Hepatocyte Growth Factor, but not Vascular Endothelial Growth Factor, Attenuates Angiotensin II–Induced Endothelial Progenitor Cell Senescence
Although both hepatocyte growth factor (HGF) and vascular endothelial growth factor (VEGF) are potent angiogenic growth factors in animal models of ischemia, their characteristics are not the same in animal experiments and clinical trials. To elucidate the discrepancy between HGF and VEGF, we compared the effects of HGF and VEGF on endothelial progenitor cells under angiotensin II stimulation, which is a well-known risk factor for atherosclerosis. Here, we demonstrated that HGF, but not VEGF, attenuated angiotensin II–induced senescence of endothelial progenitor cells through a reduction of oxidative stress by inhibition of the phosphatidylinositol-3,4,5-triphosphate/rac1 pathway. Potent induction of neovascularization of endothelial progenitor cells by HGF, but not VEGF, under angiotensin II was also confirmed by in vivo experiments using several models, including HGF transgenic mice.
Critical limb ischemia is estimated to develop in 500 to 1000 individuals per million per year.1 In a large proportion of these patients, the anatomic extent and distribution of arterial occlusive disease make the patients unsuitable for operative or percutaneous revascularization, and there is no optimal medical therapy for such critical limb ischemia.1 Therefore, therapeutic angiogenesis using angiogenic growth factors, such as vascular endothelial growth factor (VEGF),2 has been investigated. Although initial open-label clinical trials using VEGF were successful, double-blind, placebo-controlled clinical trials failed to show efficacy in peripheral arterial disease (PAD) patients.3 On the other hand, recent clinical trials using hepatocyte growth factor (HGF) plasmid DNA demonstrated a marked difference between HGF and VEGF, because there was no evidence of edema in patients transfected with the human HGF gene, in marked contrast to a VEGF trial in which 60% of patients developed moderate or severe edema. More recently, a multicenter, double-blind, placebo-controlled phase III clinical trial using HGF plasmid DNA in Japan demonstrated a significant improvement in primary end points as compared with placebo (70.4% versus 30.8%; P=0.014). In addition, a multicenter, double-blind, placebo-controlled phase II US clinical trial using HGF plasmid DNA demonstrated a significant increase in transcutaneous partial pressure of oxygen as compared with placebo.4 To elucidate the molecular difference between HGF and VEGF, we focused on the effects of HGF and VEGF on endothelial progenitor cells (EPCs) under angiotensin (Ang) II stimulation, because EPCs play a significant role in the neovascularization of ischemic tissue. However, their functional activity was markedly impaired in patients with PAD and coronary artery disease5,6 in whom the local renin-angiotensin system was activated. Because the number of circulating EPCs was negatively correlated with risk factors for PAD, such as hypertension7 and diabetes mellitus,8 these risk factors would shorten the average life span of EPCs, leading to impairment of neovascularization. Here, we demonstrated that HGF, but not VEGF, attenuated Ang II–induced senescence of EPCs through a reduction of oxidative stress by inhibition of the PIP3/rac1 pathway.
Materials and Methods
Please see the online supplemental data (http://hyper.ahajournals.org) for information on the primary antibodies and reagents used in the study.
Cell Isolation and Culture
EPCs were isolated from human peripheral blood using density gradient centrifugation with Histopaque 1077 (Sigma Chemical Co). After 24-hour starvation, several concentrations of Ang II, with or without pretreatment for 3 hours with HGF (100 ng/mL), VEGF (100 ng/mL), or superoxide dismutase (SOD; 50 U/mL), were added every 48 hours starting on day 8. The protocol of the in vitro study is shown in Figure S1A. Please see the online supplemental data for a complete description of cell isolation and culture.
HGF Transgenic Mice
We generated mice (C57BL6 background) with cardiac-specific (α-major histocompatibility complex driven) overexpression of HGF in which serum HGF derived from the heart was significantly increased (Please see the online supplemental data about the characteristics of HGF transgenic [Tg] mice).
