Targets and delivery methods for therapeutic angiogenesis in peripheral artery disease

Abstract

Therapeutic angiogenesis utilizing genetic and cellular modalities in the treatment of arterial obstructive diseases continues to evolve. This is, in part, because the mechanism of vasculogenesis, angiogenesis, and arteriogenesis (the three processes by which the body responds to obstruction of large conduit arteries) is a complex process that is still under investigation. To date, the majority of human trials utilizing molecular, genetic, and cellular modalities for therapeutic angiogenesis in the treatment of peripheral artery disease (PAD) have not shown efficacy. Consequently, the current available knowledge is yet to be translated into novel therapeutic approaches for the treatment of PAD. The aim of this review is to discuss relevant scientific and clinical advances in therapeutic angiogenesis and their potential application in the treatment of ischemic diseases of the peripheral arteries. Additionally, this review article discusses past and recent developments, such as some unconventional approaches that have the potential to be applied as therapeutic targets. The article also includes advances in the delivery of genetic, cellular, and bioactive endothelial growth factors.

Introduction

Peripheral artery disease (PAD) refers to arterial stenosis of the lower or upper extremity often secondary to atherosclerosis or thrombosis, but can also be due to other conditions such as embolitic disease, vasculitis, thomboangiitis obliterans, fibromuscular dysplasia, entrapment syndromes, and endofibrosis. PAD affects between 8 and 12 million adults in the USA alone,¹ and the prevalence is expected to rise with the aging population. Moreover, the burden of PAD and associated cardiovascular and cerebrovascular morbidity and mortality continues to increase.

Critical limb ischemia (CLI) represents a complication of PAD and is characterized by rest pain, non-healing ulcers, and gangrene of the diseased leg. CLI is a major cause of decreased mobility, quality of life (QoL), and functional capacity, as well as an increased risk of amputation or death. As first-line therapies, patients with PAD are treated with a combination of risk-factor modification, antiplatelet drugs and statins. Patients with intermittent claudication (IC) are also treated with exercise rehabilitation, and with drugs such as cilostazol to improve symptoms. Patients with CLI are treated surgically with either bypass or endovascular revascularization. However, some patients fail to respond to first-line therapies or are not candidates for endovascular or surgical procedures and have no option other than amputation. Therapeutic angiogenesis has attempted to offer an alternative treatment for this subgroup of patients.

The application of endothelial growth factors for therapeutic angiogenesis in PAD likely gained popularity with the revelation by Harold Dvorak et al. that a pro-angiogenic growth factor, later named vascular endothelial growth factor (VEGF), promotes pathologic angiogenesis.² The enthusiasm grew because there is a major unmet clinical need in a subgroup of PAD patients who cannot undergo the traditional surgical or endovascular revascularization. Hence, in the mid-1990s, Isner et al. published the first preclinical studies of therapeutic angiogenesis in the treatment of limb ischemia.^3,4 Since then, several preclinical, as well as large human trials have been conducted. Most of these studies used systemic and localized injection of proangiogenic genes or proteins in an attempt to promote physiological angiogenesis. However, these trials have shown modest and short-term benefits. Novel therapeutic modalities for vascular regeneration in cardiovascular medicine continue to be discovered. These include the use of bone marrow cells (BMCs), endothelial progenitor cells (EPCs), and the potential to apply human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs) to complement the traditional modalities of gene and recombinant endothelial growth factors. In addition, there has been advancement in the mode of gene and pro-angiogenic growth factor delivery, as well as in imaging modalities that have the potential to provide novel means to affect therapeutic angiogenesis in PAD. This review discusses the mechanisms of neovascularization in limb ischemia, the traditional pro-angiogenic methods employed so far, and the potential therapeutic targets and delivery techniques in the treatment of PAD.

Physiological neovascularization: vasculogenesis, angiogenesis and arteriogenesis

In normal physiological terms, neovascularization refers to the formation of new blood vessels and the remodeling of existing ones in order to restore the maximum vascular function and structure; and, in essence, reverse the aging process of the vasculature. Physiological neovascularization shares common pathways (similar endothelial growth factors and markers of angiogenesis) with pathological angiogenesis (tumor angiogenesis). Therefore, the aim of therapeutic angiogenesis is to recapitulate these pathways and deliver the relevant growth factors and associated genes or cells in a targeted and controlled manner so as to stimulate physiological neovascularization. Thus, physiological neovascularization is postulated to occur by the following mechanism: (1) vasculogenesis – the de novo generation of new vessels from vascular progenitor cells (hemangioblasts), with the transformation to endothelial cells (ECs), vascular smooth muscle cells, and pericytes during embryologic development;⁵ (2) angiogenesis – the sprouting of new capillaries from pre-existing vessels resulting in new capillary networks; and (3) arteriogenesis – the growth and formation of mature collateral arteries from pre-existing interconnecting arterioles after an arterial occlusion.

