Age Decreases Endothelial Progenitor Cell Recruitment Through Decreases in Hypoxia-Inducible Factor 1α Stabilization During Ischemia
Background— Advanced age is known to impair neovascularization. Because endothelial progenitor cells (EPCs) participate in this process, we examined the effects of aging on EPC recruitment and vascular incorporation.
Methods and Results— Murine neovascularization was examined by use of an ischemic flap model, which demonstrated aged mice (19 to 24 months) had decreased EPC mobilization (percent mobilized 1.4±0.2% versus 0.4±0.1%, P<0.005) that resulted in impaired gross tissue survival compared with young mice (2 to 6 months). This decrease correlated with diminished tissue perfusion (P<0.005) and decreased CD31+ vascular density (P<0.005). Gender-mismatched bone marrow transplantation demonstrated significantly fewer chimeric vessels in aged mice (P<0.05), which confirmed a deficit in bone marrow–mediated vasculogenesis. Age had no effect on total EPC number in mice or humans. Reciprocal bone marrow transplantations confirmed that impaired neovascularization resulted from defects in the response of aged tissue to hypoxia and not from intrinsic defects in EPC function. We demonstrate that aging decreased hypoxia-inducible factor 1α stabilization in ischemic tissues because of increased prolyl hydroxylase–mediated hydroxylation (P<0.05) and proteasomal degradation. This resulted in a diminished hypoxia response, including decreased stromal cell–derived factor 1 (P<0.005) and vascular endothelial growth factor (P<0.0004). This effect can be reversed with the iron chelator deferoxamine, which results in hypoxia-inducible factor 1α stabilization and increased tissue survival.
Conclusions— Aging impairs EPC trafficking to sites of ischemia through a failure of aged tissues to normally activate the hypoxia-inducible factor 1α–mediated hypoxia response.
Received May 16, 2007; accepted September 28, 2007.
Advanced age has been associated with a decreased ability to form new blood vessels in response to ischemia,1 which results in higher rates of cardiovascular complications and diminished capacity for tissue regeneration.2 Whether this is due to an intrinsic decline in the regenerative capacity of putative vascular progenitors or a decline in a proregenerative niche remains unclear. Adult neovascularization occurs by 2 distinct processes: angiogenesis (the sprouting of new blood vessels from preexisting ones) and vasculogenesis (the recruitment, proliferation, and assembly of bone marrow–derived endothelial progenitor cells [EPCs] into new vessels).3 HIF-1α (hypoxia-inducible factor 1α) is the transcription factor known to regulate both of these processes through mediators such as vascular endothelial growth factor (VEGF) and stromal cell–derived factor-1 (SDF-1α).4,5
Clinical Perspective p 2829
It has been postulated that aging results in a decline in progenitor cell function, also known as progenitor exhaustion.6,7 Current investigational stem cell therapies are based on replenishing this depleted supply of functional progenitor cells to areas of injury. Once delivered, it is hoped that progenitor cells will engraft at sites of injury and differentiate into native cell populations to regenerate tissue; however, clinical trials attempting to replenish progenitor cells after myocardial infarction have been disappointing and have often failed to maintain any lasting benefit.8–10 The inability to retain progenitor cells in the injured myocardium10 suggests that administration of progenitor cells alone is insufficient for tissue repair.
Thus, we hypothesized that defects in neovascularization are not due to a lack of functional EPCs but to the body’s diminished ability to recruit EPCs to sites of injury. We have previously shown that SDF-1α plays an important role in the recruitment of EPCs to sites of ischemia.4 Given that hematopoietic, endothelial, neural, and skeletal/smooth muscle progenitor cells11,12 all respond to SDF-1α, we propose that deficits in SDF-1α expression may represent a significant mechanism for globally impaired tissue regeneration in the aged population.
In the present study, we investigated the inherent function of young and aged EPCs to respond to an ischemic insult. We studied the effects of SDF-1α stimulation of EPCs, as well as the host’s ability to upregulate SDF-1α signaling in vivo and in vitro. Furthermore, we investigated the role of HIF-1α and prolyl hydroxylases (PHD 1, 2, and 3) on SDF-1α expression. Finally, we used deferoxamine (DFO), a known HIF-1α stabilizer, to restore the “young” environment in aged mice and reverse age-related complications.
Mouse Ischemia Model
Young (4 to 6 months, Jackson Laboratories, Bar Harbor, Me) and aged (18 to 24 months, National Institute of Aging, Bethesda, Md) C57/BL6 mice underwent ischemic flap surgery in accordance with the guidelines of the New York University and Stanford University institutional animal care and use committees. Briefly, a peninsular skin flap (2.5×1.25 cm) was elevated dorsally and resutured after a silicone sheet was placed between the flap and wound bed. This created a reproducible ischemic gradient confirmed by color laser Doppler (Moor Instruments, Devon, United Kingdom) and oxygen-sensing probe measurements (Oxford Optronix, Oxford, United Kingdom).4 (See the online-only Data Supplement for an expanded Methods section.)