Ischemic Hindlimb Animal Model and Bone Marrow Transplantation Model
Please see the online supplemental data for a complete description of the ischemic hindlimb model, bone marrow transplantation, and tissue samples used for these studies.
Protein analysis using Western blotting was performed as described previously.9
Values are expressed as means±SEs. ANOVA and t test, followed by Bonferroni adjustment for multiple comparisons, were used for comparisons of >2 groups. A P value of <0.05 was considered to indicate significance of mean differences.
Antisenescent Action of HGF
Initially, we examined the effects of HGF and VEGF on the growth of EPCs under continuous Ang II stimulation. Consistent with previous reports,10 continuous stimulation with Ang II significantly reduced the number of EPCs as compared with control, whereas HGF and VEGF significantly stimulated the proliferation of EPCs (Figure 1A). Addition of either HGF or VEGF attenuated the decrease in EPC number induced by Ang II. Then, we focused on the effects of HGF and VEGF on Ang II–induced senescence, because Ang II is known to promote senescence of EPCs mainly through oxidative stress.11 Interestingly, HGF and SOD, but not VEGF, significantly attenuated Ang II–induced senescence of EPCs as determined by senescence associated β-galactosidase staining (Figure 1B), although HGF and VEGF alone promoted EPC senescence. Moreover, HGF and SOD significantly inhibited the induction of p53, p21Cip1, and p16INK4a by Ang II (Figure S1B), which induces cell premature senescence through oxidative stress.12
Next, we examined the function of EPCs. As show in Figure S2A, S2B, and S2C, the incorporation ability into tube-like structures of human umbilical cord endothelial cells and proliferative activity were significantly decreased in the Ang II single-treated group. However, pretreatment with HGF and SOD, but not VEGF, significantly attenuated the decrease in their incorporation and proliferation.
To elucidate the role of oxidative stress, the production of superoxide was measured by dihydroethidium staining. Interestingly, the production of superoxide induced by Ang II was significantly reduced by pretreatment with HGF and SOD, but not VEGF, whereas HGF and VEGF alone significantly increased superoxide production (Figure 2A and 2B). The antioxidative effect of HGF was also confirmed in bovine aortic endothelial cells (Figure S3A). In addition, pretreatment with HGF and SOD significantly reduced the Ang II–induced gp91phox expression (Figure S3B), which is one of the components of NADPH oxidase and known to be upregulated by Ang II in EPCs.11 Although a previous report shows that VEGF accelerates EPC incorporation into the tube-like structure of endothelial cells in the acute phase,13 our data clearly showed that, because under basal conditions HGF and VEGF act as pro-oxidants, chronic stimulation with HGF and VEGF caused the decrease in EPC function. However, in the presence of Ang II, HGF, but not VEGF, can act as an antioxidant.
Upstream Molecular Mechanisms of Inhibition of Oxidative Stress by HGF
The production of reactive oxygen species (ROS) by Ang II is rac1 dependent in EPCs, because dominant-negative rac1 transfection reduced Ang II–induced ROS production (Figure 2A). Ang II increased GTP-rac1 with 2 peaks (at 10 and 60 minutes; data not shown), whereas pretreatment with HGF, but not VEGF, prevented Ang II–induced rac1 binding to GTP at 60 minutes (Figure 3A). HGF also prevented Ang II–induced rac1 translocation from the cytosol to the plasma membrane (Figure S4A). The product of phosphatidylinositol 3-kinase activity, PI (3, 4, and 5) P3 (PIP3), is required to activate a number of rac-specific guanine nucleotide exchangers in endothelial cells.14 Indeed, phosphatidylinositol 3-kinase–specific inhibitors LY294002 and wortmannin, but not the P38 mitogen-activated protein kinase inhibitor SB203580 or the mitogen-activated protein kinase kinase inhibitor PD98059, significantly reduced rac1 binding to GTP induced by Ang II (P<0.01 versus Ang II; Figure 3B). Then we examined the PIP3 production, and, finally, we found that pretreatment with HGF significantly inhibited Ang II–induced PIP3 production, whereas VEGF did not inhibit it (Figure 3C). Previous reports show that Ang II-15 and VEGF-16 induced phosphorylation of Akt depends on ROS; on the other hand, HGF-induced phosphorylation of Akt is independent of ROS.16 Consistently, pretreatment with HGF, but not VEGF, significantly inhibited Ang II–induced phosphorylation of Akt in EPCs (Figure S4B).
