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Qingting Bu,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Congye Li,

    1. Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Yaling Han,

    1. Department of Cardiology, Shenyang Northern Hospital, Shenyang, China
    Search for more papers by this author
  • Jun Ren,

    Corresponding author
    1. Center for Cardiovascular Research and Alternative Medicine, University of Wyoming, Laramie, WY, USA
    • Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author
  • Feng Cao

    Corresponding author
    • Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China
    Search for more papers by this author

Correspondence

Professor Feng Cao, Department of Cardiology & Molecular Imaging Program, Xijing Hospital, Fourth Military Medical University, 127# West Changle Road, Xi'an, Shaanxi 710032, China. Tel.: 86 29 84771024; fax: 86 29 84771170; e-mail: wind8828@gmail.com;

Professor Jun Ren, Center for Cardiovascular Research and Alternative Medicine, University of Wyoming, Laramie, WY 82071, USA; e-mail: jren@uwyo.edu

Summary

Aging population displays a much higher risk of peripheral arterial disease (PAD) possibly due to the higher susceptibility, poor prognosis, and fewer therapeutic options. This study was designed to examine the impact of combined multipotent adipose-derived stromal cells (mADSCs) and sarpogrelate treatment on aging hindlimb ischemia and the mechanism of action involved. mADSCs (1.0 × 107) constitutively expressing enhanced green fluorescent protein (eGFP) or firefly luciferase (Fluc) reporter were engrafted into the hindlimb of aged Vegfr2-luc transgenic or FVB/N mice subjected to unilateral femoral artery occlusion, followed by a further administration of sarpogrelate. Multimodality molecular imaging was employed to noninvasively evaluate mADSCs' survival and therapeutic efficacy against aging hindlimb ischemia. Aged Tg(Vegfr2-luc) mice exhibited decreased inflammatory response, and downregulation of vascular endothelial growth factor (VEGF)/vascular endothelial growth factor receptor-2 (VEGFR2) compared with young ones following hindlimb ischemia induction, resulting in angiogenesis insufficiency and decompensation for ischemia recovery. Engrafted mADSCs augmented inflammation-induced angiogenesis to yield pro-angiogenic/anti-apoptotic effects partly via the VEGF/VEGFR2/mTOR/STAT3 pathway. Nonetheless, mADSCs displayed limited survival and efficacy following transplantation. Sarpogrelate treatment with mADSCs further upregulated mammalian target of rapamycin (mTOR)/STAT3 signal and modulated pro-/anti-inflammatory markers including IL-1β/TNF-α/IFN-γ and IL-6/IL-10, which ultimately facilitated mADSCs' survival and therapeutic benefit in vivo. Sarpogrelate prevented mADSCs from hypoxia/reoxygenation-induced cell death via a mTOR/STAT3-dependent pathway in vitro. This study demonstrated a role of in vivo kinetics of VEGFR2 as a biomarker to evaluate cell-derived therapeutic angiogenesis in aging. mADSCs and sarpogrelate synergistically restored impaired angiogenesis and inflammation modulatory capacity in aged hindlimb ischemic mice, indicating its therapeutic promise for PAD in the elderly.

Introduction

Peripheral arterial disease (PAD) continues to be a highly prevalent anomaly for aging population. The prevalence of PAD was 2.5% among individuals aged < 60 years, 8.3% for 60–69 years, and an astonishing 18.8% in the elderly (> 70 years) (Criqui et al., 1997; Tendera et al., 2011). Unfortunately, the elderly display impaired vascular responses to ischemia, resulting in adverse outcomes following critical limb ischemia (CLI) (Kinnaird et al., 2008). Furthermore, many older patients with PAD are not suitable for current conventional surgical or endovascular revascularization procedures (Tendera et al., 2011). A number of factors may contribute to the contraindication of these conventional procedures including the severity and complexity of lesion in the absence or presence of other confounding risk factors.

Although the elderly are evidently susceptible to PAD, the precise mechanism of action behind the decompensation for limb ischemia has not been fully elucidated. Available therapeutic strategies are pertinent for aged PAD individuals, especially those with CLI. Previous experimental and clinical evidence from our group has depicted that the cell-based therapy may provide a promising avenue for ischemic tissue rescue (Sheikh et al., 2007; Cao et al., 2009; Sun et al., 2012). Recently, the multipotent adipose-derived stromal cells (ADSCs) have been employed to promote therapeutic angiogenesis from bench to bedside (Gimble et al., 2012; Mizuno et al., 2012). Nonetheless, neither the significance nor the mechanism of action behind ADSC-directed therapy against PAD in aging population has been well evaluated. To this end, effective approaches to improve the functional survival of donor cells are also pertinent to the success of cell-based PAD therapy (Sheikh et al., 2007).

In the present study, we assessed the in vivo kinetics of vascular endothelial growth factor receptor-2 (VEGFR2) expression, as one of biomarkers to evaluate angiogenesis in a CLI model in aged Vegfr2-luc transgenic mice using noninvasive bioluminescence imaging (BLI) technique. Murine ADSCs isolated from young donors with constitutive expression of enhanced green fluorescent protein (eGFP, mADSCsGFP+) or firefly luciferase (Fluc, mADSCsFluc+) reporter were used for hindlimb ischemia therapy. Multimodality imaging strategies were employed to visualize the survival and therapeutic efficacy of engrafted mADSCs as previously described (Cao et al., 2006; Fan et al., 2012). Sarpogrelate (Sarp), a 5-hydroxytryptamine(2A) [5-HT(2A)] receptor antagonist with proven anti-inflammatory and pro-survival properties (Rajesh et al., 2006), was also used to improve functional survival of mADSCs in vivo. We sought to assess whether synergistic mADSCs and Sarp may yield any benefit in aged CLI mice, and the underlying mechanism(s) involved.