Mouse EPC Mobilization Assay
Peripheral blood was obtained from young (n=5) and aged (n=5) mice, and erythrocytes were lysed with ammonium chloride and separated into pellets. Cells were then washed with PBS/EDTA and sorted by multichannel fluorescence-activated cell sorting for phycoerythrin-labeled Flk-1(VEGFR-2; BD Pharmingen, San Jose, Calif) and FITC-labeled CD11b (BD Pharmingen). Mouse EPCs were identified as Flk-1+/CD11b− cells, and percent mobilization of cells was sorted as previously described13,14 (see online-only Data Supplement).
Mouse Neovascularization Assay
On postoperative day 14, mice (n=4 per group) were euthanized and flap sections harvested. Frozen sections were stained for CD31 (BD Pharmingen) as described previously.14
Mouse Bone Marrow Transplantation Model
Gender-mismatched bone marrow transplantations were performed in 4 groups of mice (n=5 per group): young donor/young recipient, young donor/aged recipient, aged donor/young recipient, and aged donor/aged recipient. Bone marrow was harvested from femurs and tibia of male mice. The buffy coat was isolated with Histopaque 1083 (Sigma-Aldrich, St Louis, Mo) and washed with PBS/EDTA, and 1×106 cells in 200 μL of DMEM (Gibco [Invitrogen], Carlsbad, Calif) were injected via the femoral vein into irradiated female mice (1.6 Gy for 1.5 minutes for 2 cycles). Animals were allowed 30 days to reconstitute.
Recruitment of Mouse EPCs to Ischemic Flap
Flap sections from bone marrow–transplanted animals were deparaffinized in Citrisolv (Fisher Scientific, Fairlawn, NJ). CD31-PE (BD Pharmingen) and Y-chromosome–FITC (Cambio, Cambridge, United Kingdom) costaining was performed as described previously15; sections were counterstained with DAPI and examined under a multichannel fluorescent microscope (Olympus, Center Valley, Pa). Cells positive for CD31, Y chromosome, and DAPI markers were designated EPCs.
Deferoxamine Rescue of the Ischemic Flap
Deferoxamine (DFO; 10 mg/kg; Calbiochem, San Diego, Calif) was injected intraperitoneally into aged mice (n=4) 1 day before surgery and every other day.16 On postoperative day 7, flaps were analyzed grossly and for CD31+ staining and EPC mobilization as described above.
Enzyme-Linked Immunosorbent Assay
Quantikine human VEGF and murine SDF-1α ELISA kits (R&D Systems, Minneapolis, Minn) were used according to the manufacturer’s instructions.
EPC migration was evaluated with the NeuroPore transwell assay (8-μm pore size) as described previously.13
Human EPC Proliferation
EPCs (1×105) were plated in 12-well fibronectin-coated plates (n=12) and serum starved for 24 hours with media containing 0.5% FBS. A [3H] thymidine incorporation assay was performed as described previously.13
Determination of Human EPC Peripheral Blood Population
Peripheral blood (60 mL), isolated from younger (n=18, 18 to 35 years old) and older (n=21, 68 to 95 years old) patients in accordance with the guidelines of the New York University institutional review board, was separated with Histopaque 1077 (Sigma-Aldrich). The buffy coat was washed with PBS/10% FBS. Cells were labeled with AC133-PE17,18 (Miltenyi Biotech, Auburn, Calif) and sorted by fluorescence-activated cell sorting as described previously.13
Fibroblasts and endothelial cells were harvested from healthy younger (n=9, 18 to 35 years old) and older (n=7, 68 to 95 years old) patients undergoing melanoma excision in accordance with guidelines of the New York University institutional review board. Young and aged fibroblasts were also obtained from the National Institute on Aging’s Coriell Institute cell repository in Camden, NJ. Tissue explants were digested with 0.07% Liberase III Blendzyme (Roche, Palo Alto, Calif), filtered through a 70-μm nylon mesh, washed with PBS/0.2% BSA, and purified with CD31-coated magnetic beads. Endothelial cells were cultured on gelatin-coated plates in EGM-2 (Cambrex, East Rutherford, NJ). Unbound fibroblasts were cultured in DMEM (Gibco) with 10% FBS/1% antibiotics. Cultures were exposed to hypoxia (0.5% O2, 5% CO2) or normoxia (21% O2, 5% CO2) for 12 hours before protein and RNA harvesting. Cells obtained from the National Institute on Aging were grown in αMEM/15% FBS.
