Ординатура / Офтальмология / Английские материалы / Retinal and Choroidal Angiogenesis_Penn_2008

niche, suggesting that hypoxia may be a fundamental requirement for progenitor cell trafficking and function. As such, ischemic tissue may represent a “conditional” stem cell niche, with recruitment and retention of circulating progenitors regulated by hypoxia through regulated expression of SDF-1.46 For example, Ma and co-workers found that myocardial SDF-1 expression was increased only in the early phase post MI and that intravenous stem cell infusion was successful only if given in the early phase of MI, where it resulted in enhanced angiogenesis and improved cardiac function.47

SDF-1 functions in a paracrine manner with other inflammatory mediators. Reduced endothelial cell expression of SDF-1 by tumor necrosis factor-alpha (TNF-α) and interferon-gamma (IFN-γ) disrupts ECMdependent endothelial cell tube formation, an in vitro morphogenic process that recapitulates critical steps in angiogenesis. Replacement of SDF-1 onto the endothelial cell surface reconstitutes this morphogenic process.48 In vivo, TNF-α and IFN-γ inhibit growth factor-induced angiogenesis and SDF-1 expression in endothelial cells. These results demonstrate that SDF- 1/CXCR4 constitutes a TNF-α- and IFN-γ-regulated signaling system that plays a critical role in mediating angiogenesis inhibition by these inflammatory cytokines.48

VEGF and FGF-2 increase the expression of CXCR4 on endothelial cells, rendering these cells more responsive to SDF-1.9 Salcedo showed that prostaglandin E2 (PGE2) mediates the effects of FGF-2 and VEGF in upregulating CXCR4 expression on human microvascular endothelial cells.44 Furthermore, the ability of PGE2 to augment in vitro tubule formation in SDF-1-containing Matrigel was inhibited completely by blocking CXCR4. Moreover, they determined that augmentation of CXCR4 expression by VEGF, FGF-2, and PGE2 involves stimulation of transcription factors binding to the Sp1-binding sites within the promoter region of the CXCR4 gene. These findings indicate that PGE2 is a mediator of VEGFand FGF-2- induced CXCR4-dependent neovessel assembly in vivo and show that the angiogenic effects of PGE2 require CXCR4 expression.

De Falco et al. found that after femoral artery dissection, plasma SDF-1 levels were upregulated, while SDF-1 expression in the BM was downregulated concomitantly with the increase in the percentage of progenitor cells in the peripheral blood. An increase in ischemic tissue expression of SDF-1 at the RNA and protein levels was also observed.49

Yamaguchi et al. investigated the effect of SDF-1 on EPC-mediated vasculogenesis. They demonstrated an increase in CXCR4 expression as EPCs differentiate and showed that SDF-1 reduced apoptosis in these cells. They also injected SDF-1 into the ischemic hind limb muscle of nude mice and then provided fluorescence-labeled human EPC exogenously.

Fluorescence microscopic examination disclosed increased local accumulation of EPCs in ischemic muscle in the SDF-1 treatment group compared to controls. At day 28 after treatment, ischemic tissue perfusion was improved in the SDF-1 group, and capillary density was also increased. The authors concluded from their findings that locally delivered SDF-1 augments vasculogenesis and subsequently contributes to ischemic NV in vivo by augmenting EPC recruitment to ischemic tissues.28

Wang and co-workers explored angiogenic and signaling mechanisms generated following SDF-1 binding to CXCR4 in tumor cells. Both AKT and ERK pathways were stimulated following SDF-1 exposure. AKT activation resulted in expression of VEGF and tissue inhibitor of metalloproteinase (TIMP)-2, whereas ERK activation led to interleukin (IL)- 6 and IL-8 secretion. Expression of angiostatin was inversely related to CXCR4 levels and was inhibited by SDF-1 stimulation. These data link the SDF-1/CXCR4 pathway to changes in angiogenic cytokines by different signaling mechanisms and suggest that the delicate equilibrium between proangiogenic and anti-angiogenic factors may be achieved by different signal transduction pathways.50

4.1.1SDF-1 as a Target for Inhibiting EPC Involvement in Tumor and Ocular Angiogenesis

