Update on therapeutic vascularization strategies

Abstract

The ability to exploit angiogenesis and vascularization as a therapeutic strategy will be of enormous benefit to a wide range of medical and tissue-engineering applications. Angiogenic growth factor and cell-based therapies have thus far failed to produce a robust healing response in clinical trials for a variety of ischemic diseases, while engineered tissue substitutes are still size-limited by a lack of vascularization. The purpose of this review is to investigate current research advances in therapeutic vascularization strategies applied to ischemic disease states, tissue engineering and regenerative medicine. Recent advances are discussed that focus on better regulation of growth factor delivery and attempts to better mimic natural processes by delivering combinations of multiple growth factors, cells and bioactive materials in the right spatial and temporal setting. Some unconventional approaches and novel therapeutic targets that hold significant potential are also discussed.

A number of disease states are related to reduced vascular perfusion and could be treated with provascularization strategies. Peripheral vascular disease, ischemic heart disease, wound healing, and cell and tissue transplantation are just a few examples that would greatly benefit from proangiogenic therapies. Aging patients also suffer from slower healing responses, especially after surgery and during fracture repair [1]. Proangiogenic therapy is likely to improve the healing response of elderly patients in many situations. While research in the field of tissue engineering has been active for several decades, there are still relatively few effective clinical implementations of tissue engineering technology. Clinical use of engineered tissues and tissue substitutes is largely limited to avascular or thin tissue types such as cartilage, bladder and skin. Great progress has been attained in the development of a variety of engineered tissue types that function on a small scale in vitro, but these ultimately suffer from a lack of vascular perfusion when scaled up to a size relevant for implantation and disease treatment. Researchers have been working hard to overcome this limitation and are developing a number of innovative strategies for vascularizing engineered tissues.

Human trials in therapeutic vascularization have had limited success to date despite promising results in animals with experimentally induced ischemia [2]. Most therapeutic clinical trials have focused on delivery of a single gene or growth factor (GF), commonly various isoforms of VEGF or FGF, as discussed in detail in several review articles [3-7]. Cell transplantation therapy has also been tested clinically to treat myocardial and peripheral ischemia with promising results, reviewed in [5,7-10]. The US FDA has designated the primary end point for approval of an angiogenic agent as an improvement in exercise performance. Several clinical trials have demonstrated angiogenic therapy to alleviate certain secondary symptoms such as chest pain, but failed to demonstrate a significant difference in exercise performance. Part of the reason for this lack of enhanced performance is a strong placebo effect in control groups and difficulty in selecting an ideal patient cohort [11]. However, the results of Phase I and II trials also indicate that proangiogenic gene and protein therapy is generally safe and feasible. Critics have expressed a number of concerns for proangiogenic therapy including the potential for triggering growth of latent tumors, increasing the risk for retinopathy and promotion or destabilization of atherosclerotic plaques. In studies to date, none of these potential issues has been noted to be above baseline for the study population. Some less serious side effects associated with administration of VEGF and FGF-2 include hypotension, vascular leakage, transient tissue edema and renal insufficiency. Clinicians continue to express optimism for the future of proangiogenic therapy and a number of promising new strategies are currently under investigation. At present, a large focus of clinical and preclinical studies is focused on identifying the ideal angiogenic agent (or combination therapy), delivery strategy and dosing regimen.

Review of angiogenesis biological mechanisms

The formation of the vascular network is a finely tuned and complex process controlled by the signaling balance between integrins, angiopoietins, chemokines, junctional molecules, oxygen sensors, endogenous inhibitors and many others [12]. For a detailed description of the mechanisms of angiogenesis, the reader is referred to some excellent reviews [13,14]. There are three main mechanisms of new blood vessel growth in adults: ischemia-induced angiogenesis, arteriogenesis and vasculogenesis. Ischemia-induced angiogenesis is the best understood mechanism and involves the growth of new microvasculature and capillary beds, usually driven naturally in adults by hypoxia or wound healing but also activated in the growth of tumors. Oxygen in tissue is monitored by hypoxia-inducible factor (HIF)-1, which comprises α and β subunits. The HIF-1β subunit is relatively stable while the HIF-1α subunit is targeted for rapid degradation by the oxygen sensitive von Hippel-Lindau pathway. At low oxygen levels, HIF-1α degradation is impaired, leading to increased HIF-1 heterodimerization, DNA binding and transcription of proangiogenic genes. HIF-independent pathways also exist. For example, peroxisome-proliferator-activated receptor-γ coactivator (PGC)-1α, a major regulator of mitochondrial function in response to exercise or low-nutrient environments, is able to exert strong control over the VEGF gene and induce angiogenesis [15].