Primary murine fibroblasts were harvested in a similar fashion and subjected to hypoxia or normoxia and 100 μmol/L DFO (Calbiochem) for 18 hours. For hydroxylated HIF-1α studies, 10 μmol/L of the proteasome inhibitor MG132 (Peptides International, Louisville, Ky) was added, and protein and RNA were harvested.
Quantitative Real-Time Polymerase Chain Reaction
RNA was harvested with the RNeasy Fibrous Tissue Mini-Kit (Qiagen, Valencia, Calif) and from cell culture with the RNeasy Mini-Kit (Qiagen) according to the manufacturer’s instructions. mRNA was reverse transcribed to cDNA with the RNA PCR Core kit (Applied Biosystems, Foster City, Calif). Quantitative gene expression was determined with the Roche LightCycler 1.2 instrument with the LightCycler FastStart DNA MasterPLUS SYBR Green I kit (Roche). Absolute gene transcription was normalized to β-actin. Primer sequences are given in the Table.
A total of 50 to 80 μg of total protein extracted with RIPA buffer was separated on 7.5% SDS-PAGE, transferred to polyvinylidene difluoride membranes, and blocked with 5% milk/TBS with Tween. Protein detection was performed with primary antibodies HIF-1α (Novus, Littleton, Colo), PHD1 (Novus), PHD2 (Santa Cruz Biotechnology, Santa Cruz, Calif), PHD3 (Novus), and β-actin (Lab Vision, Fremont, Calif). Corresponding horseradish peroxidase–linked antibodies were used as the secondary antibody (Santa Cruz). Blots were developed with ECL reagent (Amersham, Piscataway, NJ) and exposed for 1 to 10 minutes on Kodak Biomax-MS film.
A previously constructed murine SDF-1 luciferase reporter construct was created by cloning the 2-kilobase SDF-1 promoter into the pGL3-basic luciferase vector (Promega, Madison, Wis). The reporter plasmid was cotransfected with a constitutively expressed Renilla luciferase construct (pHRL-TK, Promega) into primary young and aged murine fibroblasts by use of the Lipofectamine Plus reagent (Invitrogen). Twenty-four hours after transfection, cells were incubated in hypoxia (1% O2), normoxia, or normoxia with 100 μmol/L DFO for 18 hours. Luciferase activity was determined with the Dual-Luciferase System (Promega), and data were normalized to Renilla luciferase expression from at least 4 independent experiments.
All data are reported as mean±SEM, with the exception of fold induction, percentage increase, and ratios, which are reported as mean±SD. All in vitro cell experiments were repeated in triplicate, and all in vivo experiments (ischemic flap operations and bone marrow transplants) were performed with a minimum of 4 mice in each group. Statistical significance was evaluated with a paired Student t test or ANOVA, with P<0.05 considered significant.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Progenitor-Mediated Vascular Remodeling Is Impaired With Age
We examined the impact of aging on ischemia-induced vascular regeneration using a well-described soft tissue ischemia model.4 Aged mice exhibited a marked impairment in tissue survival (Figure 1A and 1B) and tissue oxygenation 7 days postoperatively compared with young mice (perfusion ratio 0.88±0.04 versus 0.58±0.07, P<0.005; Figure 1C). Aged animals also exhibited a marked decrease in mobilization of systemic Flk-1+/CD11b− progenitor cells13,14 compared with young animals, which suggests a vasculogenesis-specific impairment with age (1.4±0.2% EPCs versus 0.4±0.1% EPCs, P<0.005; Figure 1D), which also correlated with decreased vascular density (number of vessels in section B/number of vessels in section D per high-powered field: 1.96±0.24 versus 1.24±0.1, P<0.005; Figure 1E).
To assess the impact of aging on the bone marrow compartment, ischemic flaps from age-matched and gender-mismatched mice were analyzed for EPC-derived neovascularization. Colocalization of fluorescent in situ hybridization for the Y chromosome and phycoerythrin immunostaining for CD31 was used to determine the number of EPCs that incorporated into neovessels within ischemic tissue. A dramatic 5-fold reduction was observed in EPCs incorporated in vascular structures in aged versus young mice (28.8±4.4 versus 5.2±1.2 vessels per high-powered field, P<0.05; Figure 1F).
Lectin staining in situ also demonstrated patent vascular channels lined with EPCs (data not shown).4 These data suggest that aging impairs neovascularization in ischemia by reducing bone marrow–mediated vasculogenesis, which results in decreased vascular density and overall tissue survival.
Intrinsic Function of Primitive Vascular Progenitors Is Relatively Preserved With Age
It has been suggested that the observed impairments in vasculogenesis with age are due to “progenitor exhaustion.”6,7 However, examination of the mouse bone marrow compartment for primitive Sca-1+/c-kit+/lin− (SKL) vascular progenitors,19 thought to be a more primitive precursor of EPCs, demonstrated no significant differences between young and aged animals (1470±475 versus 2220±272, P>0.05; Figure 2A).