The processes of vasculogenesis and angiogenesis are distinct, based on their unique developmental mechanisms. Nevertheless, the two terms are often used as synonyms in the medical literature, and incorrectly so, to mean post-natal development of new blood vessels. During embryonic development, vasculogenesis begins with cell clusters referred to as blood islands with the participation of both the EPCs located at the periphery and the hematopoietic stem cells (HSCs) located in the center. Ultimately, the blood islands fuse and the EPCs give rise to ECs, while the HSCs give rise to mature blood cells. The fused blood islands then grow and differentiate to form capillary networks, which develop into full arteriovenous vascular structures.⁶ However, there is growing evidence that vasculogenesis does not exclusively occur during the pre-natal stage of development, but that the process is active during adulthood, the so-called ‘post-natal vasculogenesis’. This fact is supported by preclinical and clinical data, demonstrating that transplantation of peripherally derived EPCs from the bone marrow results in vasculogenesis.^7-9 Furthermore, the driving force in angiogenesis is hypoxia in the surrounding tissue. Oxygen tension plays a key role in the regulation of angiogenesis and in the expression of a number of genes, especially the hypoxia-inducible factor-1 (HIF-1), which is known to up-regulate multiple vascular endothelial growth factors, particularly the VEGF family. This canonical model induces angiogenesis, resulting in the proliferation and expansion of the capillary network that then increases the surface area available for oxygen diffusion from capillary to cell and decreases cellular ischemia.

Conversely, arteriogenesis occurs independently of hypoxia and is triggered by physical forces such as altered shear stress and pressure gradient between pre-existing arterial branches, which occur within the collateral arterioles after an occlusion of a major conduit artery.^10-12 Under normal circumstances, collateral arteries have small diameters that develop high resistance and thus exclude significant blood flow. The obstruction of a conduit artery results in an increase in pressure gradient across the collaterals, which results in an increase in the flow rate as well as arterial shear stress. The consequence is the alteration in the endothelium and the recruitment of monocyte chemoattractant protein-1 (MCP-1), and other chemokines.¹³ There is controversy about whether the processes mentioned above lead to a cascade of events resulting in the expression of endothelial growth factors such as nitric oxide (NO) and VEGF, which then activate endothelial proliferation and further angiogenesis.

Select target genes and pro-angiogenic growth factors

Several endothelial growth factors show the potential to stimulate angiogenesis in animal models, but the same success has yet to be shown in large human trials.

The endothelial growth factors that have gained particular interest are VEGF, fibroblast growth factor (FGF), hepatocyte growth factor (HGF), placental growth factor (PLGF), and developmental endothelial locus-1 (Del-1), among others. The following is a summary and description of select genes and their associated endothelial growth factors, transcription factors, chemokines, and extracellular matrix proteins, along with their functions and associated receptors (summarized in Table 1). Some relevant current and past clinical trials in therapeutic angiogenesis in PAD are also discussed below (summarized in Table 2).^14-24

Table 1.

Select factors with potential clinical utility in therapeutic angiogenesis in PAD

	Role	Isoforms	Receptor/promoter
Growth factors
VEGF	Stimulates cell proliferation, migration, and vascular formation; mobilizes and improves survival of EPCs	VEGF-A through -E, and PLGF-1 and -2	VEGFR-1 (Flt-1), VEGFR-2 (KDR/Flk-1), and VEGFR-3 (Flt-4)
FGF	Stimulates cell proliferation, migration, and differentiation; mobilizes stem cells and EPCs	23 classes (FGF-1 to FGF-23)	FGF-1, -2, -3, and -4
HGF	Stimulates cell mitogenesis, cell motility, matrix invasion, and survival; mobilizes EPCs and cardiac stem cells	HGF/NK1 and HGF/NK2	c-MET proto-oncogene
PDGF	Stimulates cell growth and division, vascular formation, pericyte and smooth muscle recruitment and proliferation	PDGF-A through -D; 5 homo/heterodimers (PDGF- AA, -AB, -BB, -CC, and -DD)	PDGFR-α and PDGFR-β
Angiopoietin	Ang-2 initiates and maintains angiogenesis; enhances vessel maturation and stability, while Ang-1 antagonizes Ang- 2 activity	Ang-1, -2, -3, and -4	TIE-1 and -2
Erythropoietin	Stimulates erythropoiesis, EC proliferation, and arteriogenesis; mobilizes HSCs and EPCs and improves survival	EPO	EpoR
GM-CSF	Stimulates stem cells to produce granulocytes and monocytes; promotes arteriogenesis	N/A	CD116
Transcription factors
HIF-1	Promotes angiogenesis by stimulating gene expression of VEGF, VEGFR-2, EPO, IGF-2, NO synthase etc., under hypoxic conditions	HIF-1α, HIF-1β (ARNT), HIF-2α, and HIF-3α	HRE
PGC-1α	Promotes gene expression for mitochondrial biogenesis and energy metabolism; simulates angiogenesis via the expression of VEGF	N/A	PPAR-γ
Chemokines
SDF-1 (CXCL12)	Attracts HSCs to the bone marrow and promotes vasculogenesis by recruiting EPCs from the bone marrow	N/A	CXCR4
Extracellular matrix proteins
Del-1	Inhibits accumulation of leukocytes in the endothelium and promotes EC attachment, migration, proliferation, and angiogenesis	N/A	Integrins (αvb3/αvb5)