As the end point of the in vitro experiment, we focused on endothelial cell tube formation under continuous Ang II stimulation. Ang II alone had a modest effect on tube formation of HUVECs after 12 hours, but the formed tubes were collapsed compared with control after 72 hours. However, interestingly, pretreatment with HGF, but not VEGF, decreased their collapse, and maintained tube formation even after 72 hours (Figure 4A and 4B).
Comparison of Neovascularization in Hindlimb Ischemia Nude Mouse Model
To confirm the function of ex vivo expanded human EPCs in vivo stimulated with in vitro study protocol, incorporation of fluorescent carbocyanine1,1′-dioctadecyl-1 to 3,3,3′3′-tetramethylidocarboyanine perchlorate (DiI) labeled human EPCs and their contribution to host neovascularization were investigated in a nude mouse hindlimb ischemia model. At 7 days after induction of ischemia, the Ang II–treated EPC group showed significantly decreased incorporation compared with the control group, whereas the Ang II+HGF group, but not the Ang II+VEGF group, showed significant attenuation of the decrease (Figure 5A and 5B). Moreover, the Ang II–treated EPC group showed significant attenuation of the increase in host mouse capillary density compared with the control group, whereas the Ang II+HGF group, but not the Ang II+VEGF group, showed a significant increase. These in vitro and in vivo data showed that HGF reduced Ang II–induced senescence of EPCs through an antioxidative effect, which maintained the function of neovascularization.
Ang II Infusion Into HGF Transgenic Mice
To further investigate the effect of HGF on Ang II–induced senescence in EPCs, we made Tg mice with cardiac-specific overexpression of human HGF in which the blood level of the human HGF protein was significantly increased (Figure S5A). We administered Ang II (0.7 mg/kg per day) to HGF Tg mice and wild-type (WT) mice. There was no significant difference in blood pressure between WT and HGF Tg mice after Ang II infusion (Figure S5B). After infusion of Ang II, peripheral blood–derived (Figure S6A) and bone marrow–derived (Figure S6B) monocytic EPCs were significantly decreased in WT mice as compared with HGF Tg mice. The rate of senescence was significantly reduced in HGF Tg mice as compared with WT mice at 4 weeks (Figure S6C). Moreover, as show in Figure S6D, HGF Tg mice showed a significant reduction of Ang II infusion–induced gp91 phox protein expression in total bone marrow mononuclear cells as compared with WT mice (P<0.01).
Based on these results, we finally used a hindlimb ischemia model after Ang II infusion for 1 week. Blood flow ratio and the relative increase in capillary density in Ang II–treated HGF Tg mice were significantly increased as compared with those in Ang II–treated WT mice (Figure S7A, S7B, S7C, and S7D). Attenuation of the impairment of neovascularization by Ang II in HGF Tg mice was accompanied by a significant increase in the number of bone marrow–derived EPCs, whereas Ang II infusion significantly decreased the number in WT mice (Figure S7E and S7F).