Progenitor cell migration in vitro has been used to evaluate cell function and has been correlated with specific risk factors and severity of cardiovascular disease.20,21 Primitive bone marrow–derived SKL vascular progenitors purified from young and aged mice migrated in similar numbers toward an SDF-1α gradient (596±84 versus 560±159, P>0.05; Figure 2B), which suggests that aged vascular progenitors functioned as well as their young counterparts.
To examine whether aged EPCs retained the ability to respond to an ischemic stimulus, systemic SDF-1α was administered intraperitoneally, and the number of mobilized EPCs was analyzed. Strikingly, both young and aged animals mobilized comparable numbers of EPCs (% mobilized: 1.30±0.16% versus 1.35±0.25%, P>0.05; Figure 2C), which suggests not only that an adequate reservoir of EPCs exists in aged bone marrow but also that EPCs were capable of responding appropriately to a systemic stimulus. These findings further suggest that a defect in the signals necessary for EPC mobilization and recruitment may contribute in part or entirely to the decreased neovascularization seen with aging.
To confirm the human relevance of these findings, we examined the number and functional capacity of EPCs in healthy younger (n=18, mean age 25.7 years, range 18 to 35 years) and older (n=21, mean age 82.1 years, range 68 to 95 years) human subjects without evidence of cardiovascular disease,22 diabetes mellitus,20 or cholesterol-lowering statin therapy.23 No significant difference was present in baseline numbers of circulating EPCs between younger and older patients presented as percentage of baseline circulating mononuclear cells (% baseline, 0.184±0.09% versus 0.183±0.010%, P>0.05; Figure 2D). Likewise, no significant difference was present in EPC colony formation (19±8 versus 18±6 colonies, P>0.05; Figure 2E), migration toward SDF-1α (604±103 versus 603±64 cells, P>0.05; Figure 2F), or hypoxia-induced proliferation (2524±400 versus 1973±79 cells, P>0.05; Figure 2G), which demonstrates that the intrinsic EPC function is preserved in humans with advancing age. Collectively, these data suggest that the decline in neovascularization that occurs with aging is not a consequence of decreased EPC number or function.
Aging Neovascular Phenotype Corresponds With Altered Environmental Signals
To confirm the importance of progenitor depletion and peripheral tissue recruitment, we performed age- and gender-mismatched reciprocal bone marrow transplantations, which yielded 2 control groups (young donor/young recipient and aged donor/aged recipient) and 2 experimental groups (young donor/aged recipient and aged donor/young recipient). Interestingly, aged bone marrow transplanted into a young ischemic environment demonstrated mobilization of aged EPCs comparable to that in young controls (1.42±0.34% versus 1.53±0.46%, P>0.05; Figure 3A). This resulted in similar increases in tissue perfusion ratio (0.81±0.11 versus 0.84±0.03, P>0.05; Figure 3B) and vascular density (number of vessels in section B/number of vessels in section D: 1.42±0.15 versus 1.56±0.25, P>0.05; Figure 3C). The number of CD31+/Y-chromosome+ chimeric neovessels (Figure 3E) was similar to that in young animal controls (23.4±3.1 versus 28.8±4.4, P>0.05; Figure 3D), which demonstrates normal targeting and incorporation of aged EPCs into the host vasculature in response to ischemia. Collectively, this adds further support to the theory that intrinsic EPC function remains intact even with aging given appropriate environmental signals.
However, young progenitors in an aged host exhibited marked deficiencies in mobilization from the bone marrow after an ischemic stimulus, similar to aged controls (0.41±0.04 versus 0.42±0.09, P>0.05; Figure 3A). This resulted in markedly fewer CD31+/Y-chromosome+ chimeric neovessels (7.6±2.8 versus 5.2±1.2, P>0.05; Figure 3E), reduced flap perfusion (0.57±0.05 versus 0.50±0.08, P>0.05), and decreased vessel density (0.92±0.08 versus 1.01±0.10, P>0.05; Figure 3B and 3C), similar to aged controls. Thus, these results also indicate the age-associated decline in vasculogenesis results from failure of peripheral tissues to generate a suitable signal for EPC recruitment rather than from primary EPC depletion or dysfunction.
Aging Impairs Systemic and Local Hypoxic Responses Required for Progenitor-Mediated Vascular Remodeling
We have previously demonstrated that SDF-1α is critical to EPC mobilization and trafficking to sites of neovascularization.4,24 HIF-1α mediates ischemia-induced SDF-1α expression,4 and the ability to upregulate HIF-1α is thought to be impaired in aging.25 ELISA demonstrated a significant reduction in SDF-1α protein in ischemic flaps harvested from aged animals (347.4±24.3% versus 52.1±32.2% in section B, P<0.005; Figure 4B), which correlated with a marked decrease in HIF-1α protein stabilization (Figure 4A). Primary aged murine fibroblasts also demonstrated decreased HIF-1α expression compared with young cells (Figure 4C). This suggests that the observed deficiencies in SDF-1α within aged ischemic tissues may result from decreased HIF-1α stabilization.