Guleng et al. showed that inhibition of the SDF-1/CXCR4 pathway decreases the growth of subcutaneous gastrointestinal tumors through the suppression of tumor neoangiogenesis. CD31+ tumor capillaries were reduced to 45% and intratumor blood flow was decreased to 65% by blockade of CXCR4. These findings show that the anti-angiogenic effects of blocking CXCR4 are related to a reduction in the establishment of tumor endothelium.51

Very little is known about SDF-1 action in the eye and its relevance to ocular angiogenesis. This area has been a recent focus of our investigative efforts. We proposed that chemokines such as SDF-1 may be responsible for development of diffuse macular edema (DME) and/or aberrant NV in patients with proliferative diabetic retinopathy (PDR). SDF-1 is a potent stimulator of VEGF expression, the main effector of NV and the key inducer of vascular permeability associated with DME. In a prospective study, we investigated the relationship between SDF-1 and VEGF in the vitreous of patients with varying degrees of diabetic retinopathy and DME before and after intraocular injection of triamcinolone acetonide, which is used to treat refractory DME.52 Thirty-six patients were included and observed for 6 months. Vitreous VEGF and SDF-1 levels were measured in samples obtained immediately before and 1 month after injection of triamcinolone.

We found that both VEGF and SDF-1 were significantly higher in patients with PDR than in patients with nonproliferative diabetic retinopathy. Levels of SDF-1 were markedly increased in patients with DME compared with those without DME. VEGF correlated with SDF-1 levels and disease severity. Both vitreous VEGF and SDF-1 levels declined significantly after triamcinolone treatment. We concluded that the elimination of DME with

regression of NV after triamcinolone injection may be due to the suppression of VEGF and SDF-1.52

In subsequent studies we found that SDF-1 induces human retinal endothelial cells to increase expression of VCAM-1. This could facilitate EPC attachment to resident endothelial cells, which is a required step prior to extravasation and migration into the ischemic tissue. SDF-1 also influenced cellular tight junctions by reducing occludin expression. Both changes could serve to recruit HSCs or hematopoietic cells (choose one) and EPCs along an SDF-1 gradient.53

Using a murine model of retinal angiogenesis, we showed that the majority of new vessels formed in response to oxygen starvation originated from HSC-derived EPCs. The levels of SDF-1 found in the vitreous of patients with PDR were able to induce retinopathy in our murine model. Intravitreal injection of blocking antibodies to SDF-1 prevented retinal NV in our murine model, even in the presence of exogenous VEGF. Our data suggest that SDF-1 plays a role in ocular angiogenesis and disruption of

blood-retina barrier function and may be an ideal target for the prevention of PDR.53

The wet form of age-related macular degeneration is characterized by choroidal neovascularization (CNV). We demonstrated that antibodies to SDF-1 reduced NV following laser-induced rupture of Bruch’s membrane. Antibody treatment reduced the degree of stem cell recruitment and incorporation into the CNV lesions, compared with the control, and reduced the size of the CNV lesions. These studies indicate that targeting SDF-1 may represent a strategy to prevent CNV.54

4.2EPC Response is Influenced by the Type and Severity of Injury

EPCs arise from a common progenitor, the hemangioblast, during embryogenesis.55 We recently showed that adult murine BM-derived HSCs possess hemangioblast activity.34 From this work, one would predict that the hemangioblasts give rise to circulating EPCs that participate in vessel homeostasis throughout adult life. Understanding the roles of hemangioblasts and EPCs in the repair of damaged vessels or in pathogenic tumor neoangiogenesis has been complicated by apparent differences in the

involvement of these cells depending upon the type of host injury or tumor model used.

We used a combination of ischemic and traumatic injury with VEGF overexpression to induce retinal NV in the adult mouse. The combination of insults was required to induce a robust proliferative retinopathy similar to that seen in human PDR. Recently we examined the role of the nitric oxide (NO) pathway in modulating vessel repair by transplanted populations of green fluorescence protein (gfp)-expressing BM cells. Recipient mice deficient in inducible nitric oxide synthase (iNOS–/–) or endothelial nitric oxide synthase (eNOS–/–) received transplants of congenic BM-derived gfp+ cells enriched for HSCs. At three months following transplantation, an adeno-associated viral vector expressing VEGF was injected into the vitreous of the test eye of each host. One month later, laser photocoagulation of the vessels juxtaposed to the optic nerve was performed, leading to an ischemic injury to nearly one-half of the treated retina. The non-laser-treated eye served as a control.