Results

Restoration of hindlimb ischemia in aged CLI model following mADSCs transplantation

In this study, mADSCs isolated from allogeneic young and healthy donors were employed. mADSCs could be abundantly isolated (~106 cells per gram raw adipose tissue). mADSCsGFP+ exhibited distinctive fibroblastoid morphology and eGFP expression in vitro (Fig. 1A). Fluorescence imaging (FRI, Fig. 1B) exhibited a robust linear correlation between mADSCsGFP+ number and eGFP optical intensity (r2 = 0.930, Fig. 1C). mADSCsGFP+ possessed multipotency and other characteristics of mADSCs as previously described (Fan et al., 2012). Laser Doppler perfusion imaging (LDPI) spatiotemporally visualized the change in peripheral blood perfusion (Fig. 1D). Mice in the Aged group exhibited a lower level of perfusion recovery in ischemic hindlimb compared with the Young group, suggesting an impaired self-compensation of aged mice following ischemia onset (Fig. 1E). Either mADSCsGFP+ or Sarp administration remarkably promoted hindlimb blood flow in aged CLI mice. Combined therapy of mADSCsGFP+ and Sarp provided more benefit in hindlimb perfusion restoration, as the perfusion ratio (PR) in Aged + Dual group was significantly greater than Aged + ADSCs group between postoperative day 7 (POD7) and POD21 (0.681 ± 0.011 vs. 0.521 ± 0.024 on POD14, < 0.001). Blind scoring depicted that combined therapy of mADSCsGFP+ and Sarp further improved manifestation (Fig. S1A) and ambulatory function (Fig. S1B) of the ischemic hindlimb.

Figure 1.

Blood perfusion imaging and angiogenesis assessment following hindlimb ischemia. (A) Fibroblastoid morphology and green fluorescence of mADSCsGFP+ (3rd passage). (B) Fluorescence imaging exhibits (C) linear correlation between cell quantity and eGFP average radiance, with correlation coefficient r2 value and linear function. Colored scale bar represents fluorescence average radiance in P s−1 cm−2 sr−1. Wild-type mADSCs (mADSCsWT) were used as negative control. (D) Laser Doppler perfusion imaging (LDPI) visualized dynamic changes of hindlimb perfusion. Colored scale bar represents blood flow velocity in LDPI index. (E) Blood perfusion was quantified using perfusion ratio (PR), that is, the ratio of average LDPI index of ischemic (red arrows) and contralateral nonischemic hindlimb. = 20 for each. (F) Immunohistochemical staining for CD31(PECAM-1)-positive capillaries (red arrows) in the ischemic gastrocnemius muscle section on POD1 and POD14. (G) Quantitative analysis demonstrated the ratio of CD31-positive capillaries (red arrows)/muscle fibers in ischemic hindlimb on POD14. (H) Confocal microscopy for gastrocnemius sections with triple immunofluorescence staining of von Willebrand factor (vWF, endothelial cell marker, red), GFP (green) and 4′,6-diamidino-2-phenylindole (DAPI, blue). (I) More GFP-positive cells were observed in the Aged + Dual group compared with the Aged + ADSCs group on POD14. = 20 random fields. Scale bars represent 50 μm for low-power and 5 μm for magnification (Mag, yellow square). Error bars: mean ± SD. *< 0.001 vs. Young, #< 0.001 vs. Aged, †< 0.001 vs. Aged + ADSCs. ADSC, adipose-derived stromal cell; Sarp, sarpogrelate; eGFP, enhanced green fluorescent protein.

Engrafted mADSCs promoted hindlimb angiogenesis

Histological analysis revealed that the aged CLI mice exhibited reduced angiogenic capacity compared with the young ones. The CD31 (PECAM-1)-positive capillaries/muscle fibers ratio within ischemic hindlimb was significantly lower in the aged compared with the young mice on POD14 (although comparable on POD1, Fig. 1F,G). mADSCsGFP+ augmented hindlimb angiogenesis in aged CLI model, while combined therapy of mADSCsGFP+ and Sarp further promoted microvasculature formation. More CD31+ or von Willebrand factor-positive (vWF+) blood vessels were found within the same ischemic region in the Aged + Dual group compared with the Aged + ADSCs group on POD14 (< 0.001 for Fig. 1G and Fig. S2). Although Sarp treatment failed to elicit any significant pro-angiogenic effect, more GFP+ cells were observed in the Aged + Dual group compared with the Aged + ADSCs group on POD14 (< 0.001, Fig. 1H,I). Accordingly, Sarp treatment overtly facilitated mADSCsGFP+ survival, which may contribute to enhanced therapeutic angiogenesis in Aged + Dual group. No engrafted mADSCsGFP+ expressed vWF, indicating little chance for the injected cells to be differentiated or incorporated into vascular cells. However, few GFP+ cells were observed within ischemic muscles after POD35, indicating death of donor mADSCsGFP+.