These experiments were repeated in human microvascular endothelial cells and fibroblasts harvested from surgical specimens of healthy younger (n=9, age 18 to 35 years, mean 24.6 years) and older (n=7, age 68 to 95 years, mean 84.5 years) human subjects. Under hypoxia, aged human fibroblasts exhibited decreased HIF-1α stabilization (Figure 4D, bottom), which translated into a blunted ability to upregulate SDF-1α mRNA expression measured by real-time polymerase chain reaction (fold induction: endothelial cells 1.85±0.21 versus 0.67±0.35, P<0.005, and fibroblasts 1.90±0.21 versus 0.90±0.17, P<0.008; Figure 4D, top left). ELISA analysis of VEGF, another downstream target of HIF-1α, also demonstrated reduced baseline levels in aged cells (2.67±0.52 versus 124.93±14.98 pg/mL protein, aged versus young cells, P<0.0004) and an inability to upregulate VEGF levels in response to hypoxia (25.30±5.30 versus 615.85±35.75 pg/mL protein, aged versus young, P<0.00001; Figure 4D, top right). Although the fold increase was higher in aged cells, the very low baseline levels of VEGF in cells from older subjects greatly magnified small changes in VEGF levels. Even with hypoxic stimulation, VEGF levels in aged cells were below baseline levels in young cells.
Aging Increases HIF-1α Degradation Through Increased PHD Activity
To elucidate the mechanism for decreased HIF-1α stability in aged cells, we investigated the contribution of PHDs, which hydroxylate proline residues (p402 and p564) in the presence of oxygen and iron,26 targeting HIF-1α for ubiquitin-mediated proteasome degradation.26 Additionally, HIF-1α function is mediated through the asparaginyl hydroxylase, factor-inhibiting HIF, which on hydroxylation of an asparagine group prevents the binding of the necessary transcriptional cofactor p300.26
Primary aged murine fibroblasts demonstrated 170% to 250% higher baseline transcription of all hydroxylases compared with young controls (% increase aged/young: PHD1 171±10.7%, PHD2 253±41.6%, PHD3 215±39%; factor-inhibiting HIF 257±43.9%; Figure 5A), which correlated with decreased HIF-1α protein expression (data not shown). These data were further confirmed by examination of protein expression of PHD1, PHD2, and PHD3, which was also significantly higher in aged tissues (Figure 5B through 5D).
To establish the relevance of increased PHD levels with HIF-1α degradation in aging, we performed Western blots that specifically targeted the proline 564 residue. Because hydroxylated HIF-1α is degraded rapidly, the proteasome inhibitor MG132 was administered. Aged murine fibroblasts demonstrated increased levels of hydroxylated HIF-1α (Figure 5E). In addition, although hypoxic conditions decreased levels of hydroxylated HIF-1α in aged cells, these levels never returned to the baseline levels seen in young cells (Figure 5E). This demonstrates that aged cells exhibit increased baseline PHD1–3 function and increased HIF-1α degradation.
DFO Promotes Ischemic Flap Survival Through Increased EPC Mobilization Via Increased HIF-1α and SDF-1α Levels
To demonstrate the causal importance of decreased HIF-1α stabilization in vascular defects observed with aging, we performed a rescue experiment. DFO inhibits PHD hydroxylation of HIF-1α proline residues through sequestration of iron, a necessary cofactor in the reaction.16,26,27 Western blots on young and aged primary murine fibroblasts confirmed the ability of DFO to increase HIF-1α stabilization (Figure 6A) with downstream increases in SDF-1α and other HIF-responsive genes. Augmentation of HIF-1α stabilization with DFO in aged cells surpassed hypoxic HIF-1α stabilization in young cells (data not shown). Furthermore, although aged cells were unable to increase SDF-1α luciferase activity in hypoxia, DFO-treated aged cells demonstrated significant upregulation of luciferase activity that surpassed even that of young cells (fold induction 0.87±0.03 for aged, 1.67±0.09 for young, and 2.08±0.17 for aged+DFO; Figure 6B).