Both HGF and VEGF are potent angiogenic growth factors in animal models of ischemia, but their characteristics are not the same in animal experiments and clinical trials. A multicenter, double-blind, placebo-controlled phase III clinical trial in Japan and a US phase II clinical trial of HGF gene therapy for PAD demonstrated a significant improvement in primary end points or an increase in transcutaneous partial pressure of oxygen even after 1 year compared with placebo,4 whereas effectiveness of VEGF treatment for PAD has not yet been shown. On the other hand, Asahara et al17 discovered that adult bone marrow–derived EPCs played a significant role in neovascularization of ischemic tissue. Since then, many researchers have considered the potential use of EPCs for cell therapy to augment neovascularization.18 However, atherosclerosis risk factors are known to negatively correlate with the number and function of EPCs6; moreover, ex vivo cultivated EPCs exhibit rapid onset of senescence because of very low proliferative capacity and activity of telomerase.19,20 To consider their clinical application, enhancement of EPC activity under such high-risk conditions might be necessary. Thus, we focused on the effects of HGF and VEGF on EPC function under continuous Ang II stimulation to clarify the discrepancy between animal experiments and clinical trials of HGF and VEGF.
In this study, we clearly demonstrated that Ang II accelerated EPC senescence through oxidative stress, as reported previously,11 which aggravated the impairment of EPC function in vivo and in vitro. Ang II alone had a modest effect on tube formation of HUVECs in the short term (12 hours); moreover, Ang II+VEGF showed the tendency to form tubular formation compared with VEGF alone after 12 hours, as described previously,13 but the formed tubes collapsed compared with control in the long term (72 hour). We also demonstrated that, under basal conditions, HGF and VEGF significantly stimulated proliferation of EPCs and tube formation of HUVECs (12 hours), but they also increased ROS production, leading to replicative senescence and ROS-induced premature senescence of EPC. Interestingly, finally, we also found that pretreatment with HGF, but not VEGF, significantly reduced Ang II–induced ROS and senescence in EPCs through a reduction of Ang II–induced PIP3/rac1 activity, leading to maintained function of EPCs and tube formation of HUVECs, even after 72 hours.
Consistent with these findings, pretreatment with HGF, but not VEGF, decreased Ang II–induced Akt phosphorylation (Figure S4B). As for cell senescence and Akt activation, Miyauchi et al21 showed that Akt activity increased with cellular senescence, and inhibition of Akt extended the life span of endothelial cells; thus, constitutive activation of Akt might promote senescence.
Recently, it has become more clear that VEGF is not only an angiogenic growth factor but also a proinflammatory cytokine.22,23 VEGF leads to increased monocyte adhesion to endothelial cells, which is mediated by ROS (rac1)-nuclear factor κB activation24,25 and promotes atherosclerotic plaque formation without increased microvessel development.26 From this viewpoint, HGF is unique, because HGF is known to act as an anti-inflammatory cytokine22,27 and as an antioxidative stress cytokine.28 Min et al22 showed that HGF prevented the increase in expression of intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 induced by VEGF, through inhibition of nuclear factor κB activation. The present study also indicates the possibility that HGF might inhibit VEGF-induced inflammation through a reduction of oxidative stress.
We showed different effects of HGF and VEGF on PIP3/rac1 activity under Ang II stimulation but could not show its upstream mechanism. Recent articles revealed that SH2 domain–containing inositol 5-phosphatase binds to the epidermal growth factor receptor, which contributes to PIP3/Akt activation, and SH2 domain–containing inositol 5-phosphatase also binds to c-Met directly and is involved in HGF-induced cell scattering. In addition, Ang II–induced rac1 activation and ROS production are mediated by epidermal growth factor receptor.29 These reports and our experiment suggest the possibility that translocation of SH2 domain–containing inositol 5-phosphatase between c-Met and epidermal growth factor receptor controls their signaling and ROS production. Careful research is necessary to clarify this mechanism in the future.
Certainly, our experiment demonstrated that HGF, but not VEGF, significantly attenuated the Ang II–induced impairment of neovascularization by reducing oxidative stress in EPCs. On the other hand, Ushio-Fukai and colleagues30,31 reported that neovascularization in the ischemic hindlimb is significantly impaired in gp91phox−/− mice, and the ROS-generating agent LY83583 improved their neovascularization. This and our experiment suggest that conditions of excess or insufficient ROS may also impair neovascularization.