mADSCs induced Vegfr2-luc expression in aged CLI model

To explore the mechanism of action behind mADSC-mediated angiogenesis, in vivo BLI was employed to monitor Vegfr2-luc expression for 5 weeks (Fig. 2A). The Fluc signal in hindlimb area was detectable on POD1 before reaching a peak between 7 and 14 days following ischemia induction (Fig. 2B). VEGFR2 (KDR/Flk-1) is the major receptor of vascular endothelial growth factor (VEGF) for pro-angiogenesis (Olsson et al., 2006). Aged Tg(Vegfr2-luc) mice displayed lower VEGFR2 transcriptional levels following hindlimb ischemia compared with the young ones, the effect of which was reversed by engrafted mADSCsGFP+. More surviving mADSCsGFP+ in Aged + Dual group further upregulated VEGFR2 expression, as the Fluc signal was still detectable at a later stage of tissue repair (POD35: < 0.001 vs. Aged + ADSCs group). The mADSC-induced Vegfr2-luc expression was validated by the ex vivo Fluc assay (Fig. 2C) and reverse transcription polymerase chain reaction (RT-PCR, Fig. 2D-a,E) analysis.

Figure 2.

Imaging Vegfr2-luc expression following hindlimb ischemia. (A) In vivo bioluminescence imaging (BLI) tracked the spatiotemporal kinetics of Vegfr2-luc expression following hindlimb ischemia induction in Tg(Vegfr2-luc) mice. Colored scale bar represents bioluminescence average radiance in P s−1 cm−2 sr−1. = 20 for each. (B) Quantification of BLI signals and (C) ex vivo luciferase assays demonstrated the mADSC-directed VEGFR2 expression in ischemic hindlimbs, validated using (D-a) RT-PCR with (E) quantitative analysis. (D-b, F) Western blot and (G) ELISA determined mADSC-induced upregulation of VEGFR2(Tyr951)/VEGFR2 and VEGF in the ischemic hindlimb on POD10. LDPI monitored that combined treatment of (H) anti-VEGF monoclonal neutralizing antibody (mAb) or (I) VEGFR2 inhibitor Ki8751 with mADSCs counteracted mADSC-induced hindlimb ischemia restoration, compared with combined nonspecific IgG or vehicle treatment. = 5 for each. Error bars: mean ± SD. *< 0.05 vs. Young, #< 0.001 vs. Aged, †< 0.001 vs. Aged + ADSCs, ††< 0.01 vs. Aged + ADSCs, ‡< 0.01 vs. ADSCs + anti-VEGF mAb, ‡‡< 0.05 vs. PBS + nonspecific mAb, ‖< 0.01 vs. ADSCs + Ki8751, ¶< 0.05 vs. PBS + vehicle. ADSC, adipose-derived stromal cell; VEGF, vascular endothelial growth factor; VEGFR2, VEGF receptor-2; LDPI, laser Doppler perfusion imaging.

Western blot results demonstrated that aging significantly suppressed the level of VEGFR2 and phospho-VEGFR2 in Tg(Vegfr2-luc) mice, the effect of which was restored by mADSCsGFP+ (Fig. 2D-b,F). Furthermore, enzyme-linked immunosorbent assay (ELISA) revealed a remarkably decreased VEGF expression in the aged mice compared with the young ones, the effect of which was also reconciled by mADSCsGFP+ (Fig. 2G). Activation of VEGF/VEGFR2, as potent pro-angiogenic signaling molecules (Olsson et al., 2006), was permissive to mADSC-mediated angiogenesis. Longitudinal and serial LDPI displayed that VEGF neutralizing antibody (Fig. 2H) or selective VEGFR2 kinase inhibitor Ki8751 (Fig. 2I) treatment with mADSCs abrogated mADSC-mediated ischemia restoration to the level equal to models without mADSCs.

Sarp facilitated mADSC-mediated anti-apoptosis through regulation of inflammation

Our data revealed that aged Tg(Vegfr2-luc) mice exhibited higher level of postischemic apoptosis in the hindlimb compared with the young ones, using terminal deoxy-nucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) apoptotic assay on POD14 (Fig. 3A). Either mADSCsGFP+ or Sarp treatment significantly inhibited apoptosis. Reduced apoptosis index (AI) was assessed in treated muscles in the Aged + ADSCs or Aged + Sarp group compared with the Aged group (10.50 ± 1.80% or 8.33 ± 2.63% vs. 17.42 ± 2.25%, < 0.001 respectively, Fig. 3B). Moreover, Sarp plus mADSCs treament provided an incremental benefit in mADSC-mediated anti-apoptosis.

Figure 3.

Apoptosis and inflammation level in aged hindlimb ischemic mice. (A) Confocal microscopy of the gastrocnemius muscle sections stained with TUNEL reagent for apoptotic cells (green) and merged 4′,6-diamidino-2-phenylindole (DAPI) for total nuclei (blue) on POD14. (B) TUNEL-positive apoptotic cells were quantified using apoptosis index (AI), as the number of TUNEL-positive cells divided by total cells per field. = 20 random fields. Expression of (C) interleukin-1β (IL-1β), (D) tumor necrosis factor-alpha (TNF-α), (E) interferon-γ (IFN-γ), (F) IL-6, (G) IL-10, and (H) VEGF were assessed using ELISA. = 5 for each. Scale bar represents 100 μm. Error bars: mean ± SD. *< 0.01 vs. Young, ‡< 0.01 vs. Aged, #< 0.05 vs. Aged, †< 0.001 vs. Aged + ADSCs. ADSC, adipose-derived stromal cell; VEGF, vascular endothelial growth factor.