Ang, angiopoietin; ARNT, aryl hydrocarbon receptor nuclear translocator; c-MET, transmembrane receptor tyrosine kinase; CXCL12, chemokine ligand 12; CXCR4, chemokine receptor 4; Del-1, developmental endothelial locus-1; EC, endothelial cell; EPCs, endothelial progenitor cells; EPO, erythropoietin; EpoR, erythropoietin receptor; FGF, fibroblast growth factor; Flk, fetal liver kinase; Flt, fms-related tyrosine kinase; GM-CSF, granulocyte/macrophage-colony stimulating factor; HGF, hepatocyte growth factor; HIF, hypoxia-inducible factor; HRE, hypoxia response element; HSCs, hematopoietic stem cells; IGF-1; insulin growth factor-1; KDR, kinase insert domain receptor; N/A, not applicable; NK1, N-terminal, first Kringle domains; NK2, N-terminal, first two Kringle domains; NO, nitric oxide; PDGF, platelet-derived growth factor; PDGFR, platelet-derived growth factor receptor; PGC-1α, peroxisome proliferator-activated receptor gamma coactivator 1-α; PLGF, placental growth factor; PPAR-γ, peroxisome proliferator-activated receptor gamma; SDF-1, stromal cell-derived factor-1; TIE, tyrosine kinase with immunoglobulin-like and EGF-like domains; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.

Table 2.

Relevant human trials investigating the effects of protein and gene therapy in PAD

Trial (phase)	Factor	Control	Delivery	No. of patients	Resultsa
N/A (I) Isner, 199814	VEGF-165 (plasmid)	N/A	IM	6 (7 limbs) with Buerger’s disease	Positive perfusion, edema in 3/7 limbs
VEGF-PVD (II) Makinen, 200215	VEGF-165 (adenovirus/plasmid)	Ringer’s lactate	IA	54 (18: adVEGF, 17: pVEGF/liposome, 19: placebo)	Positive vascularity; negative restenosis rate, Rutherford class and ABI
TRAFFIC (II) Lederman, 200216	FGF-2 (recombinant protein)	Placebo + 1 dose FGF-2	IA	174 (62: FGF-2 day 1 + placebo day 30, 54: FGF-2 days 1 and 30, 58: placebo days 1 and 30)	Positive PWT and ABI at day 90, but not at day 180, no difference with repeat doses at day 30
RAVE (II) Rajagopalan, 200317	VEGF-121 (adenovirus)	Placebo	IM	105 (40: high dose, 32: low dose, 33: placebo)	Negative PWT, ABI, COT, and QoL
GRONINGEN (II) Kusumanto, 200618	VEGF-165 (plasmid)	Saline	IM	54 diabetic patients (27: VEGF-165, 27: placebo)	Negative amputation rates; improvement in ulcer healing and ABI
DELTA-1 (IIa) Grossman, 200719	Del-1 (plasmid)	Poloxamer-188	IM	105 (52: Del-1 + poloxamer-188, 53: placebo)	Positive symptomatic improvement in both groups, but no difference in PWT, and ABI
HGF-STAT Powell, 200820	HGF (plasmid)	Saline	IM	104 (26: low dose, 25: middle dose, 27: high dose, 26: placebo)	Positive perfusion (TcPO2); negative amputation rates, wound healing, and ABI
TALISMAN201 (II) Nikol, 200821	FGF-1 (plasmid) NV1FGF	Saline	IM	125 (59: NV1FGF, 66: placebo)	Negative ulcer healing; reduced risk of major amputation and death
Shigematsu (III), 201022	HGF (plasmid)	Placebo	IM	40 (27: HGF, 13: placebo)	Improvement in ulcer healing and QoL; no improvement in ischemic rest pain, and ABI
TAMARIS (III) Belch, 201123	FGF-1 (plasmid) NV1FGF	Saline	IM	525 (259: NV1FGF, 266: placebo)	Negative reduction in the rates of amputation or death
WALK (II) Creager, 201124	HIF-1α/VP16 (adenovirus)	Placebo	IM	289 (74: low dose, 74: medium dose, 65: high dose, 76: placebo)	Negative PWT, COT, ABI, and PWD