In sprouting angiogenesis, new blood vessels are formed by sprouting and migration from existing capillaries. Migrating endothelial cells lead the growth of a new sprouting vessel by degrading the basement membrane and laying down a provisional extracellular matrix. Endothelial cells proliferate and form connections to neighboring vessels, develop lumens and split existing vessels though a process known as intussusception [16]. An intricate process of sprouting, branching, intussusception and regression plays out to shape the ultimate pattern of a capillary plexus. Following tubule formation, newly formed vessels are sealed by the formation of endothelial cell–cell junctions and stabilized by the addition of mural cells or pericytes and smooth muscle cells. Sprouting vessels grow down gradients of VEGF and other factors led by an endothelial tip cell, which is tuned to a variety of positive and negative signals coming from the surrounding matrix and other cells. The process of sprouting angiogenesis is illustrated in Figure 1. This process appears to by driven in a very similar way to neuron guidance and many of the neural guidance proteins are being discovered to play pivotal roles in vascular morphogenesis [16-18]. The four major families of guidance proteins in neural and vascular sprouting are semaphorins, ephrins, slits and netrins, along with their associated receptors. The regulation of neural and vascular guidance is a fascinating field of study, but is outside the focus of this review.

Figure 1. Sprouting angiogenesis.

(A) Endothelial cell is activated in response to a growth factor gradient; tip cell is selected. (B) Tip cell degrades basement membrane and migrates down gradient, following positive and negative guidance cues in the matrix; stalk cells form behind leading tip cell. (C) Guidance cues control pathfinding and morphogenesis, large vacuoles form in stalk cells and merge to make tubules. (D) Recruitment of pericytes and deposition of basement memebrane stabilizes newly formed endothelial tubules.

The VEGF family of GFs has been identified as critically important to angiogenesis and comprises several isoforms. VEGF-A (commonly referred to as simply VEGF) is a potent stimulator of endothelial cell mitogenesis, cell migration, vasodilation and a mediator of microvascular permeability. PlGF, another potent VEGF isoform, is capable of stimulating angiogenesis and collateral vessel growth while avoiding side effects common to VEGF such as edema and hypotension [19]. While being a strong vascular stimulator, VEGF signaling primarily serves to drive expansion of the capillary bed and it appears that VEGF administered alone has limited ability to induce the growth of larger vessels. Long-duration exposure is necessary to produce stable microvasculature that does not resorb after withdrawal of the VEGF stimulus [20]. Since VEGF-driven angiogenesis does not consistently result in the formation of functional, stable vasculature, VEGF therapy alone may not be the ideal treatment for increasing blood perfusion to an ischemic tissue [3,21]. Administration of additional supporting GFs such as FGF-2, angiopoeitins, PlGF or PDGF [21] at the appropriate time point help to stabilize newly formed endothelial tubules by recruiting pericytes and smooth muscle cells, perhaps resulting in a more predictable therapeutic response.

In contrast to ischemia-induced angiogenesis, arteriogenesis usually occurs proximally to the ischemic tissue and is often linked to atherosclerosis when it occurs naturally in adults, although the specific mechanisms are less well understood. Mechanical stresses at the site of arterial stenosis activate endothelial cells [22], upregulating expression of the monocyte chemoattractant protein (MCP)-1, FGF, PDGF, VEGF, matrix metalloproteinase (MMP), cell-adhesion molecules and nitric oxide (NO). This mixture of factors promotes inflammatory and remodeling responses. MCP-1 attracts monocytes that enter the vessel wall and differentiate to macrophages, producing inflammatory cytokines such as TNF-α. Studies of hind-limb ischemia in animals indicate that arteriogenesis is more likely to occur by widening of existing vessels rather than formation of new vessels [23], although this has not been confirmed for all situations. The process of arteriogenesis has the ability to increase blood flow to distal tissues 20–30 times [11], a much larger volume of flow than possible by microvascular expansion. A few large vessels are capable of delivering a much larger volume of blood than many small high-resistance capillaries according to the Poiseuille's law [24] and, as such, therapeutic vascularization should aim to develop collateral vessels. At the same time, capillary formation cannot be ignored as it is crucial to improving oxygenation and gas exchange in ischemic tissues [25] and it may prove true that stimulation of both capillaries and collateral vessels is required for the best healing response.

Vasculogenesis, once believed to only occur in the developing embryo, refers to the de novo generation of new vessels from the migration and differentiation of endothelial progenitor cells (EPCs) found in the bone marrow and circulation. Vasculogenesis in adults has been an area of intense study and controversy in the recent literature with conflicting studies published as to the sourcing and contribution of progenitors in the growth of new vessels. Some evidence indicates that progenitor cells do not directly contribute to the endothelium and are located only perivascularly [26]. Nevertheless, a growing body of preclinical [27] and clinical trials [28,29] indicates that transplantation of blood-derived or bone marrow-derived progenitor cells beneficially affects cardiac function after myocardial infarction as well as promotes neovascularization in ischemic tissue. An increasingly well-characterized cell population expressing CD34 as well as other markers and considered to include a hematopoietic cell population is known as EPCs. These endothelial progenitors have the ability to home to sites of neovascularization and differentiate into endothelial cells [30].

Growth factor delivery

Early Phase I and II human clinical trials with VEGF, FGF and hepatocyte GF were promising, but larger randomized placebo-controlled trials failed to demonstrate significant benefits in the approved end points. One of the proposed problems associated with the limited success of early clinical trials is insufficient delivery or dosing of GFs due to a short half-life in the body. Bolus injection of soluble GF, either systemically or locally administered by one of the classical delivery methods including intra-arterial administration, systemic administration (intravenous) and direct intramuscular injection, results in a rapidly depleted, poorly controlled delivery with the resulting vascular growth often disorganized, poorly functional and transient. New studies are aimed at developing a better understanding of the specific concentration ranges, gradients and exposure duration to elicit a more controlled angiogenic response.