Assessment of cytokine further revealed that aged Tg(Vegfr2-luc) mice displayed lower postischemic inflammation level compared with the young ones. Following hindlimb ischemia induction, expression of inflammatory cytokines including interleukin-1β (IL-1β, Fig. 3C), tumor necrosis factor-α (TNF-α, Fig. 3D), interferon-γ (IFN-γ, Fig. 3E), IL-6 (Fig. 3F), IL-10 (Fig. 3G), and angiogenic cytokine VEGF (Fig. 3H) was significantly lower in the aged mice than the young mice on POD14. Engrafted mADSCsGFP+ triggered local inflammatory responsiveness, as evidenced by higher IL-1β, TNF-α, IL-6, IL-10 and VEGF levels in the Aged + ADSCs group compared with the Aged group. Sarp treatment effectively attenuated the expression of pro-inflammatory markers IL-1β, TNF-α, and IFN-γ, which are known to induce acute cell death and defective transplanted stem cells (Suzuki et al., 2004; Ishii et al., 2010; Liu et al., 2011). Meanwhile, Sarp promoted IL-6, IL-10 and exerted a nonsignificant effect on VEGF expression, which might protect donor mADSCsGFP+ in response to ischemia (Yao et al., 2006; Xie et al., 2007; Holladay et al., 2011).

Sarp prolonged mADSCs' survival via mTOR/STAT3 in aged CLI mice

We went on to demonstrate that Sarp facilitated engrafted mADSCsFluc+ survival in the ischemic hindlimb of aged FVB/N mice using the noninvasive BLI. mADSCsFluc+ displayed similar characteristics of mADSCsGFP+, with the exception of a stable Fluc expression (Fig. 4A–C). In vivo BLI longitudinally tracked a progressive death of mADSCsFluc+ within a 3-week posttransplant period in the ADSCs group (Fig. 4D), resulting in a decay of BLI signal from (7.13 × 107 ± 1.66 × 106) P s−1 cm−2 sr−1 on POD0 to (6.86 × 106 ± 4.10 × 105) P s−1 cm−2 sr−1 on POD14, and down to background level after POD21 (Fig. 4E). Combined Sarp treatment with mADSCsFluc+ promoted cell survival, as evidenced by a ~40% donor mADSCsFluc+ survival on POD14 (< 0.001 vs. ADSCs group), while the BLI signal was still detectable on POD28. The combined treatment of PP242 (an ATP-competitive inhibitor of mammalian target of rapamycin, mTOR) or S3I-201 (an selective inhibitor of signal transducer and activator of transcription 3, STAT3) suppressed mADSCsFluc+ survival and further counteracted the Sarp-mediated mADSCsFluc+ protection in vivo, which was confirmed by the ex vivo Fluc assay (Fig. 4F).

Figure 4.

Sarp facilitated mADSC survival via mTOR/STAT3 in aged hindlimb ischemic mice. (A, B) mADSCsFluc+ exhibited a linear correlation with firefly luciferase (Fluc) signal and (C) a stable Fluc expression in eight passages. (D) BLI longitudinally tracked the survival and kinetics of mADSCsFluc+ in vivo. = 20 for each. (E) Progressive decay of bioluminescence intensity over time indicates donor cell death, validated using (F) ex vivo luciferase assays on POD21. Western blotting of (G) mTOR(Ser2448)/mTOR and (H) STAT3(Tyr705)/STAT3 expression in the ischemic hindlimb on POD7, quantified using integrated optical density (IOD) ratio of each pair. (I) ELISA quantification (POD7) with (J) run charts demonstrated the level and trend of inflammatory cytokines expression following hindlimb ischemia onset. = 5 for each. Error bars: mean ± SD. *< 0.05 vs. PBS, †< 0.05 vs. ADSCs, **< 0.05 vs. PBS, ††< 0.01 vs. ADSCs, ‡< 0.001 vs. ADSCs, #< 0.001 vs. ADSCs + Sarp, ‖< 0.001 vs. ADSCs, ¶< 0.001 vs. ADSCs + Sarp, ##< 0.05 vs. Sham. ADSC, adipose-derived stromal cell; mTOR, mammalian target of rapamycin; STAT3, signal transducer and activator of transcription 3; BLI, bioluminescence imaging.

Immunoblotting data revealed that mADSCsFluc+ significantly augmented postischemic mTOR (Fig. 4G) and STAT3 (Fig. 4H) phosphorylation, and triggered inflammatory response in ischemic hindlimb (Fig. 4I,J). Sarp further promoted mTOR/STAT3 activation and modulated inflammation, leading to downregulation of IL-1β/TNF-α/IFN-γ and upregulation of IL-6/IL-10. PP242 treatment abolished mTOR activation and partially inhibited STAT3 phosphorylation, while promoting IL-1β/TNF-α/IFN-γ and suppressing IL-6/IL-10 expression. Moreover, S3I-201 selectively inhibited STAT3 phosphorylation, en route to upregulated TNF-α/IFN-γ and downregulated IL-6/IL-10 expression. These data suggested that mTOR/STAT3 activation favored a pro-survival effect on donor mADSCs through regulation of inflammation.