In this experiment, to reproduce the conditions of a clinical trial, we compared the effect of high-dose HGF (100 ng/mL) and VEGF (100 ng/mL) on neovascularization. However, Ehrbar et al32 showed that controlling the level of VEGF with special delivery systems (α2PI1-8-VEGF121) results in nonleaky vessels and increased vessel formation more potently than did native VEGF121. We should consider the possibility that slow-release, low-dose VEGF might be effective in clinical trials.
Overall, the present study demonstrated that HGF, but not VEGF, attenuated senescence induced by Ang II in EPCs. These protective effects of HGF on senescence and oxidative stress in EPCs and endothelial cells might contribute to the clinical use of therapeutic angiogenesis using the HGF gene observed in clinical trials. The present study suggests ideal features of future angiogenesis therapy using angiogenic growth factors. In the future, HGF gene therapy could be applied for ischemic heart disease.
We thank Aiko Kikuchi for her expert technical assistance.
Sources of Funding
This work was partially supported by a Grant-in-Aid from the Organization for Pharmaceutical Safety and Research, a Grant-in-Aid from the Ministry of Public Health and Welfare, a Grant-in-Aid from Japan Promotion of Science, and through special coordination funds of the Ministry of Education, Culture, Sports, Science and Technology, of the Japanese government.
- Received July 30, 2008.
- Revision received August 24, 2008.
- Accepted November 4, 2008.
Baumgartner I, Pieczek A, Manor O, Blair R, Kearney M, Walsh K, Isner JM. Constitutive expression of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation. 1998; 97: 1114–1123.
Rajagopalan S, Mohler E III, Lederman RJ, Saucedo J, Mendelsohn FO, Olin J, Blebea J, Goldman C, Trachtenberg JD, Pressler M, Rasmussen H, Annex BH, Hirsch AT. Regional angiogenesis with vascular endothelial growth factor (VEGF) in peripheral arterial disease: design of the RAVE Trial. Am Heart J. 2003; 145: 1114–1118.
Powell RJ, Simons M, Mendelsohn FO, Daniel G, Henry TD, Koga M, Morishita R, Annex BH. Results of a double-blind, placebo-controlled study to assess the safety of intramuscular injection of hepatocyte growth factor plasmid to improve limb perfusion in patients with critical limb ischemia. Circulation. 2008; 118: 58–65.
Yu Y, Fukuda N, Yao EH, Matsumoto T, Kobayashi N, Suzuki R, Tahira Y, Ueno T, Matsumoto K. Effects of an ARB on endothelial progenitor cell function and cardiovascular oxidation in hypertension. Am J Hypertens. 2008; 21: 72–77.
Tepper OM, Galiano RD, Capla JM, Kalka C, Gagne PJ, Jacobowitz GR, Levine JP, Gurtner GC. Human endothelial progenitor cells from type II diabetics exhibit impaired proliferation, adhesion, and incorporation into vascular structures. Circulation. 2002; 106: 2781–2786.
Iekushi K, Taniyama Y, Azuma J, Katsuragi N, Dosaka N, Sanada F, Koibuchi N, Nagao K, Ogihara T, Morishita R. Novel mechanisms of valsartan on the treatment of acute myocardial infarction through inhibition of the antiadhesion molecule periostin. Hypertension. 2007; 49: 1409–1414.
Chen J, Goligorsky MS. Premature senescence of endothelial cells: Methusaleh’s dilemma. Am J Physiol Heart Circ Physiol. 2006; 290: H1729–H1739.
Taniyama Y, Ushio-Fukai M, Hitomi H, Rocic P, Kingsley MJ, Pfahnl C, Weber DS, Alexander RW, Griendling KK. Role of p38 MAPK and MAPKAPK-2 in angiotensin II-induced Akt activation in vascular smooth muscle cells. Am J Physiol Cell Physiol. 2004; 287: C494–C499.
Abid MR, Spokes KC, Shih SC, Aird WC. NADPH oxidase activity selectively modulates vascular endothelial growth factor signaling pathways. J Biol Chem. 2007; 282: 35373–35385.