In vivo, mice were also treated with DFO. Fluorescence-activated cell sorting analysis demonstrated a marked increase in systemic Flk-1+/CD11b− EPCs in DFO-treated aged mice compared with aged controls (1.43±0.08% versus 0.56±0.12% EPCs, P<0.02) and young controls (Figure 6C). This correlated with increased vascular density (1.96±0.24 versus 1.24±0.1 vessels in section B/vessels in section D per high-powered field, P<0.005; Figure 6D, left) and with significantly improved flap survival compared with their untreated aged counterparts, which approached the survival seen in young mice (Figure 6E and 6F). In contrast, we found no change in vascular density in normal skin (5.4±0.51 versus 4.8±0.38, P>0.05; Figure 6D, right) or serum SDF-1α and VEGF levels in DFO-treated animals (data not shown). These experiments confirm the well-established role of DFO to stabilize HIF-1α and reverse the age-related decline in HIF-1α and improve tissue survival.
Progenitor-mediated regeneration occurs through differentiation of tissue-resident progenitor cells28 or trafficking of bone marrow–derived progenitor cells3,29 to sites of injury. Both require the local environment to generate appropriate chemotactic signals and progenitor cells to respond appropriately. Recently, there has been speculation about the hypothesis that aging may result from primary progenitor cell dysfunction or exhaustion.6,7 However, it has been demonstrated that senescent bone marrow cells retain the capacity to repopulate depleted bone marrow over multiple successive generations,30 which suggests that progenitor cells remain fully functional even with aging. In the present study, intrinsic EPC function and number remained intact both in humans and mouse model systems that used well-described assays of progenitor function.4,14 This led us to examine whether age-related defects in neovascularization could be attributed to a lack of hypoxia-induced signals necessary for EPC mobilization.
The repair and regeneration of the vascular system requires local vessel repair through angiogenesis and vasculogenesis. We have previously demonstrated that newly formed blood vessels in injured tissues can be composed entirely of EPCs.13 It has also been proposed that bone marrow cells recruited to areas of neovascularization are perivascular recruited bone marrow–derived circulating cells (RBCCs) that function as “helper” cells. These cells are postulated to augment neovascularization by secreting SDF-1α in response to upregulated local VEGF production.31 Regardless of the mechanism or cell type recruited, it is clear that the process of vasculogenesis requires an elaborate cascade of signaling events capable of mobilizing, homing, and retaining these cells. In the present study, using our bone marrow transplantation model, we demonstrate that vasculogenesis is markedly reduced with aging.
During development, gradients of oxygen tension regulate gene expression and determine spatial patterns of tissue organization via specific chemokines and growth factors that guide primordial progenitor cells.32 In the adult, these oxygen gradients result from injury rather than tissue growth and can be demonstrated experimentally by reorientation of blood vessels to align with decreasing oxygen tensions and increasing chemokine concentrations.4,13 Central to the process of angiogenesis is the ability of local cells to sense these conditions of hypoxia and stimulate VEGF production through stabilization of intracellular HIF-1α, thus inducing migration of endothelial cells.33 We have demonstrated that vasculogenesis functions in a similar fashion in which local cells increase HIF-1α stabilization, which leads to augmented SDF-1α production and EPC recruitment.4
The present investigation into HIF-1α stability as a common mediator for both angiogenesis and vasculogenesis indicates that aging prevents upregulation of HIF-1α because of constitutively active PHDs that increase HIF-1α degradation. This is clearly demonstrated by the inability to stabilize HIF-1α even under hypoxia, which leads to decreased levels of SDF-1α, impaired EPC mobilization, decreased vasculogenesis, and increased tissue necrosis. These data indicate an inability to mount a normal hypoxic response critical for the guidance and retention of reparative EPCs.
Interestingly, systemic administration of exogenous SDF-1α in aged animals promoted mobilization of EPCs but failed to stimulate targeted EPC recruitment and increase tissue survival. Presumably, this results from limited tissue-specific SDF-1α expression to recruit and maintain EPCs. This suggests the stimulation of EPC-mediated repair is unlikely to be recaptured by chemokine monotherapy. However, global augmentation of HIF-1α with DFO promoted neovascularization only in areas of ischemia and did not produce neovascularization in uninjured skin. This selective effect has been confirmed by Duffy et al,34 who demonstrated no significant effect of DFO on human endothelial cells in the absence of ischemic coronary artery disease. The reasons for this are unclear, but recent work has examined the complex interplay between intracellular iron, PHD activity, and HIF stabilization35 and suggests that iron becomes limiting only in conditions of low oxygen availability.27 From this, one would predict that in vivo administration of DFO would not completely inhibit PHD function in normoxia but would linearly decrease PHD activity as oxygen availability decreases in the ischemic tissue, as we observed in the present study.
The findings of the present study have been confirmed by Rohrbach et al,36 who demonstrated an upregulation of PHD3 in aged cardiac tissue; however, unlike the present study, they found no difference in PHD1 or PHD2 in aged tissues. This difference may be due to the well-described tissue specific alterations in PHD expression37–39 or to differences in the classification of the younger and older groups. In the study by Rohrbach et al,36 younger and older groups were 18 to 55 and 55 to 75 years old, respectively, whereas specimens in the present study were harvested from chronologically distinct groups at the extremes of age (18 to 35 years and 68 to 95 years, respectively). This may have provided a more robust determination of differences in PHD1 and PHD2. Nonetheless, their data corroborate the present findings that aging increases PHD-mediated HIF-1α degradation.