Sarp enhanced mADSCs' survival over hypoxia/reoxygenation via mTOR/STAT3 in vitro

In vitro BLI (Fig. 5A) and TUNEL assay (Fig. S3) revealed that Sarp significantly reinforced the mADSC-associated change in cell survival (Fig. 5B) and anti-apoptosis (Fig. 5C) over H/R. Activation of mTOR (Fig. 5D) and STAT3 (Fig. 5E) was also noted in Sarp-treated mADSCsFluc+ following H/R. Sarp remarkably reduced TNF-α levels and promoted IL-6/IL-10 secretion of mADSCsFluc+ during H/R (Fig. 5F–H). Inhibition of mTOR/STAT3 using PP242 or S3I-201 promoted TNF-α elevation and IL-6/IL-10 suppression, leading to the mADSC-associated cell death and apoptosis during H/R.

Figure 5.

Sarp promoted mADSC survival over hypoxia/reoxygenation in vitro. (A) In vitro BLI and (B) quantification revealed mADSCsFluc+ survival pre-/post- hypoxia/reoxygenation (H/R) treatment. Colored scale bar represents bioluminescence average radiance in P s−1 cm−2 sr−1. (C) Quantitative analysis of apoptotic mADSCs over H/R using apoptosis index (AI), as assessed by TUNEL assay. = 20 random fields. Western blotting demonstrated (D) mammalian target of rapamycin (mTOR) (Ser2448)/mTOR and (E) STAT3(Tyr705)/STAT3 expression of mADSCs following H/R, quantified using IOD ratio of each pair. ELISA documented levels of (F) TNF-α, (G) IL-6 and (H) IL-10 within mADSCsFluc+ supernatants over H/R. = 5 for each. Error bars: mean ± SD. *< 0.001 vs. Control (Normoxia), †< 0.01 vs. H/R, ‡< 0.05 vs. H/R, #< 0.001 vs. H/R + Sarp, ‖< 0.001 vs. H/R, ¶< 0.001 vs. H/R + Sarp. ADSC, adipose-derived stromal cell; BLI, bioluminescence imaging.

Discussion

Although multiple factors may substantially contribute to the higher risk of senile PAD, the age-related angiogenic dysfunction, which results in inadequate collateral development, is sufficient to prompt the adverse prognosis of aged CLI patients. In the present study, aging depressed inflammation and angiogenesis level following CLI onset, as evidenced by overtly reduced pro-/anti-inflammatory cytokines and pro-angiogenic factor VEGF. Moderate inflammatory responsiveness is often necessary for angiogenesis and tissue repair (Imhof & Aurrand-Lions, 2006). The balance between inflammation and angiogenesis, recently considered together as ‘inflammation-induced angiogenesis’ (West et al., 2010), is pinned to several regulators, VEGF/VEGFR2, for example. A number of inflammatory cytokines, for example IL-6, may promote postischemic angiogenesis by activating VEGF/VEGFR2 (Imhof & Aurrand-Lions, 2006; Olsson et al., 2006; Yao et al., 2006). Such process is believed to serve as a link between inflammatory and vascular endothelial cells. Not surprisingly, suppression of inflammatory response following CLI may constitute a significant crux for the angiogenic dysfunction in aged PAD animals, leading to decompensation in perfusion restoration.

Vascular endothelial growth factor receptor-2 tyrosine kinase acts as a crucial regulator of angiogenesis to compensate postischemic cardiovascular dysfunction (Olsson et al., 2006; Lakshmikanthan et al., 2011). VEGF primarily utilizes VEGFR2 to induce pro-angiogenic and anti-apoptotic responses via several signaling cascades, such as Akt/mTOR/STAT3 (Olsson et al., 2006). In this study, we evaluated the spatiotemporal kinetics of VEGFR2 in vivo, as a possible hallmark of compensatory and therapeutic angiogenesis in the context of aging. Using BLI, the longitudinal transcriptional downregulation of VEGFR2 gene, along with a decline in pan protein expression and phosphorylation of VEGFR2, was demonstrated in aged hindlimb ischemic Tg(Vegfr2-luc) mice. Previous work also demonstrated that aging-related defects in response to VEGF and other pro-angiogenic genes may have contributed to the suboptimal compensation in aged hindlimb ischemic mice (Wang et al., 2011). Thus, regulation of VEGF/VEGFR2 activity may represent a definite therapeutic approach for aging CLI.

Recent evidence has shed some light onto the cell-based PAD therapy, although its precise impact on aging PAD was not well defined. In our hands, mADSCs derived from the young, healthy allogeneic mice donors facilitated the recovery of hindlimb angiogenesis and perfusion in aged CLI mice. Engrafted mADSCs promoted the inflammatory responsiveness and upregulated VEGF/VEGFR2 in the ischemic hindlimb. A variety of pro-/anti-inflammatory cytokines were elevated following mADSCs transplantation, which may stimulate the postischemic angiogenesis of aging hindlimb partly via VEGF/VEGFR2. Interestingly, mADSCs were found adjacent to the host microvasculature rather than incorporated or differentiated into it, although the multipotency of mADSCs has been identified in vitro. The para-vascular functional survival of mADSCs is more in accordance with the putative paracrine action of mADSCs by previous report (Rehman et al., 2004). Furthermore, several mADSC-induced cytokines, for example VEGF/IL-10, possess pro-survival potential. Our earlier data demonstrated that VEGF was capable of facilitating stem cell survival in ischemic myocardium, while nascent vasculature was being formed (Xie et al., 2007). ADSC was also shown to respond to a hypoxic condition by regulating VEGF expression in vitro, in a cell-protective mechanism (Rehman et al., 2004). IL-10 is a confirmed anti-inflammatory cytokine to improve cell survival in vivo (Holladay et al., 2011). As a result, mADSCs prevented native myocytes from the postischemic apoptosis in aged CLI mice.