^aIndicates the defined primary or secondary end points.

ABI, ankle–brachial index; (ad), adenovirus; AI, intra-arterial; COT, claudication onset time; Del-1, developmental endothelial locus-1; FGF, fibroblast growth factor; HGF, hepatocyte growth factor; HIF, hypoxia-inducible factor; IM, intramuscular; N/A, not applicable; NV1FGF, non-viral fibroblast growth factor-1 plasmid; (p), plasmid; PWD, perceived walking distance; PWT, peak walking time; QoL, quality of life; TcPO₂, transcutaneous oxygen pressure; VEGF, vascular endothelial growth factor; VP16, viral particle 16.

A. Vascular endothelial growth factor (VEGF) and target receptors

The VEGF family is the most widely studied and characterized endothelial growth factor. It is comprised of seven major isoforms: VEGF-A through E, and PLGF-1 and 2. The term VEGF is often used to refer to the VEGF-A isoform, known to play a crucial role in cell migration, proliferation, and survival. There are at least four other well-characterized isoforms of VEGF-A: VEGF-121, VEGF-165, VEGF-189, and VEGF-206, with VEGF-165 being the most biologically active.²⁵ VEGF interacts with several EC surface receptors, including VEGFR-1 (Flt-1) and VEGFR-2 (KDR/Flk-1). Although the role of VEGFR-1 is not yet well understood at the time of this writing, VEGFR-2 is known to interact with VEGF-A as well as with splice variants of VEGF-C and D in signaling angiogenesis.²⁵ The third receptor VEGFR-3 (Flt-4) interacts with VEGF-C and D and is involved with lymphangiogenesis.²⁵ Trials using the various isoforms of VEGF have been previously reported, and the relevant outcomes are summarized in Table 2. However, most of these clinical trials have been negative, except for the study done by Makinen et al., which only showed improvement in the primary end point.¹⁵

B. Fibroblast growth factor (FGF) and associated genes

The FGF family consists of at least 23 structurally related proteins, of which FGF-1 (also known as acidic, aFGF) and FGF-2 (also known as basic, bFGF) are the best characterized. FGF is a robust regulator of angiogenesis and a potent mediator of EC migration, differentiation, and overall survival.²⁶ FGF is also known to bind heparin receptors with high affinity, which then acts as a co-receptor for one of the four known families of FGF-receptors.²⁷ FGF was first shown to improve perfusion in an animal model of limb ischemia by Asahara et al.²⁸ As a result, there have been several phase I/II human trials evaluating the benefit of using FGF-based modalities in the form of genes or proteins to improve perfusion in PAD.^16,21,29 For example, in a phase II trial named TALISMAN, Nikol et al. evaluated the safety of intramuscular (IM) injection of NV1FGF (a non-viral naked FGF plasmid DNA) versus placebo in the treatment of patients with CLI.²¹ A total of 125 subjects were randomized to receive multiple injections of NV1FGF or saline; and at 52 weeks of follow-up there was no significant difference in ulcer healing (primary end point) between the two groups. However, there was a significant reduction in all amputation rates and overall mortality (secondary end points). It is likely that the severity of baseline ulcers, a small sample size, and the diverse nature of the study subjects might have contributed to the negative primary end point. However, recently, TAMARIS, a larger phase III trial that randomized 525 patients to receive IM injection of NV1FGF versus placebo, found no evidence that NV1FGF is effective in reducing the time to major amputation or death (primary end point) in patients with intractable CLI.²³ The secondary end points were also negative (all amputations, rest pain, ulcer healing, and functional status). These findings are disappointing, especially after the phase II trial (TALISMAN) gave some optimism. So why is the end point of major amputation and death positive in TALISMAN but negative in TAMARIS? Several reasons can be given: (1) the benefit seen in TALISMAN could have occurred by chance, given the small sample size (n = 125); (2) the heterogeneity of patients in the TAMARIS trial could have influenced the negative outcome; (3) the high percentage of co-morbid conditions such as diabetes and hyperlipidemia in the TAMARIS trial could have affected response to therapy; and (4) the TAMARIS trial was underpowered.