The precise mechanism for constitutively activated PHDs in aging remains unknown. Current theories of aging include cellular senescence, accumulation of somatic mutations that lead to cellular apoptosis,40 and free radical accumulation that causes direct cellular damage.40 One possible mechanism for the observed alteration in HIF-1α stability is through increased accumulation of reactive oxygen species with age. It has been shown the expression of PHDs, which are highly dependent on binding of oxygen to iron,41 is enhanced by reactive oxygen species.42 Therefore, it is possible that long-term exposure to reactive oxygen species may permanently upregulate PHD expression and function, thereby destabilizing HIF-1α. The present data demonstrating increased levels of hydroxylated HIF-1α with aging support the idea that reactive oxygen species may indeed play a role in increased HIF-1α degradation.
The potential of stem cell–based therapies has led to an explosion in the number of clinical trials that use stem cells.43 Using bone marrow cell transplantation, several phase I trials for the treatment of myocardial infarction have shown only modest results.8,9 Although these studies demonstrated short-term improvements in cardiac function, these improvements could not be sustained.10 Two possible explanations are the inability to retain stem cells in those areas and the complex temporal coordination between cell delivery and optimal chemokine production.8,10 The present demonstration of the dysfunctional hypoxia response in aged tissue with subsequent impairments in downstream gene expression (SDF-1α and VEGF) offers a potential explanation for the disappointing outcomes of clinical trials to date. Therefore, therapeutic interventions designed to restore the “young” hypoxic response may be a powerful strategy to prevent age-associated disease and functional decline by recruiting and retaining native or delivered stem cells to areas of injury.
Sources of Funding
This study was supported in part by the National Institute on Aging (R01-AG025016-03 to Dr Gurtner), an American Heart Association postdoctoral fellowship (Dr Loh), and the National Institutes of Health Loan Repayment Program (Dr Edward I. Chang).
Lakatta EG, Levy D. Arterial and cardiac aging: major shareholders in cardiovascular disease enterprises, part I: aging arteries: a “set up” for vascular disease. Circulation. 2003; 107: 139–146.
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.
Rauscher FM, Goldschmidt-Clermont PJ, Davis BH, Wang T, Gregg D, Ramaswami P, Pippen AM, Annex BH, Dong C, Taylor DA. Aging, progenitor cell exhaustion, and atherosclerosis. Circulation. 2003; 108: 457–463.
Schachinger V, Erbs S, Elsasser A, Haberbosch W, Hambrecht R, Holschermann H, Yu J, Corti R, Mathey DG, Hamm CW, Suselbeck T, Assmus B, Tonn T, Dimmeler S, Zeiher AM. Intracoronary bone marrow-derived progenitor cells in acute myocardial infarction. N Engl J Med. 2006; 355: 1210–1221.
Peled A, Petit I, Kollet O, Magid M, Ponomaryov T, Byk T, Nagler A, Ben-Hur H, Many A, Shultz L, Lider O, Alon R, Zipori D, Lapidot T. Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR4. Science. 1999; 283: 845–848.
Tepper OM, Capla JM, Galiano RD, Ceradini DJ, Callaghan MJ, Kleinman ME, Gurtner GC. Adult vasculogenesis occurs through in situ recruitment, proliferation, and tubulization of circulating bone marrow-derived cells. Blood. 2005; 105: 1068–1077.
Lyden D, Hattori K, Dias S, Costa C, Blaikie P, Butros L, Chadburn A, Heissig B, Marks W, Witte L, Wu Y, Hicklin D, Zhu Z, Hackett NR, Crystal RG, Moore MA, Hajjar KA, Manova K, Benezra R, Rafii S. Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat Med. 2001; 7: 1194–1201.
Wang GL, Semenza GL. Desferrioxamine induces erythropoietin gene expression and hypoxia-inducible factor 1 DNA-binding activity: implications for models of hypoxia signal transduction. Blood. 1993; 82: 3610–3615.
Yin AH, Miraglia S, Zanjani ED, Almeida-Porada G, Ogawa M, Leary AG, Olweus J, Kearney J, Buck DW. AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood. 1997; 90: 5002–5012.
Gehling UM, Ergun S, Schumacher U, Wagener C, Pantel K, Otte M, Schuch G, Schafhausen P, Mende T, Kilic N, Kluge K, Schafer B, Hossfeld DK, Fiedler W. In vitro differentiation of endothelial cells from AC133-positive progenitor cells. Blood. 2000; 95: 3106–3112.