The dosage response of VEGF is highly sensitive: low doses result in increased vascular permeability and overdoses result in hemangioma formation [31] and fatal vascular leakage [31]. Most research demonstrates that sustained stimulation with a high level of VEGF is required for formation of stable vasculature in vivo [32], but optimal dosing schedules remain unproven [33]. Researchers are investigating ways to regulate the exposure time [34] and local delivery dosage of angiogenic GFs, as well as optimize safe and effective dosages for use in humans. In mice, VEGF dosages of 150 ng/day delivered by osmotic pump consistently induced high degrees of vascularization with vessels stable for at least 80 days after withdrawal of the GF [15]. Vessels induced by higher concentrations of VEGF resorbed within 20 days of GF withdrawal while lower concentrations of VEGF failed to induce a high degree of vascularization. Optimal dosing in human patients may be difficult to determine since physiologically relevant dosages can differ from animal models and patients in need of angiogenic therapy may have an impaired angiogenic response due to various disease states and metabolic disorders whereas most animal models use healthy test subjects.

Besides dose response, scientists are also discerning the role of directionality and gradient in angiogenic signaling. A microcarrier-based angiogenic sprouting assay, an effective in vitro screening technique [35,36], was recently used to examine the effects of VEGF gradient and concentration on endothelial sprouting [37]. Endothelial cells seeded on microcarriers sprouted tubules aligned with VEGF presented in a gradient, with maximal alignment occurring in a 0–100 ng/ml gradient. It was also seen that cells at the tip of the sprout bound significantly more VEGF then cells in the body of the sprout. At very high VEGF concentrations, binding was saturated and sprouting directionality was lost. This finding supports previous data that the tip cells of vascular sprouts aid in directionality and propagation of new microvessels down a GF gradient [38]. When translated in vivo to a hind-limb ischemia model, scaffolds containing gradients of VEGF were better at restoring perfusion and avoiding necrosis than nongradient controls. It is thought that forces such as interstitial flow contribute to the formation of natural GF gradients in vivo and have been investigated for directing vascular morphogenesis [39-41]. When considered in the context of vascular architecture, gradient formation makes sense to drive directionality of vessel growth. Naturally derived vascular beds have an organized hierarchical structure while tumor vasculature is randomly organized. Gradient delivery of GFs may be a means of promoting better architecture in induced vascular beds.

While VEGF and FGF have received much attention in the literature, PlGF may prove worthwhile to study in greater detail. PlGF has been shown to be equally as potent as VEGF at promoting angiogenesis while also promoting arteriogenesis [42]. PlGF actually promotes increased expression of VEGF in ischemic tissue and its effects are increased when administered in synergy with VEGF [43]. Notably, PlGF may be a better candidate for arteriogenic therapy than FGF or PDGF, which preferentially recruit either endothelial and mural cells or inflammatory cells, respectively, while PlGF can recruit all three cells types on its own. All three cells types are cited as required for arteriogenesis [44].

Research has demonstrated that controlled GF release, correct dosage and exposure time is highly beneficial to inducing the growth of functional vasculature in animal models of ischemia [45,46]. Stabilization and molecular regulation of nascent blood vessels is also critical to the achieving a correct angiogenic response [47]. Preclinical data demonstrate that delivery of basic FGF or VEGF eventually results in unstable vessel growth that resembles immature tumor vasculature [48]. Tumor vessels are chaotic and do not follow the hierarchical branching pattern of normal vascular networks [14,47]. Several novel approaches are under investigation to more precisely regulate the spatial temporal presentation of the right combination of factors to induce more appropriate vessel architecture and stability [49]. One such approach is the design of polymeric scaffolds that stagger exposure to two or more GFs such as VEGF and PDGF [49,50], FGF and HGF [51], or VEGF and angiopoietin-1 [52] to create more stable vessels with regular architecture by more closely mimicking the biological mechanism of vessel induction followed by stabilization. Many other stabilization and growth ‘on’ and ‘off’ signals exist, including activators and inhibitors of the Wnt and Notch pathways [53] and neural guidance cues (e.g., netrin and ephrin). More research is needed in the area of branching morphogenesis [54] to solve the intricate programming structure of vascularization.

A different approach to single or multiple staggered GF delivery is to induce the expression of upstream activators of a large number of angiogenic regulators such as HIF-1α. Since HIF-1α is naturally degraded, a number of techniques have been employed to ensure its stabilization in vivo. Delivery of the gene for a physiologically stable HIF-1α/VP16 fusion protein has been demonstrated to promote recovery of peripheral limb ischemia in animals [55,56] and was recently tested in a Phase I trial with a percentage of patients with peripheral limb ischemia showing pain resolution and ulcer healing [57]. Another method for HIF-1α stabilization is expression of a mutant HIF-1α that lacks an oxygen degradation domain [58].