Regulation of inflammation-induced angiogenesis via paracrine action appears to represent the underlying mechanism behind mADSC-directed CLI therapy in the elderly. Previous studies have demonstrated that a single administration of VEGF exhibited insufficiency to promote the maturation of vessels to provide adequate blood perfusion to ischemic lesions (Kidoya et al., 2010). VEGF treatment also failed to yield significant benefits in early clinical trials (Kidoya et al., 2010). The mADSC-based strategy potentiated spatiotemporal presentation of VEGF together with other pro-angiogenic factors, and further enabled activation of VEGFR2. As the harvest of ADSC is relatively easy from an allogeneic donor, young ADSC may thus offer an alternative therapeutic avenue for clinical aging PAD (Gimble et al., 2012; Mizuno et al., 2012). In the current study, however, in vivo survival and therapeutic function of mADSCs were evidently limited. Using the noninvasive BLI, we followed up with the abbreviated lives of mADSCs in the ischemic hindlimb of aged mice. Accordingly, facilitation of the functional survival of engrafted mADSCs is a definite prerequisite prior to introduction of the cells.

The viability of engrafted cells can be severely limited by postischemic microenvironment in vivo (Sheikh et al., 2007), especially in an aging ischemic hindlimb with insufficient angiogenesis and incremental cell apoptosis compared with the young. Although inflammatory response may potentiate cell-induced angiogenesis (Laflamme & Murry, 2011), the overexpression of pro-inflammatory cytokines could further trigger pro-apoptotic signal cascade and suppress the survival of donor cells (Liu et al., 2011). Aging-associated inflammatory disorder also contributes to the high risk and progression of senile PAD (Erren et al., 1999; Bartlett et al., 2012). Our present data indicated Sarp may enable a favorable niche for the functional survival of mADSCs through regulation of inflammation. Sarp treatment effectively modulated the expression of pro-/anti-inflammatory cytokines and reduced cell apoptosis, although it yielded nonsignificant effect on VEGF/VEGFR2 expression and angiogenesis. The levels of IL-1β/TNF-α/IFN-γ were moderately attenuated by Sarp plus mADSCs transplantation. IL-1β/TNF-α/IFN-γ is known to contribute to cell apoptosis and dysfunction (Suzuki et al., 2004; Ishii et al., 2010). TNF-α/IFN-γ severely undermines the functional survival of donor stem cells (Liu et al., 2011). Sarp further promoted the expression of IL-10/IL-6. Although the debate of IL-6 as pro- or anti-inflammatory cytokine is not fully elucidated, a shred of evidences show that IL-6 may promote cell survival and prevent pro-inflammatory cytokines-induced cell death (Ahmed & Ivashkiv, 2000). Moreover, single Sarp treatment also recovered blood perfusion and ambulation of ischemic hindlimb through angiogenesis-independent mechanism. The therapeutic benefit of Sarp possibly derived from its vasodilation stimulation efficacy (Tanaka et al., 1998).

It has been demonstrated that 5-HT(2A) participates in vascular inflammation and cytokine production process, while selective 5-HT(2A) antagonism using Sarp is determined to exert anti-apoptotic and cardiovascular protective effects (Rajesh et al., 2006). Our current experiment further demonstrated mTOR/STAT3 pathway was involved in the inflammatory regulation mechanism of Sarp. Recent data show that mTOR orchestrates pro-/anti-inflammatory cytokines and plays a role in stem cell survival (Laplante & Sabatini, 2012). STAT3 can promote IL-6/IL-10 expression, which further activates STAT3 in a feedforward manner (Ahmed & Ivashkiv, 2000). Our recent work demonstrated that mTOR/STAT3 could mediate cardioprotection (Ge & Ren, 2012). In the present study, activation of mTOR/STAT3 was necessary for the mADSCs' survival in aged CLI mice. Sarp treatment upregulated phosphorylation of mTOR/STAT3 and elicited anti-inflammatory reaction, which prevented mADSCs' death following transplantation in vivo and over hypoxia/reoxygenation in vitro. Inhibition of mTOR or STAT3 exaggerated pro-inflammatory response and counteracted the cell-protective effect of Sarp. Sarp-induced mTOR/STAT3 activation could also mediate inflammatory homeostasis, which is essential to maintain moderate inflammation-induced angiogenesis in aged CLI mice. Overall, Sarp plus mADSCs transplantation augmented cell-derived therapeutic angiogenesis and attenuated tissue loss, which ultimately benefited in improved prognosis of aged CLI mice. The putative mechanism involved in sarpogrelate plus mADSCs therapy for aged CLI in this study was illustrated in Fig. 6.

Figure 6.