C. Hepatocyte growth factor (HGF) and associated genes

HGF is a cytokine originally described as a mesenchymal-derived factor. HGF is known to regulate cell growth, cell motility, morphogenesis, and angiogenesis through activation of its receptor, the transmembrane tyrosine kinase encoded by the c-Met proto-oncogene.^30,31 In an animal model, HGF promoted EC proliferation and angiogenesis in hind-limb ischemia.³² Further, several human trials have been performed using plasmid DNA encoding HGF in the treatment of limb ischemia. For instance, Powell et al. showed in the phase II HGF-STAT randomized trial that patients with rest pain or ischemic ulcers receiving the highest of the three escalating doses of IM injections of HGF plasmid DNA had improvement in limb perfusion (primary end point) as compared to placebo at 6 months.²⁰ However, there were no significant differences in amputation rates, wound healing, and ankle–brachial index (ABI) (secondary end points). Morishita et al. were able to show in two phase II trials that IM injections of HGF plasmid in patients with limb ischemia were safe, feasible, and promoted clinical improvement.^33,34 In a phase III trial, Shigematsu et al. demonstrated improvement in the primary end point (ulcer healing) with no improvement in ischemic rest pain and ABI (secondary end points) at 12 weeks in patients receiving IM injections of HGF plasmid DNA as compared to placebo for the treatment of CLI.²² Following these developments, a global multi-site phase III trial is ongoing with the aim to recruit over 500 PAD subjects who have no other therapeutic options.

D. Developmental endothelial locus-1 (Del-1)

Del-1 is an extracellular matrix protein expressed during early development and influences vasculogenesis.³⁵ Del-1 is known to inhibit inflammatory cell recruitment and could provide a basis for targeting leukocyte–endothelial interactions in vascular disease. Further, experimental data indicate that interaction of the αvb3/αvb5 integrin receptor complexes with Del-1 is critical in tumor-induced angiogenesis through a cell survival function.³⁶ In experimental animal models, Del-1 was shown to stimulate potent angiogenic response and promote functional recovery in hind-limb ischemia, although a cardiac ischemia model has shown mixed results.^37,38 The phase IIa DELTA-1 trial randomized 105 subjects with moderate to severe bilateral PAD to receive either IM injection of VLTS-589, a plasmid DNA encoding Del-1, mixed with poloxamer-188, or placebo (poloxamer-188 alone). The result indicated that Del-1 expressing plasmid DNA and the control resulted in significant improvement in exercise capacity and ABI compared to baseline at 30, 90, and 180 days; but there was no difference between groups in the primary or secondary end points as well as in outcome measures.¹⁹ A subsequent phase IIb trial was investigated to determine the possibility that poloxamer-188 had a placebo effect by comparing poloxamer-188 to saline, and that study too was negative.

E. Hypoxia-inducible factor-1 (HIF-1)

HIF-1 is a heterodimeric transcription factor that mediates adaptive responses to hypoxia and ischemia in eukaryotic cells. HIF-1 consists of oxygen (O₂)-regulated HIF-1α and constitutively expressed HIF-1β (ARNT) subunits.⁵ Under low oxygen tension, HIF-1α translocates to the nucleus, dimerizes with HIF-1β to transcribe genes encoding angiogenic cytokines such as VEGF, platelet-derived growth factor (PDGF), angiopoietin-2 (Ang-2), stromal cell-derived factor-1 (SDF-1), PLGF, and stem cell factor (SCF), all of which have been shown to promote neovascularization in EC culture.^39,40 These pro-angiogenic factors then bind to target receptors expressed on the surface of vascular ECs and vascular pericytes/smooth muscle cells, promoting angiogenesis.⁵ However, under normoxic conditions, HIF-1α is degraded by the prolyl-4-hydroxylase-2 (PHD2), which modifies HIF-1α by the addition of hydroxyl groups at proline residue 402 and/or 564, using oxygen as the rate-limiting substrate.⁴¹ Following the prolyl hydroxylation of HIF-1α, the von Hippel-Lindau tumor suppressor (pVHL) protein then binds to HIF-1α, triggering recruitment of E3 ubiquitin-protein ligase and resulting in HIF-1α ubiquitination and proteasomal degradation.⁴² Data from animal models indicate that HIF-1α mRNA and protein levels are elevated in the ischemic limb on day 3 post femoral artery ligation, coinciding with the expression of mRNA encoding some of the angiogenic growth factors mentioned.⁴³ These responses are impaired in Hif-1α (+/−) mice, which are heterozygous for a knockout allele at the locus encoding HIF-1α.⁴³ Further, a rabbit model of limb ischemia demonstrated that intramuscular injection of a constitutively active HIF-1α promoted arteriogenesis and angiogenesis.⁴⁴