There are certain disadvantages to delivering recombinant proteins, including a short half-life in vivo and expense and difficulty in manufacturing. Scientists have investigated using gene therapy to deliver a more prolonged and targeted delivery of angiogenic GFs. Both viral and nonviral vectors have been tested clinically for delivery of angiogenic factors, with results similar to those seen by delivering GFs. Owing to safety concerns with viral vectors, nonviral vectors are initially more attractive for human use. However, nonviral vectors are generally less efficient at inducing expression. Novel gene therapy delivery mechanisms are focused on improving nonviral vector delivery. Electroporation has recently been used to deliver complementary DNA for HIF-1α to improve wound healing in elderly diabetic mice [59]. Other delivery mechanisms such as liposome complexes [60] are under investigation to improve delivery of nonviral vectors and have been tested preclinically for delivery of angiogenesis related genes. Delivery of VEGF-activating transcription factors is also under investigation [61,62]. The concept of engineered cell therapy has been proposed as a solution to problems with direct gene therapy and is briefly discussed further in the next section.

Cell therapy

Another approach to treating ischemic disease and promoting angiogenesis is the transplantation of autologous cells, which has been studied in large animals [27,63] and in clinical trials [28,29]. Some of the cell sources under investigation include bone-marrow stromal cells, mesenchymal stem cells and EPCs. A number of trials under development and currently ongoing will test the safety and efficacy of autologous cell transplantation in a variety of ischemic diseases (Table 1). Some of the early cell transplantation trials used skeletal myoblasts and had mixed results at treating ischemic heart disease [28,64-66]. This may be due to a reduced ability of myoblasts to promote neovascularization and reperfusion. Transplantation of less-differentiated cell types, such as marrow-derived stromal cells, vascular/EPCs and mesenchymal stem cells, may lead to a stronger healing response. Transplantation of EPCs has shown success in treating peripheral limb ischemia by promoting collateral vessel formation in both humans and animals [67,68], as well as treating myocardial ischemia [69,70]. Transplanted mesenchymal stem cells and marrow stromal cells promote wound healing through differentiation and release of proangiogenic factors [71]. EPCs have also been studied for improving bone regeneration and healing [72]. Current myocardial delivery mechanisms for cell-based therapy that have proven somewhat effective include intracoronary and intramyocardial delivery [8]. The population of EPCs found in the bone marrow, circulation or other tissues, first isolated primarily by Asahara and colleagues in 1997 [73], is one of the most promising cell sources owing to their regenerative capacity and ability to home to sites of ischemia. One comparative study between mesenchymal progenitors and endothelial progenitors for myocardial infarct regeneration showed better neovascularization and contractility for treatment with endothelial progenitors over mesenchymal cells [74]. The EPC population is described in greater detail in [30,75]. Stem cell homing has also been investigated for gene delivery [76].

Table 1.

Selection of ongoing and recently completed clinical trials for treatment of ischemic conditions.