Mechanism involved in sarpogrelate plus mADSCs combination therapy for aged hindlimb ischemia in this study. Pathways in the outer rounded rectangle represent the putative regulatory mechanism of ‘inflammation-induced angiogenesis’ by sarpogrelate plus mADSCs therapy. Engrafted mADSCs triggered inflammatory cytokines (inner rounded rectangle) and angiogenesis partly via pathways regulated by the VEGF/VEGFR2 and mTOR/STAT3. mTOR/STAT3 pathway was also permissive to sarpogrelate-mediated inflammatory modulation for cell survival and anti-apoptosis. These pathways can be manipulated to inhibit by the experimental interventions in the boxes. The figure only shows selected pathways and is simplified for clarity. ADSC, adipose-derived stromal cell; mTOR, mammalian target of rapamycin; VEGF, vascular endothelial growth factor; VEGFR2, vascular endothelial growth factor receptor-2.

In conclusion, the present study has provided in vivo evidence of mADSC-directed therapeutic angiogenesis in an aging PAD model. mADSCs and sarpogrelate may synergistically rejuvenate the inflammation-induced angiogenesis, which benefits in limb salvage for the aged CLI mice. As the field of cell-based PAD therapy continues to mature, inflammation-induced angiogenesis would be more effectively intervened using cautious and optimized strategies, which can bring more benefits to the elderly PAD patients.

Experimental procedures

Animals

Male Vegfr2-luc mice [FVB/N-Tg(Vegfr2-luc)-Xen, FVB/N background, n = 240; Caliper Life Sciences, Hopkinton, MA, USA], which carry a transgene containing 4.5 kb murine VEGFR2 promoter and a modified Fluc DNA (pGL-3), and FVB/N mice [Tg(Vegfr2-luc) inbred strain, no luciferase expression, n = 240] were grouped based on age including the following: young (8–10 weeks old, 22–28 g, specific pathogen free, SPF) and aged (18 months old, 22–28 g, SPF). Young (8–10 weeks old, 22–28 g, SPF) ß-Actin-luc transgenic mice [FVB/N-Tg(ß-Actin-luc)-Xen, FVB/N background; Caliper Life Sciences] and Cag-egfp transgenic mice [FVB.Cg-Tg(Cag-egfp)B5Nagy/J:003516, FVB/N background; The Jackson Laboratory, Bar Harbor, ME, USA], which constitutively express Fluc or eGFP in all tissues and organs, were used for mADSCs isolation. Animal ethics and anesthesia are described in the Data S1 (Supporting Information).

Cell culture and identification

mADSCsGFP+ and mADSCsFluc+ were isolated from Tg(Cag-egfp) and Tg(ß-Actin-luc) mice, respectively. Methods for cell culture, immunophenotype identification, and multilineage differentiation are described in the Data S1 (Supporting Information).

CLI model and cell delivery

Aged Tg(Vegfr2-luc) mice (= 200) were randomized into five groups (40 each matched for weight): (i) CLI + PBS (Aged); (ii) CLI + Sarp (Aged + Sarp); (iii) CLI + ADSCs (Aged + ADSCs); (iv) CLI + ADSCs + Sarp (Aged + Dual); (v) Sham. Young Tg(Vegfr2-luc) mice (= 40) were treated with CLI + PBS (Young) as control. For in vivo cell tracking investigation, aged FVB/N mice (= 240) were randomized into eight groups (30 each matched for weight): (i) CLI + PBS (PBS); (ii) CLI + ADSCs (ADSCs); (iii) CLI + ADSCs + Sarp (ADSCs + Sarp); (iv) CLI + ADSCs + PP242 (ADSCs + PP242); (v) CLI + ADSCs + Sarp + PP242 (ADSCs + Sarp + PP242); (vi) CLI + ADSCs + S3I-201 (ADSCs + S3I); (vii) CLI + ADSCs + Sarp + S3I-201 (ADSCs + Sarp + S3I); (viii) Sham. CLI model was constructed using ligation and excision of left femoral artery with all superficial/deep branches. Surgical procedure is described in the Data S1 (Supporting Information). Sham-operated mice received incision without artery ligation. Tg(Vegfr2-luc) and FVB/N mice were respectively subjected to mADSCsGFP+/mADSCsFluc+ delivery on POD1. Cells (3rd passage, 1.0 × 107) suspended in 30-μL phosphate-buffered saline (PBS) were injected into left gastrocnemius using a 29-gauge insulin syringe (BD Biosciences, San Jose, CA, USA), while other groups received 30-μL PBS only. Sarp-treated mice received a supplemented diet containing sarpogrelate hydrochloride (5 mg kg−1 day−1). Inhibitor administration was performed using intramuscular injection of PP242 (200 nm kg−1 day−1; Sigma-Aldrich, St. Louis, MO, USA) or S3I-201 (sc-204304, 100 nkg−1 day−1; Santa Cruz Biotechnology, Santa Cruz, CA, USA) into left gastrocnemius.

Reporter gene imaging and assays

Reporter gene imaging was performed to monitor Vegfr2-luc transcription level and mADSCs survival in vivo (Cao et al., 2006; Fan et al., 2012). For in vitro BLI, mADSCsFluc+ of different quantities were respectively incubated with the probe D-luciferin (150 ng μL−1; Invitrogen, Carlsbad, CA, USA), and sequentially imaged using a charge-coupled device (CCD, dual-modality) camera within Xenogen in vivo Imaging System (IVIS; Caliper Life Sciences), with the following parameters: binning: 4, F/Stop: 1, exposure time: 1 min. For in vitro FRI, mADSCsGFP+ were directly imaged by CCD with its excitation wavelength at 465/430 nm and emission filter at 560 nm. Wild-type mADSCs (mADSCsWT) were used as negative control. For in vivo BLI, mice were anesthetized and injected intraperitoneally with 150 mg kg−1 D-luciferin. BLI was conducted at 3-min intervals until the peak signal was observed. Peak BLI signal was quantified by average radiance (photons second−1cm−2 steridian−1, P s−1 cm−2 sr−1) from a fixed-area region of interest (ROI) over left hindlimb using livingimage 4.2 (Caliper Life Sciences). Methods for luciferase assay were described in the Data S1 (Supporting Information).