Following the above experimental models, two human clinical trials have been investigated. A recent phase I dose escalation clinical trial using adenovirus to deliver a constitutively active HIF-1α plasmid (Ad2/HIF-1α/VP16) in patients with CLI was found to be safe and well tolerated.⁴⁵ The result of a similar phase II, randomized, controlled, placebo trial named ‘WALK’ was recently published.²⁴ In this study, a dose escalation of Ad2/HIF-1α/VP16 viral particles (2 × 10⁹, 2 × 10¹⁰, 2 × 10¹¹, or placebo) was delivered intra-muscularly (1:1:1:1 ratio) in a double-blinded fashion in 289 patients with severe bilateral claudication. Graded treadmill tests were performed at baseline and 3, 6, and 12 months post-treatment. The primary end point was the change in peak walking time (PWT) from baseline to 6 months. The secondary end point was the change in claudication onset time (COT), and the tertiary end points were the changes in ABI, perceived walking distance (PWD), and QoL measurers at baseline to 12 months. The results did not show any difference in the primary, secondary, or tertiary end points compared to placebo and the conclusion was that this modality is not effective in the treatment of patients with severe claudication. Several reasons are possible why this trial failed: (1) it is not clear whether the constitutively expressed HIF-1α has any clinical relevance in normoxic conditions with claudication alone, as was in this trial, as opposed to CLI, which has shown success in animal models of limb ischemia; (2) the mode of delivery (viral vector) may be inefficient in the skeletal muscle of patients and might have been degraded before it reached its target tissues; (3) the duration required for adequate gene expression coupled with subjects’ genetic characteristics may affect clinical outcomes; and (4) the ideal dose of viral particles may be inadequate to ascertain a therapeutic relevance. Further trials are needed, especially in patients with CLI, perhaps comparing two well-studied delivery methods so as to alleviate some of the above confounding variables.

Potential target genes and small molecules

There are several unconventional target genes and molecules that show promising results in experimental studies. Interestingly, some of these potential targets are regulated by oxygen tension and hence activate the HIF-1α pathway to initiate angiogenesis.

A. Prolyl-4-hydroxylase inhibitors

The prolyl 4-hydroxylases (PHD1, PHD2, and PHD3) are a family of enzymes that target HIF-1 for degradation and are the key physiological regulators of cellular hypoxia. As mentioned, the PHD2 (EGLN1) plays a major role in the proteasomal degradation of HIF-1α under normoxic conditions through hydroxylation of proline residue 402 and/or 564 utilizing α-oxoglutarate and O₂ as substrates. The hydroxylation reaction then initiates binding of HIF-1α by the pVHL protein, which then triggers the recruitment of E3 ubiquitin-protein ligase, resulting in HIF-1α ubiquitination and proteasomal degradation.

A number of small molecules inhibit PHD enzymes.^46,47 For example, PR39 is a proline and arginine rich peptide that stimulates angiogenesis indirectly by inhibiting PHD2, hence preventing the proteolytic degradation of HIF-1α.⁴⁸ This pathway provides a potential alternative in further prolonging the bioavailability of HIF-1α in ischemic conditions. Co-administration of HIF-1α, in addition to PR39 via viral vector DNA delivery, has shown some cardioprotection in animal models of cardiovascular disease.^49,50 Another approach to inhibiting PHDs has also been explored by the construction of small interfering RNA (siRNA) that attenuates the expression of PHD2, thus increasing HIF-1α stability and activity in a murine model of ischemia.⁵¹ It is not yet clear if the application of PHD inhibitors will be clinically efficacious in the treatment of PAD. Concerns abound as to whether there will be increased risks of promoting tumor growth. However, further insight may be obtained from the various ongoing clinical trials that utilize the inhibitors to treat conditions such as anemia.