The gfp+ HSC-enriched cells contributed to new vessel formation in the laser-treated, but not the control, eye of normal congenic hosts. Similarly, gfp+ HSC-enriched cells contributed to new vessel formation in the iNOS–/– hosts only in the laser-injured retina, with little contribution in the contralateral non-laser-treated retina. Surprisingly, gfp+ HSC-enriched cells robustly contributed to NV in both test and control eyes of eNOS–/– hosts, and gfp+ cells populated large and small vessels in all tissues examined. This result was in stark contrast to the paucity of gfp+ cell contribution to systemic vascular tissues in the wild-type and iNOS–/– hosts. Thus, eNOS–/– mice appear to display a systemic vascular dysfunction in which BM-derived progenitor cells are extensively recruited and incorporated into the vessels. eNOS–/– mice have hypertension, hyperglycemia, and widespread vascular dysfunction. Thus, it is not surprising that gfp+ BM-derived cells reendothelialize the entire vasculature of these mice. We propose that eNOS deficiency may lead to a compensatory overexpression of vascular iNOS and subsequently to pathological vascular endothelial cell turnover. NO appears to have an important role in the mechanisms of EPC mobilization and engraftment in certain neoangiogenic sites.56

There are considerable data to support the idea that altered EPC function may play a role in atherosclerosis as well as in diabetic complications. It is predicted that EPC dysfunction is the cause of both PDR and the marked increased risk of atherosclerosis suffered by individuals with diabetes. It may seem incongruous to discuss commonality in the cause of PDR, a condition with excessive NV, with the condition of diabetic peripheral vascular disease and atherosclerosis, which are associated with reduced re-endothelialization following injury.57 However the entire diabetic endothelium suffers damage

as a result of oxidative stress and the hyperglycemic state. Injured macrovasculature endothelium, if not repaired, leads to a propensity for atherosclerosis. With regard to the retina, this same endothelial damage results in capillaries lacking an intact endothelium, i.e. acellular capillaries. EPCs are critical for endothelial repair of both the macroand microvasculature. We have characterized a defect in migration of EPCs from individuals with diabetes that could result in an impaired ability to repair endothelial injury in either the microor macrovasculature.

In the retina, a defect in the EPCs would prevent repair of endothelial injury early on, leading to acellular capillary lesions and retinal ischemia. In the macrovasculature, the inability to repair the endothelium results in an increase in monocyte chemoattractant protein-1 (MCP-1) and upregulation of adhesion molecules with an influx of lipoprotein, monocytes, and T cells, initiating the atherosclerotic lesion.57 Thus, the initial cause of PDR and atherosclerosis may be the same: lack of EPC repair of the endothelium (see Figure 3). However, the microvasculature within the retina is unique, since the vitreous has been shown to be a “sink” for chemokines and growth factors secreted by the retina.52 Endothelial injury, acellular capillaries, and retinal ischemia all lead to the compensatory release of chemokines and

growth factors such as VEGF, SDF-1, and MCP-1. These are all found at increased levels in the vitreous of patients with PDR.52,58 Although we state

that the diabetic EPCs are defective in migration, the markedly increased levels of chemokines and growth factors within the vitreous of the diabetic patient may ultimately overcome the EPC migratory defect. This migration leads to abnormally located, pre-retinal vessels that are poorly developed and prone to bleeding. We are currently testing the hypothesis that if we reverse the diabetic EPC defect in migration early on, we can repair the acellular capillaries, prevent ischemia and the subsequent retinal expression of chemokines and growth factors that accumulate in the vitreous, and ultimately prevent the development of PDR.