Therapy	Title	Conditions	Interventions	Phase	Enrollment	NCT ID	Ref.
Cell	Stem cell study for patients with heart disease	Coronary artery disease	Genetic: Cell therapy – autologous CD34+ cells	I	24	NCT00081913	[125]
Cell	EPC by intracoronary injection in patients with chronic stable angina	Coronary artery disease	Device: Vescell™ – autologous EPCs/angiogenic cell precursors	II	24	NCT00384514
Cell	Bone marrow transfer to enhance ST-elevation infarct regeneration	Myocardial infarction	Procedure: Intracoronary bone marrow cell transfer	I	60	NCT00224536	[126]
Cell	Intramyocardial injection of autologous aldehyde dehydrogenase-bright stem cells for therapeutic angiogenesis	Coronary artery disease	Procedure: Stem cell injection under electromechanical guidance	I	60	NCT00314366
Cell	Autologous stem cells for cardiac angiogenesis	Ischemic heart disease	Device: Intramyocardial injection of stem cells via noga mapping	I	30	NCT00203203
Cell	Combination stem cell therapy for the treatment of severe leg ischemia	Limb ischemia	Biological: Mesendo	I	10	NCT00518401
Cell	Stem cell study for patients with leg ulcer/gangrene	Limb ischemia	Genetic: Autologous peripheral blood CD34+ cell therapy	I, II	15	NCT00221143	[127]
Cell	Autologous bone marrow for lower extremity ischemia treating	Limb ischemia	Procedure: Bone marrow aspiration injection of isolated CD133+ cells	II	42	NCT00753025
Cell	ACPs in severe PAD/CLI by direct intramuscular injection	Limb ischemia	Procedure: ACPs or Vescell	I	6	NCT00523731
Cell	Autologous bone marrow transplanted via transendocardial catheter to chronic myocardial infarct border zone	Limb ischemia	Procedure: Transendocaridal transplantation of autologous bone marrow	I	20	NCT00507468
Cell	Autologous stem cell transplantation in acute myocardial infarction	Myocardial infarction	Gene transfer: Intracoronary autologous stem cell transplantation	II	100	NCT00199823	[128]
Cell	CD133+ autologous cells after myocardial infarction	Myocardial infarction	Procedure: CD133+ cell intracoronary administration	I, II	15	NCT00400959
Cell	Autologous bone marrow transplanted via transendocardial catheter to chronic myocardial infarct border zone	Limb ischemia	Procedure: Transendocaridal transplantation of autologous bone marrow	I	20	NCT00507468
Cell	Autologous stem cell transplantation in acute myocardial infarction	Myocardial infarction	Gene transfer: Intracoronary autologous stem cell transplantation	II	100	NCT00199823	[128]
Cell	CD133+ autologous cells after myocardial infarction	Myocardial infarction	Procedure: CD133+ cell intracoronary administration	I, II	15	NCT00400959
Cell	Autologous bone marrow transplanted via transendocardial catheter to chronic myocardial infarct border zone	Limb ischemia	Procedure: Transendocaridal transplantation of autologous bone marrow	I	20	NCT00507468
Cell	Autologous stem cell transplantation in acute myocardial infarction	Myocardial infarction	Gene transfer: Intracoronary autologous stem cell transplantation	II	100	NCT00199823	[128]
Cell	CD133+ autologous cells after myocardial infarction	Myocardial infarction	Procedure: CD133+ cell intracoronary administration	I, II	15	NCT00400959
Cell	Bone marrow cells in myocardial infarction	Myocardial infarction	Procedure: Intracoronary injection of bone marrow cells	II, III	100	NCT00363324
Cell	Stem cell therapy to improve myocardial function in patients undergoing CABG	Myocardial infarction	Procedure: Bone marrow stem cell therapy combined CABG	I, II	50	NCT00395811
Cell	Efficacy study of intramuscular or intracoronary injection of autologous bone marrow cells to treat scarred myocardium	Myocardial infarction	Control procedure: intramuscular administration of bone marrow cells procedure: intracoronary administration of bone marrow cells	II	63	NCT00560742
Cell	REPAIR-AMI: intracoronary progenitor cells in AMI	Myocardial infarction	Drug: Intracoronary infusion of enriched bone marrow-derived progenitor cells	III	200	NCT00279175	[129]
Cell	Safety study of bone marrow derived cells to treat damaged heart muscle	Myocardial infarction	Drug: NX-CP105	I	18	NCT00361855
Cell	Safety study of adult MSCs to treat acute myocardial infarction	Myocardial infarction	Drug: Provacel™	I	48	NCT00114452	[130]
Cell	Bone marrow-derived stem cell transfer in acute myocardial infarctions	Myocardial infarction	Procedure: Bone marrow-derived stem cell transfer	II	68	NCT00264316	[131]
Cell	Myocardial stem cell administration after acute myocardial infarction (MYSTAR) study	Myocardial infarction	Procedure: Bone marrow-derived stem cells implantation	II	116	NCT00384982
Cell	Intracoronary stem cells in large myocardial infarction	Myocardial infarction	Procedure: Intracoronary infusion of autologous bone-marrow derived stem cells	II		NCT00389545
Cell	Long-term follow-up of autologous bone marrow mononuclear cells therapy in STEMI	Myocardial infarction	Procedure: Saline infusion. Procedure: Autologous bone marrow mononuclear cells infusion	I, II	37	NCT00626145
Cell	ACT34-CMI adult autologous CD34+ stem cells	Myocardial infarction	Device: Stem cell injection	II	150	NCT00300053
Cell	Stem cell therapy in chronic ischemic heart failure	Myocardial infarction	Procedure: Bone marrow transplantation	II	35	NCT00235417
Cytokine	Stem cell mobilization to treat chest pain and shortness of breath in patients with coronary artery disease	Coronary artery disease	Procedure: Stem cell mobilization	II	35	NCT00043628
Cytokine	A Phase I–II safety study of filgrastim (Neupogen®) to improve left ventricular function after severe acute myocardial infarction	Myocardial infarction	Drug: Filgrastim	I, II		NCT00215124	[132]
Cytokine	Stem cells in myocardial infarction	Myocardial infarction	Drug: G-CSF (Neupogen)	II	78	NCT00135928
Cytokine	Myocardial regeneration and angiogenesis in myocardial infarction with G-CSF and intracoronary stem cell infusion-3-DES	Myocardial infarction	Drug: G-CSF Procedure: Collection of mobilized peripheral blood stem cells Procedure: Intracoronary infusion of mobilized cells	II	96	NCT00291629	[133]
Cytokine	Bone marrow stem cell mobilisation therapy for AMI (REVIVAL-2)	Myocardial infarction	Drug: G-CSF Other: Placebo	IV	114	NCT00126100	[134]
Cytokine	Effects of recombinant human erythropoietin on platelet function in patients with acute myocardial infarction	Myocardial infarction	Drug: recombinant human erythropoietin-α	II	48	NCT00367991
Gene	The effect of mobilized stem cell by G-CSF and VEGF gene therapy in patients with stable severe angina pectoris	Ischemic heart disease	Gene transfer: VEGF-A165 plasmid	I, II	48	NCT00135850
Gene	VEGF gene transfer for critical limb ischemia	Limb ischemia	Genetic: pVGI.1 (VEGF-2)	I		NCT00304837

ACP: Angiogenic cell precursor; AMI: Acute myocardial infarction; CABG: Coronary artery bypass grafting; CLI: Critical limb ischemia; DES: Drug-eluting stent; EPC: Endothelial progenitor cell; G-CSF: Granulocyte-colony stimulating factor; MSC: Mesenchymal stem cell; PAD: Peripheral artery disease; REPAIR-AMI: Reinfusion of Enriched Progenitor And Infarct Remodeling in Acute Myocardial Infarction. Data from [201].