B. The thioredoxin system

The thioredoxin (Trx) system is composed of redox proteins that function to reduce oxidized cysteins in the endothelium through a nicotinamide adenine dinucleotide phosphate (NADPH)-dependent reaction. The classic oxidation–reduction reaction catalyzed by the Trx system involves the recycling of unpaired electrons between the Trx reductase (TrxR) and NADPH oxidase, leading to the transformation of Trx from its inactive oxidized form to its active reduced form. The active form of Trx then functions as a reducing agent for various endothelial peroxidases, other cellular antioxidants, as well as transcription factors, thus maintaining oxidative homeostasis. Oxidative stress in the endothelium is mainly mediated via the reactive oxygen species (ROS). Although physiological levels of ROS are required for angiogenesis, excessive ROS production is known to induce ischemia and apoptosis, especially in patients with cardiovascular risk factors.⁵² Therefore, the Trx system acts to balance the ROS system to the desired physiological level, which results in enhanced cell survival, function and overall positive effect in post-natal angiogenesis. For example, in an in vitro study, overexpression of Trx was associated with increased expression of VEGF and HIF-1α and, therefore, angiogenesis. Conversely, transfection of cells with antisense to Trx markedly reduced the expression of VEGF and HIF-1α as well as angiogenesis.⁵³ The Trx system has two isoforms: the cytosolic and nuclear form (Trx-1) and the mitochondrial form (Trx-2); Trx-1 being the best well characterized.⁵⁴ The vascular endothelium also has numerous ROS producing systems, including NADPH oxidases and xanthine oxidase, among others. The NADPH oxidases are known to generate the most abundant ROS in the pathogenesis of cardiovascular disease.⁵⁵

As mentioned, the cellular and molecular mechanisms by which the Trx system, and especially Trx-1, stimulates angiogenesis involve the reduction of ROS, thereby stimulating cellular proliferation and migration by activating a variety of transcription factors, including HIF-1α. Trx-1 is known to inhibit the pVHL-mediated degradation of HIF-1α, which leads to VEGF expression and further activation of pro-angiogenic factors such as NO.⁵³ Moreover, in pathologic angiogenesis, overexpression of Trx-1 is associated with increased levels of HIF-1α and VEGF expression, and cancer cells overexpressing Trx-1 are associated with decreased survival as well as poor response to chemotherapy.^54,56 The small molecules, PX-12 and pleurotin, inhibit Trx-1 and prevent the expression of HIF-1α under hypoxic conditions, correlating with decreased expression of VEGF.⁵⁷ Several animal models have been explored to investigate the potential utility of Trx-1 in promoting angiogenesis in the cardiovascular system. For instance, adenovirus-mediated transfection of Trx-1 in the myocardium of diabetic rats resulted in increased angiogenesis and reduced left ventricular remodeling, further re-enforcing the cardioprotective role of Trx-1.⁵⁸ The application of the above knowledge regarding the Trx system provides yet another potential alternative that could be applied in the treatment of ischemic diseases. However, human clinical trials dedicated to vascular disease will need to be explored.

C. The PGC-1α pathway

The peroxisome proliferator-activated receptor gamma coactivator 1-α (PGC-1α) protein, is a transcriptional coactivator abundant in the mitochondria and is known to regulate metabolic function during exercise, low oxygen or nutrient state, as well as to stimulate VEGF gene expression and angiogenesis.^59,60 Furthermore, PGC-1α strongly induces other angiogenic factors such as platelet-derived growth factor subunit B (PDGFB) and Ang-2 in cultured muscle cells,⁶¹ and mice lacking PGC-1α in the skeletal muscle fail to increase capillary density in response to exercise.⁶⁰ Also, homozygous knockout mice (PGC-1α −/−) demonstrate the inability to reconstitute normal blood flow after limb ischemia, which is reversed after transgenic expression of the PGC-1α gene.⁵⁹ The exact mechanism by which PGC-1α induces VEGF expression and angiogenesis is uncertain but has been investigated. For example, in a mouse model of hind-limb ischemia, Arany et al. reported that the mechanism of induction of VEGF expression by PGC-1α is not dependent on the canonical hypoxia response pathway regulated by HIF. Instead, their research shows that PGC-1α coactivates a separate nuclear receptor, the estrogen-related receptor-α (ERRα) found on the promoter and on a newly identified enhancer in the first intron of the VEGF gene.⁵⁹ Thus, they conclude that PGC-1α and ERRα control a novel, HIF-independent angiogenic pathway that delivers needed oxygen and substrates in the face of ischemia.⁵⁹ These findings strongly suggest that PGC-1α has a major role in the regulation of angiogenesis in the muscle in response to ischemic insult. In addition to pathologic ischemia, PGC-1α also plays a role in angiogenesis under physiologic conditions. The expression of PGC-1α is strongly induced by exercise in rodents and humans, and PGC-1α is a powerful promoter of both mitochondrial biogenesis and angiogenesis in skeletal muscle.⁶² Therefore, it is likely that PGC-1α promotes physiological and ischemia-induced angiogenesis. A recent in vitro study seems to implicate HIF-2α as a major PGC-1α target in skeletal muscle and this pathway is stimulated by exercise and increased β-adrenergic tone. There is further suggestion that PGC-1α might play a role in the improvement of mitochondrial function, which is often reduced in PAD patients.⁶³ However, with the limited clinical data, it is premature to predict the exact role of the PGC-1α system in therapeutic angiogenesis.