5.PHARMACOLOGICAL THERAPY TO MANIPULATE EPC NUMBER AND FUNCTION

Hydroxymethylglutaroyl coenzyme A (HMG-CoA) reductase inhibitors (statins) have been shown to enhance EPC mobilization and function and improve endothelial function independent of cholesterol reduction.59

Dimmeler et al. and Llevadot et al. both demonstrated that statins increased EPC differentiation in vitro and enhanced mobilization of EPCs in vivo.60,61

These effects were believed to be mediated via the phosphatidyl inositol-3- kinase/Akt pathway, which is also known to be activated by angiogenic growth factors and to influence HSC differentiation.62 Statins promoted angiogenesis in animals that did not have hypercholesterolemia. They also promoted the accumulation of BM-derived cells during corneal NV.63 Four weeks of atorvastatin treatment resulted in a 3-fold increase in circulating EPCs in patients with stable coronary artery disease patients.64 Statins increase the number of circulating EPCs that participate in repair after ischemic injury and may be responsible for the enhancement of coronary blood flow observed in patients with stable coronary artery disease following treatment. Statins accelerate re-endothelialization after vascular injury.8 Statins increased EPC mobilization and incorporation into damaged endothelium and minimized intimal hyperplasia.

Integrin receptor expression was influenced by statins, altering cell adhesiveness and promoting EPC homing.8 In order to allow homing to and incorporation into sites of vascular injury, the adhesiveness of EPCs may be altered. Exposure of human EPCs to simvastatin causes an increase in the expression of integrins α5β1 and αvβ5, which play a role in angiogenesis.65 These simvastatin-treated EPCs displayed increased incorporation into the neoendothelium of balloon-injured carotid arteries, which was abrogated on integrin receptor blockade.8 These results suggest that homing to and incorporation into foci of ischemic or vascular injury is determined not only by the number of circulating EPCs but also by the adhesiveness of EPCs, which changes during maturation. The mechanisms mediating EPC homing and differentiation are just beginning to be elucidated. Future work toward understanding these pathways may lead to the ability to maximize the efficiency of EPC-mediated NV. Statins may positively modulate vascular repair, limiting neointimal formation and occlusion of diseased vessels, by enhancing EPC function.8

6.EPCs AS BIOMARKERS OF DISEASE PATHOLOGY

Vasa et al. reported an inverse correlation between the number of cardiovascular risk factors and the number and migratory activity of EPCs.64 Smoking reduced EPC numbers, whereas hypertension impaired EPC migration. These authors speculated that smoking and hypertension increased apoptosis of EPCs or interfered with pathways regulating differentiation and mobilization.64 Thus, it has been suggested that the levels of EPCs may serve as a predictor for vascular function and cumulative cardiovascular risk.66

7.EPCs IN REGENERATIVE MEDICINE

With the identification of EPCs as important players in adult NV, several studies have attempted to utilize EPCs from healthy donors to restore blood flow to ischemic tissue. Most have involved the transplantation of EPCs expanded ex vivo. The transplantation of EPCs significantly improved blood flow recovery and capillary density in myocardium67 and in animal ischemic hind limbs.68 Ex vivo expanded human EPCs promote NV of ischemic hind limbs in athymic nude mice. Mice receiving human EPCs have increased capillary density and blood flow in the ischemic limb after transplantation, leading to increased limb salvage.68 Furthermore, EPC transplantation induces blood flow recovery in the ischemic hind limbs of both diabetic mice and rats, suggesting that EPC-mediated NV can still occur under disease conditions, and can thus be applied as a therapeutic treatment in the patients who would benefit most.68,69 Subsequent studies showed that autologous BM cell transplantation in a rat model and BM mononuclear cell transplantation

in a rabbit model of hind limb ischemia were able to improve collateral vessel formation and blood perfusion in the ischemic limb.22,70

Just as progenitor cell transplantation restored blood flow to ischemic hind limbs, EPC transplantation after myocardial infarction also induced NV. Kawamoto et al. demonstrated that transplanted, ex vivo expanded EPCs had a favorable impact on the preservation of left ventricular function.71 After the induction of myocardial ischemia, labeled EPCs were injected intravenously. The EPCs were shown to accumulate in the ischemic area and to participate in myocardial NV. Echocardiography revealed ventricular dimensions that were significantly smaller and fractional shortening that was significantly greater in the EPC transplant. The transplant group also had less ventricular scarring and better-preserved wall motion.72