An alternative to delivering cells directly is the administration of cytokines that attract progenitor cells to sites of ischemia. After myocardial infarction there is a mobilization of bone marrow-derived stem/progenitor cells to the circulation [77,78]. Clinical trials investigating the use of granulocyte colony-stimulating factor, a cytokine shown to promote the mobilization of bone marrow stem/progenitor cells and subsequent accumulation in ischemic tissue, have so-far failed to demonstrate significant beneficial effects in the ischemic heart [79-81]. However, other techniques are underway to capture or home endothelial progenitors to sites of ischemia. Systemic administration of anti-α4 integrin antibody was recently demonstrated to promote the mobilization and functional incorporation of bone-marrow EPCs [82], in a method that mimics observed downregulation of α4 integrin in natural progenitor cell mobilization [83,84]. Asahara and colleagues recently discovered a key piece of the endothelial progenitor homing puzzle, demonstrating that specific Jag-1-derived Notch signaling is required for endothelial progenitor-mediated vasculogenesis [85], although much work remains to be carried out in this area. According to a recent review article, endothelial progenitors hold much promise for treatment of ischemic disorders, and better techniques for cell isolation, expansion, mobilization, recruitment and transplantation are under development [86].

Engineered cell therapy is one prospective method to resolve problems related to direct gene delivery in humans. Engineered cell therapy to promote vascularization is growing in popularity and has been used to deliver non-naturally secreted proteins to the heart [87-89] and for delivery of VEGF in tissue-engineered bone repair scaffolds [90].

Cell transplantation is also being investigated as a means of vascularizing tissue-engineered constructs. Several examples have been published of microvascular tubules forming in implanted constructs containing endothelial cells. The most successful constructs, in terms of linkage to host vasculature, vessel architecture and vessel longevity, include at least two cell types in an appropriate matrix that allows for cell migration and promotes the endothelial cell phenotype. A particularly successful model constructed by Jain and colleagues incorporated a co-culture of human umbilical vein endothelial cells and 10T1/2 mesenchymal precursor cells in a fibronectin-type I/collagen gel implanted in a mouse [91]. Patent vasculature connected to the host circulatory system was formed that was stable for at least a year and was responsive to the vasoconstrictor endothelin. In other studies, human dermal microvascular endothelial cell spheriods and preadipocytes were transplanted into a fibrin matrix on a chick chorioallentoic membrane [14] and fibroblast sheets co-cultured with endothelial progenitors were shown to improve cardiac function in infracted hearts [92]. In these examples the transplant formed a patent microvasculature connected to the host system without exogenous angiogenic GFs or transient transfection. Incorporating appropriate autologous cell types in well-designed matrices may prove a safe and effective means of vascularizing tissue-engineered implants, and has been used for developing engineered vascularized skeletal muscle [93].

Matrix interaction

Much progress has been made towards the development of vascular-inductive tissueengineering matrices. Early work in this area involved passive adsorption or bulk incorporation of GFs in porous or degradable scaffolds. While somewhat effective at producing initial vascular growth, quick release profiles and rapid diffusion of GFs do not result in the desired response of functional, stable vasculature. Researchers are investigating covalently tethering GFs to matrices and incorporating GF release mechanisms tied to angiogenic activity such as MMP-degradable sites [94,95]. In addition to GFs, extracellular matrix proteins that regulate factors such as cell adhesion and migration have an impact on cell function and gene regulation. Different extracellular matrix proteins and ligands have been shown to modulate the angiogenic response [96], often through integrin activation [97,98]. Recent research has taken an integrative approach to combine angiogenic GFs with the appropriate matrix signals to create controlled biomimetic analogs to natural extracellular matrices.

A new generation of bioactive materials is under development, designed to more closely mimic the natural extracellular matrix while providing greater control over the cellular response [99].

Fibrin matrices loaded with a modified version of VEGF that directly binds fibrin and is subsequently locally released in a proteolyticdependent manner have been shown to induce local and controlled blood vessel growth in animal models [100,101]. Recent advances have been made in the design of bioartificial matrices, or artificial scaffold materials that incorporate bioactive motifs, illustrated in Figure 2. Conjugation of bioadhesive signals such as the integrin-binding peptide RGD to surfaces of artificial materials has long been established. Hubbell and West have succeeded in developing 3D artificial matrices that incorporate adhesive signals as well as GFs such as VEGF [101,102] and epidermal GF [103]. The concept of a bioartificial matrix is attractive because it allows the presentation of a controlled and tailored environment. In the case of a polyethylene glycol-based matrix, intrinsically resistant to nonspecific protein adsorption and cell adhesion, biofunctionality can be built onto a ‘clean-slate’ background material. This system can be used to test the functionality of diverse proteins in a controlled environment, such as ephrin, which was shown to promote the formation of endothelial tubules without additional adhesive signals [104]. Another attractive element of bioartificial matrices is the ability to spatially control presentation of bioactive ligands with fine precision. Photopatterning techniques have been utilized to construct 3D patterned matrices [105-107]. Geometric constraints such as line width have been demonstrated to effect formation of endothelial tubules [108]. In theory, photopattening techniques could be used to control the architecture of a vascular network and has been investigated on 2D surfaces by our laboratory.