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.
Heeschen C, Lehmann R, Honold J, Assmus B, Aicher A, Walter DH, Martin H, Zeiher AM, Dimmeler S. Profoundly reduced neovascularization capacity of bone marrow mononuclear cells derived from patients with chronic ischemic heart disease. Circulation. 2004; 109: 1615–1622.
De Falco E, Porcelli D, Torella AR, Straino S, Iachininoto MG, Orlandi A, Truffa S, Biglioli P, Napolitano M, Capogrossi MC, Pesce M. SDF-1 involvement in endothelial phenotype and ischemia-induced recruitment of bone marrow progenitor cells. Blood. 2004; 104: 3472–3482.
Rivard A, Berthou-Soulie L, Principe N, Kearney M, Curry C, Branellec D, Semenza GL, Isner JM. Age-dependent defect in vascular endothelial growth factor expression is associated with reduced hypoxia-inducible factor 1 activity. J Biol Chem. 2000; 275: 29643–29647.
Appelhoff RJ, Tian YM, Raval RR, Turley H, Harris AL, Pugh CW, Ratcliffe PJ, Gleadle JM. Differential function of the prolyl hydroxylases PHD1, PHD2, and PHD3 in the regulation of hypoxia-inducible factor. J Biol Chem. 2004; 279: 38458–38465.
Hirsila M, Koivunen P, Xu L, Seeley T, Kivirikko KI, Myllyharju J. Effect of desferrioxamine and metals on the hydroxylases in the oxygen sensing pathway. FASEB J. 2005; 19: 1308–1310.
Simper D, Stalboerger PG, Panetta CJ, Wang S, Caplice NM. Smooth muscle progenitor cells in human blood. Circulation. 2002; 106: 1199–1204.
Duffy SJ, Biegelsen ES, Holbrook M, Russell JD, Gokce N, Keaney JF Jr, Vita JA. Iron chelation improves endothelial function in patients with coronary artery disease. Circulation. 2001; 103: 2799–2804.
Qutub AA, Popel AS. A computational model of intracellular oxygen sensing by hypoxia-inducible factor HIF1 alpha. J Cell Sci. 2006; 119: 3467–3480.
Hirsila M, Koivunen P, Gunzler V, Kivirikko KI, Myllyharju J. Characterization of the human prolyl 4-hydroxylases that modify the hypoxia-inducible factor. J Biol Chem. 2003; 278: 30772–30780.
Takeda K, Cowan A, Fong GH. Essential role for prolyl hydroxylase domain protein 2 in oxygen homeostasis of the adult vascular system. Circulation. 2007; 116: 774–781.
Weinert BT, Timiras PS. Invited review: theories of aging. J Appl Physiol. 2003; 95: 1706–1716.
Berchner-Pfannschmidt U, Yamac H, Trinidad B, Fandrey J. Nitric oxide modulates oxygen sensing by hypoxia-inducible factor 1-dependent induction of prolyl hydroxylase 2. J Biol Chem. 2007; 282: 1788–1796.
Recent clinical trials of stem and progenitor cell treatments for ischemic disease have been disappointing. This has led to a reappraisal of the potential determinants of stem cell activity in vivo. Among investigators in the field, appreciation is growing that a receptive environment or “soil” is critical for stem or progenitor cells to exert a proregenerative effect. Advanced age is a well-established risk factor for increased morbidity and mortality after myocardial infarction and poor tissue regeneration after an ischemic insult. In the present report, we demonstrate a profound age-related decrease in the signals necessary for endothelial progenitor cell recruitment and vasculogenesis in both human and experimental systems. Interestingly, we were unable to elucidate any differences in intrinsic progenitor cell function in younger and older humans, which suggests that the progenitor cells themselves are unaffected by aging. These “soil” abnormalities were subsequently traced to increased degradation of the transcription factor hypoxia-inducible factor 1 α (HIF-1α), which resulted in reduced activation of the hypoxia response genes VEGF (vascular endothelial growth factor) and SDF-1α (stromal cell–derived factor 1). This ultimately results in impaired neovascularization and increased tissue necrosis. Because VEGF is also an HIF-1α–dependent gene, these findings have broader implications for ischemia-induced angiogenesis as well. These age-related effects can be reversed by treatment with deferoxamine, which increases HIF-1α stabilization and subsequent VEGF and SDF-1α gene expression. The present report provides a mechanistic window into the well-established effects of aging on neovascularization and suggests novel therapeutic strategies for increasing vascular growth in elderly patients.
↵*Drs Chang and Loh contributed equally to this work.
The online-only Data Supplement, consisting of an expanded Methods section, is available with this article at http://circ.ahajournals.org/cgi/content/ full/CIRCULATIONAHA.107.715847/DC1.