Similar results were found by Kocher et al. after the transplantation of G- CSF-mobilized CD34+ human cells containing both HSCs and EPCs.31 Transplantation improved myocardial function, prevented cardiomyocyte apoptosis, and limited pathological myocardial remodeling. Endothelial cells of human origin localized at the infarct core, whereas rat-derived endothelial cells contributed to new vessel formation at the infarct border. It is interesting to speculate from these data that BM-derived progenitor cells may directly contribute to in situ vessel formation in the core region and stimulate angiogenic sprouting from existing endothelium at the infarct border. Similar results were found using a porcine myocardial infarction model.72 Three weeks after implantation, regional blood flow, capillary density, and the number of visible collateral vessels were significantly higher in transplant recipients compared with controls. The authors conclude that BM implantation represents a novel and safe strategy to achieve therapeutic angiogenesis. Beneficial effects are likely to be mediated by the transplanted cells incorporating into new vessels and promoting angiogenesis through growth factor expression. Additional mechanisms may include transdifferentiation of EPCs into functionally active cardiomyocytes and assisting with cardiomyocyte regeneration after ischemia.73

Clinical trials assessing safety and feasibility of autologous progenitor cell transplantation are promising. Tateishi-Yuyama and collaborators in the Therapeutic Angiogenesis Using Cell Transplantation Study demonstrated favorable results with autologous implantation of BM mononuclear cells in patients with ischemic limbs.74 Four weeks following random injection of BM mononuclear cells into the gastrocnemius of one leg and peripheral blood mononuclear cells into the other leg as a control in patients with bilateral leg ischemia, various outcomes were measured. Legs into which the BM mononuclear cells were injected had a significantly improved anklebrachial index, an improved transcutaneous oxygen pressure, a significant reduction in rest pain, and an increase in pain-free walking time compared to legs receiving peripheral blood mononuclear cells. Improvements were sustained for six months with no serious complications. The authors attributed the positive outcome to the ability of marrow cells to initiate therapeutic angiogenesis by supplying both EPCs and angiogenic factors.74

Intracoronary infusion of autologous progenitor cells after AMI appears to be safe and effective in limiting post-infarction remodeling processes. The Transplantation of Progenitor Cells and Regeneration Enhancement in Acute Myocardial Infarction (TOPCARE-AMI) Trial was established to assess the safety and feasibility of autologous progenitor cell transplantation in patients with ischemic heart disease.75 At four months, transplantation of progenitor cells resulted in increased global left ventricular ejection fraction, improved regional wall motion in the infarct zone, reduced end-systolic left ventricular

356 S. Caballero et al.

volumes, and increased myocardial viability in the infarct zone compared with a nonrandomized matched reference group. In this study, however, no differences were detected between the blood-derived and BM-derived progenitor cell groups.

In another study, ten patients with AMI were treated by intracoronary transplantation of autologous mononuclear BM cells in addition to standard therapy.76 After three months, the patients receiving cell transplantation showed a significant decrease in infarct region size and a significant increase in infarct wall movement velocity compared with AMI patients treated with standard therapy alone. The results of these studies suggest that intracoronary infusion of progenitor cells can improve outcome after AMI.76

Dobert and co-workers examined the effect of local intracoronary progenitor cell infusion on the regeneration of infarcted cardiac tissue after AMI. They used 18F-fluorodeoxyglucose positron emission tomography (PET) and 201Tl single-photon emission computed tomography (SPECT) to evaluate patients. The patients underwent intracoronary infusion of either BM-derived or circulating EPCs 4 ± 2 days after acute myocardial infarction. There were no significant differences in myocardial viability and perfusion between the two types of infusions. Their results also showed that coronary stenting and transplantation of progenitor cells result in a significant increase in myocardial viability and perfusion.77

Aoki and co-workers utilized a “pro-healing” approach for prevention of post-stenting restenosis rather than cytotoxic or cytostatic local pharmacological therapies. EPC capture stents have been developed using immobilized antibodies targeted at EPC surface antigens. The HEALINGFIM (Healthy Endothelial Accelerated Lining Inhibits Neointimal GrowthFirst In Man) registry is the first clinical investigation using this technology. The investigators in this study demonstrated that the EPC capture coronary stent was safe and feasible for the treatment of de novo coronary artery disease. Their results suggest that further developments in this technology are warranted to evaluate the efficacy of this device for the treatment of coronary artery disease.78

8.CONCLUSIONS

There are considerable data to support the belief that altered EPC function may play a role in vascular disease such as atherosclerosis and diabetic complications associated with reduced re-endothelialization following injury. In this review, we discussed key signaling factors, with special emphasis on SDF-1, which are involved in orchestrating the stem cell-driven repair process of the vasculature. Many factors are known for their