Figure 2. Bioartificial matrix.

A bioartificial matrix contains a number of bioactive components that enables an artificially synthesized material such as PEG hydrogel to exhibit properties of natural extracellular matrix in a controlled manner. The bioactive components may include MMP-degradable crosslinkers, adhesive ligands such as RGD, and growth factors. A-PEG groups allow for photopolymerization of modular components to form a matrix.

A-PEG: Acrylate-PEG; MMP: Matrix mettaloproteinase.

The concept of photopatterning guidance channels mimics the natural phenomenon of vascular memory. Vascular memory, important to the study of tumor angiogenesis, occurs when regressed microvasculature leaves behind empty sleeves of basement membrane and associated pericytes [109]. When conditions permit revascularization, such as reversal of anti-VEGF therapy, a rapid repopulation of vessels along basement membrane sleeves is observed [109,110], similar to the way nerves regenerate along pre-existing pathways [111]. Vascular memory has also been noted to occur in normal human vasculature [112,113]. Further study of vascular and neural pathfinding in development and healing will undoubtedly produce future medical benefits.

Metabolic control

Some novel unconventional approaches have been proposed to increase vascular perfusion in ischemic tissue. Scientists looking for new approaches have been driven mainly by technical difficulties in delivering proangiogenic factors locally to ischemic tissue for sufficiently long periods of time, often confounding requirements for multiple factors delivered in the right spatial and temporal dosage, and the risk of undesirable side effects from systemic administration of angiogenic factors. One novel approach capitalizes on a large body of research that implicates NO as a stimulator of angiogenesis [114-116]. NO increases the expression of VEGF and other angiogenic factors, recruits pericytes and improves blood perfusion by inducing vessel dilation. NO can also protect tissue against ischemic damage by reprogramming the cell's metabolism to tolerate a lower oxygen environment through nitrosylation of oxygen sensor Phd1 proteins and complex I proteins in the electron-transfer chain to reduce the generation of proinflammatory reactive oxygen species. It is important to note that in higher concentrations, or in certain physiological conditions, NO can have the opposite effect. Kumar and coworkers demonstrated the use of nitrite (NO²⁻) as a proangiogenic molecule, considering the fact that nitrite is reduced to NO in ischemic conditions but is oxidized into harmless nitrate (NO³⁻) in oxygenated tissue [117]. They demonstrated that administration of sodium nitrite by intraperitoneal injection significantly restored ischemic hind-limb blood flow, vascular density and endothelial cell proliferation. Administration of carboxy-PTIO, a NO scavenger, abolished the positive effects, bolstering their argument that NO is responsible for the changes. These results are exciting since they open the possibility of an effective, yet inexpensive treatment option with few side effects and simple administration options allowing for long-term treatment.

Traditional angiogenic factors such as VEGF and FGF stimulate vascularization but do not promote additional metabolic changes [118]. Recent studies are solidifying the link between metabolic demand for oxygen and angiogenesis. Expression of the transcriptional coactivator PGC-1α, a potent metabolic sensor and regulator, is induced by a low-nutrient and low-oxygen environment such as during aerobic exercise training. PGC-1α was recently demonstrated to be a potent stimulator of VEGF expression angiogenesis in ischemic tissues but does not involve the traditional HIF-1 pathway [119]. Activating the PGC-1α pathway may be a novel method of inducing angiogenic response. Another interesting idea is to induce tolerance of a hypoxic environment to stave off tissue damage and necrosis after an infarction [120], allowing time for a subsequent proangiogenic therapy to take effect. Aragones and coworkers show that Phd1, an oxygen-sensitive enzyme that controls the stability of HIFs, plays a key role in regulating metabolism and reducing oxygen requirements in muscle tissue [120]. While it might appear counter-intuitive to induce hypoxia tolerance, as Phd1-null mice showed reduced exercise performance, the benefit lies in reducing damage. Oxidative consumption under hypoxic conditions has been linked to generation of reactive oxygen species and limiting oxygen consumption when oxygen levels are low leads to less cell death [121]. Another potential metabolic control involves the experimental concept of ischemia preconditioning, where tissues are rendered resistant to ischemia/reperfusion injury, and was first shown to result in reduction of myocardial infarct sizes in 1986 [122]. Ischemia precondition is traditionally induced by brief repeated episodes of coronary artery occlusion prior to a major prolonged occlusion. New research has demonstrated that ischemic preconditioning results in a robust activation of HIF-1α and that pretreatment with the HIF activator dimethyloxalylglycine or selective siRNA repression of Phd2 resulted in cardioprotection similar to that of traditional ischemic preconditioning [123]. One could imagine a treatment scenario where hypoxia tolerance is initially induced, followed by switching on angiogenesis mechanisms and gradually restoring oxygen consumption back to normal levels as the tissue is revascularized. These new studies demonstrate that targeting oxidative metabolism and capitalizing on its natural link to angiogenesis is a valid and novel therapeutic tool for treating ischemic disease.