Serial peripheral perfusion imaging and functional assessment

The LDPI was used to serially monitor hindlimb perfusion recovery after ischemia. Mice were placed on a 37.4–38.0 °C heating pad to minimize temperature variation, and then imaged using LDPI analyzer (PeriScan-PIM3; Perimed AB, Järf-E4lla, Sweden). The blood flux was quantified using perfusion ratio (PR) [ratio of average LDPI index of ischemic/nonischemic (contralateral, self-control) hindlimb] by LDPIwin3.1.3 (Perimed AB). A cohort of mice in Aged or Aged + ADSCs group were subjected to LDPI following intramuscular injection of VEGF neutralizing monoclonal antibody (anti-VEGF mAb, = 10, 2.0 μg kg−1 day−1; R&D system, Minneapolis, MN, USA) or Ki8751 (sc-203090, = 10, 100 nkg−1 day−1; Santa Cruz Biotechnology) into gastrocnemius till POD21. Nonspecific isotype IgG or vehicle were administered as control (= 5, respectively). Semi-quantitative functional scoring of ischemic hindlimbs was performed as described in the Data S1 (Supporting Information).

Histological analysis of angiogenesis

Immunohistochemistry analysis was performed to visualize CD31+ vessels on POD1/POD14. Immunofluorescence assay were further adopted to display mADSCsGFP+ and vWF+ microvasculature on POD14. Histological analysis were performed as described in the Data S1 (Supporting Information).

In vitro hypoxia/reoxygenation model of mADSCs

In vitro hypoxia/reoxygenation (H/R) model was performed to further evaluate mADSCs' survival in temporary anoxic condition. mADSCs were uniformly plated in 24-well plates (5.0 × 104 cells per well) and randomized into seven groups: (i) Normoxia (Control); (ii) H/R + common medium (H/R); (iii) H/R + Sarp (5 nm mL−1); (iv) H/R + PP242 (4 nm mL−1); (v) H/R + Sarp + PP242; (vi) H/R + S3I-201 (2 nm mL−1); (vii) H/R + Sarp + S3I-201. To mimic hypoxia/reoxygenation, the cells were incubated in anoxic chamber (95%N2/5%CO2) at 37 °C for 21 h, and subsequently moved into a normoxic incubator (95% air/5%CO2) at 37 °C for 3 h. The control group was cultured in standard conditions for 24 h.

Apoptosis assessment in vivo and in vitro

Left gastrocnemius muscle tissues or cultured mADSCs following hypoxia/reoxygenation treatment were collected for TUNEL apoptotic assay using a TUNEL Apoptosis Assay Kit (Promega, Madison, WI, USA) according to the manufacturer's instructions. Sections were further stained with 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen) for total nuclei quantification. Apoptosis Index, that is, the number of TUNEL-positive cells divided by the total cells per field, was assessed in 20 randomly selected fields.

Analysis of signal pathway and cytokines

Left gastrocnemius tissues or cultured mADSCs were harvested and pulverized to extract RNA or protein for RT-PCR, Western blotting and ELISA. The protocol was described in the Data S1 (Supporting Information).

Statistics

Results are expressed as mean ± standard deviation (SD). spss 17.0 (SPSS Inc., Chicago, IL, USA) and prism 5.0 (GraphPad Software, La Jolla, CA, USA) were used to perform the one-way analysis of variance (anova) for evaluating the differences of bioluminescence radiance, LDPI index, semi-quantitative scores, vascular density, apoptosis index, cell number, integrated optical density, and cytokines concentration, among different experimental groups and different time points within each group. Pairwise multiple comparisons were to identify the parameters differences between two groups using anova-conjuncted Tukey's test. A two-tailed P-value < 0.05 was considered significant. Polynomial regression analysis was performed to evaluate the correlation between cell numbers and optical radiance in vitro.

Acknowledgments

This work was supported by National Basic Research Program of China [2012CB518101]; National Natural Science Foundation of China [No. 30970845, No. 81090274, No. 81090270, No. 81130072]; FCao [BWS12J037]; Inovation Team Development Grant by China Department of Education [2010CXTD01]; China's Ministry of Science and Technology 863 Program [2012AA02A603].

Conflict of interest

The authors declare no conflict of interest.

Author contributions

The specific contribution of each author is listed below: Weiwei Fan for acquiring data and writing the manuscript; Shenxu Wang and Hu Da for acquiring data; Xing Qin, Kang Cheng, Ri Zhou and Chao Tong for analyzing data; Xiujuan Li, Congye Li, Qingting Bu and Yaling Han for critically revising the manuscript; Jun Ren for enhancing its intellectual content; Feng Cao and Chengxiang Li for conception, design, and final approval of manuscript. Weiwei Fan and Chengxiang Li contributed equally to this article.

Ancillary