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Lunde K, Solheim S, Aakhus S, Arnesen H, Abdelnoor M, Egeland T, Endresen K, Ilebekk A, Mangschau A, Fjeld JG, Smith HJ, Taraldsrud E, Grogaard HK, Bjornerheim R, Brekke M, Muller C, Hopp E, Ragnarsson A, Brinchmann JE, Forfang K. Intracoronary injection of mononuclear bone marrow cells in acute myocardial infarction. N Engl J Med. 2006; 355: 1199–1209.
Feugier P, Jo DY, Shieh JH, MacKenzie KL, Rafii S, Crystal RG, Moore MA. Ex vivo expansion of stem and progenitor cells in co-culture of mobilized peripheral blood CD34+ cells on human endothelium transfected with adenovectors expressing thrombopoietin, c-kit ligand, and Flt-3 ligand. J Hematother Stem Cell Res. 2002; 11: 127–138.
Assmus B, Urbich C, Aicher A, Hofmann WK, Haendeler J, Rossig L, Spyridopoulos I, Zeiher AM, Dimmeler S. HMG-CoA reductase inhibitors reduce senescence and increase proliferation of endothelial progenitor cells via regulation of cell cycle regulatory genes. Circ Res. 2003; 92: 1049–1055.
Miyauchi H, Minamino T, Tateno K, Kunieda T, Toko H, Komuro I. Akt negatively regulates the in vitro lifespan of human endothelial cells via a p53/p21-dependent pathway. EMBO J. 2004; 23: 212–220.
Min JK, Lee YM, Kim JH, Kim YM, Kim SW, Lee SY, Gho YS, Oh GT, Kwon YG. Hepatocyte growth factor suppresses vascular endothelial growth factor-induced expression of endothelial ICAM-1 and VCAM-1 by inhibiting the nuclear factor-kappaB pathway. Circ Res. 2005; 96: 300–307.
Abid MR, Schoots IG, Spokes KC, Wu SQ, Mawhinney C, Aird WC. Vascular endothelial growth factor-mediated induction of manganese superoxide dismutase occurs through redox-dependent regulation of forkhead and IkappaB/NF-kappaB. J Biol Chem. 2004; 279: 44030–44038.
Lucerna M, Zernecke A, de Nooijer R, de Jager SC, Bot I, van der Lans C, Kholova I, Liehn EA, van Berkel TJ, Yla-Herttuala S, Weber C, Biessen EA. Vascular endothelial growth factor-A induces plaque expansion in ApoE knock-out mice by promoting de novo leukocyte recruitment. Blood. 2007; 109: 122–129.
Gong R, Rifai A, Dworkin LD. Anti-inflammatory effect of hepatocyte growth factor in chronic kidney disease: targeting the inflamed vascular endothelium. J Am Soc Nephrol. 2006; 17: 2464–2473.
Seshiah PN, Weber DS, Rocic P, Valppu L, Taniyama Y, Griendling KK. Angiotensin II stimulation of NAD(P)H oxidase activity: upstream mediators. Circ Res. 2002; 91: 406–413.
Ushio-Fukai M, Tang Y, Fukai T, Dikalov SI, Ma Y, Fujimoto M, Quinn MT, Pagano PJ, Johnson C, Alexander RW. Novel role of gp91(phox)-containing NAD(P)H oxidase in vascular endothelial growth factor-induced signaling and angiogenesis. Circ Res. 2002; 91: 1160–1167.
Tojo T, Ushio-Fukai M, Yamaoka-Tojo M, Ikeda S, Patrushev N, Alexander RW. Role of gp91phox (Nox2)-containing NAD(P)H oxidase in angiogenesis in response to hindlimb ischemia. Circulation. 2005; 111: 2347–2355.
M, Djonov VG, Schnell C, Tschanz SA, Martiny-Baron G, Schenk U, Wood J, Burri PH, Hubbell JA, Zisch AH. Cell-demanded liberation of VEGF121 from fibrin implants induces local and controlled blood vessel growth. Circ Res. 2004; 94: 1124–1132.