Conclusion

The therapeutic vascularization field experienced initial setbacks in the failure of several clinical trials with GF therapy going quickly from benchtop to bedside before precise delivery methods and controls were elucidated. A growing body of evidence is beginning to suggest that single GF therapy may be insufficient for restoring a large volume of blood flow to an ischemic tissue, especially in a diseased state where vessels may be less responsive. A great deal of effort has gone into identifying upstream activators that induce expression of multiple angiogenic factors as well as alternative pathways to the traditional angiogenic signaling paradigm. Cell-based therapy looks very promising, but continues to suffer from optimal sourcing and delivery strategies. A new generation of therapies is needed to better address the problems with current angiogenesis strategies.

Future perspective

Future approaches to therapeutic vascularization will have a better chance of success if a more targeted and careful approach than that of earlier trials is utilized. A clinically successful vascularization therapy is likely to be a combination therapy. Examples of combination therapies such as engineered cell therapy, multiple GF release, or growth factor–cell–matrix combinations [124] are beginning to emerge. We must also consider the importance of inducing arteriogenesis, angiogenesis, neovascularization or some combination. Future therapies need to consider the end-goal in terms of reperfusion and what type of therapy or combination strategy is best-suited to obtain both the necessary increase in blood-flow volume and ability to deliver oxygen. Ongoing research is helping us to understand the complex signaling and pathfinding mechanisms that lead to branching morphogenesis. Knowledge divulged from studies on development of vascular architecture and pathfinding, such as GF gradients, vascular memory and morphogenesis signaling, needs to be applied to create more physiologically accurate and better functioning induced vasculature. More work is needed to sort out the extremely complex angiogenic molecular program. Metabolic control over angiogenesis is an interesting new approach and further research in this area may provide alternative therapies to patients who do not respond to traditional GF therapy or that have metabolic disease-related ischemic disorders. Additionally, advances in engineered bioartificial and bioactive matrices are illustrating higher levels of control in tissue engineering and regenerative medicine, paving the way towards large-scale engineered tissue substitutes. As work progresses in all of these areas, we can expect more functional therapies to become available to patients. Based on the impressive progress thus far, we can expect the future of therapeutic vascularization to be quite bright.

Executive summary.

Biological mechanisms

• ■ Sprouting angiogenesis is the best understood mechanism of new blood vessel growth in adults and involves the formation of new capillaries in response to ischemia or other stimuli.

• ■ Arteriogenesis involves widening of existing arteries and formation of collateral vessels triggered by mechanical stresses often linked to atherosclerosis. Arteriogenesis is capable of restoring a large volume of blood flow to a distal ischemic tissue, although will not necessarily increase oxygen delivery without simultaneous capillary expansion in the ischemic tissue itself.

• ■ Vasculogenesis refers to de novo generation of vessels from migrating or transplanted endothelial progenitor cells.

Growth factor delivery

• ■ Early trials with VEGF, FGF and HGF were promising but failed to demonstrate significant benefit in exercise stress tests.

• ■ Better understanding of growth factor (GF) dosage and spatial and temporal control of delivery are needed to produce better clinical results.

• ■ New strategies stagger exposure of inductive followed by stabilizing factors to generate more functional vasculature.

• ■ Delivery of upstream activators that engage more components of the angiogenic machinery such as hypoxia-inducible factor (HIF)-1α and PlGF may provide novel therapies.

• ■ Gene therapy remains an option to combat problems with protein delivery such as production expense and short half-life in vivo.

Cell therapy

• ■ Transplantation of endothelial progenitors or mesenchymal stem cells has benefits for ischemic heart disease.

• ■ Administration of attractive cytokines is a possible alternative to or augmentation for cell transplantation but has not yet been very effective in clinical studies.

• ■ Engineered cell therapy is an attractive alternative to direct gene therapy.

• ■ Endothelial progenitors in combination with well-designed matrices are proving successful for tissue engineering.

Matrix interaction

• ■ A new generation of bio-active materials is being designed that closely mimics the natural extracellular matrix but provides a high level of control over the signaling environment.

• ■ Patterning techniques could be employed to control architecture of induced vasculature.

Metabolic controls

• ■ Administration of nitrite has been studied to selectively deliver proangiogenic nitric oxide to ischemic tissues while leaving healthy tissue unaffected.

• ■ Expression of peroxisome-proliferator-activated receptor-γ coactivator-1α, a metabolic sensor, activates a non-HIF-1α angiogenic pathway.

• ■ Reducing prolyl-hydroxylase-1 expression results in a hypoxia-tolerant state, reducing ischemic damage to tissue.

• ■ Pharmaceutical methods for simulating ischemic preconditioning might be possible.

Future perspective

• ■ Effective future approaches are likely to be combination therapies with components containing multiple GFs, upstream activators, cell-transplantations or novel matrices.