D. MicroRNAs and oligonucleotides

MicroRNAs (miRNAs) represent a class of ~22 nucleotide non-coding RNAs that regulate gene expression by targeting messenger RNAs (mRNAs) for cleavage or translational repression.⁶⁴ Recent data suggest that miRNAs contribute to ischemia–reperfusion injury by altering key signaling elements related to the vasculature, thus making them potential targets for inhibition or enhancement for therapeutic angiogenesis.⁶⁵ For example, miR-92a, a component of a cluster of miRNAs, is implicated in the process of inhibiting angiogenesis.^66,67

In an animal model, forced overexpression of miR-92a in ECs blocked angiogenesis in vitro and in vivo. It is postulated that miR-92a acts through effects on expression of integrins, proteins involved in cell adhesion and migration.^66,67 Inhibition of integrin expression results in the accumulation of leukocytes in the endothelium, which leads to increased inflammation, endothelial dysfunction, and apoptosis. In mouse models of heart and limb ischemia, therapeutic inhibition of miR-92a led to an increase in blood vessel density in the damaged tissues and enhanced functional recovery.⁶⁷ In another study, Ghosh et al. found that the expression of HIF-1α is upregulated under hypoxia in human ECs transfected with miR-424, leading to increased expression of VEGF, glucose transporter-1 (GLUT-1), erythropoietin (EPO), and angiogenesis.⁶⁸ The same study was able to correlate similar findings in a mouse model of limb ischemia. Interestingly, miR-92a levels were significantly decreased under this model, further emphasizing the role of miR-424 in promoting angiogenesis. The mechanism by which miR-424 induces expression HIF-1α is via the inhibition of cullin-2 (CUL2) mRNA translation. The CUL2 molecule is known to play a crucial role in the assembly of the E3 ubiquitin-protein ligase system, resulting in the proteolytic degradation of HIF-1α.⁶⁹ A comprehensive discussion of the various miRNAs and oligonucleotides is beyond the scope of this review. However, it remains crucial to perfect this modality in human trials in order to further elucidate the clinical importance of miRNAs as potential targets for therapeutic angiogenesis in PAD.

Cells as targets for therapeutic angiogenesis in PAD

Cell-based therapy is an attractive approach that has the potential to be applied in therapeutic angiogenesis. Animal data of hind-limb ischemia models and human trials of PAD done thus far have employed the transplantation of either autologous or allogeneic cells. Some of the cell sources under investigation include EPCs, BMCs, mesenchymal stem cells (MSCs), adipose-derived stem cells (ADSCs), hESCs, and iPSCs. These cell sources are discussed below, including some relevant current and previous clinical trials (summarized in Table 3).^70-76

Table 3.

Relevant human trials of cell-based therapy in therapeutic angiogenesis in PAD

Trial (phase)	Treatment	Control	Delivery	No. of patients	Resultsa
START (II) Royen, 200570	GM-CSF	Placebo	SQ	40 (20: GM-CSF, 20: placebo)	Negative ABI and PWT; positive microvascular flow
EPOCH-CLI (I/IIa) Kawamoto, 200971	EPCs (CD34)	N/A	IM	17	Improvement in exercise capacity (TWD, PFWD), ulcer healing, and hemodynamics (ABI, TBI and TcPO2)
Burt (I), 201072	EPCs (CD34/CD133)	N/A	IM	9	Improvement in amputation-free survival, exercise capacity, pain relief, collateral formation, perfusion, and QoL
Lasala (I), 201073	MNCs + MSCs	N/A	IA	10	Improvement in exercise capacity, hemodynamic measurements, collateral formation, and QoL
PROVASA (II) Walter, 201174	BM-MNCs	Placebo	IA	40 (19: BM-MNCs, 21: placebo); 3 months’ follow-up, randomized-start crossover	Positive ulcer healing, negative limb perfusion, and ABI
RESTORE-CLI (II) Powell, 201175	BM-MNCs	Placebo	IM	46 (32: BM-MNCs, 14: placebo)	Improvement in amputation-free survival and ulcer healing
Idei (I) 201176	BM-MNCs	Placebo	IM	97 (51: BM-MNCs, 46: placebo)	Improvement in amputation-free survival, cumulative survival, and hemodynamic measurements
Harvest Technologies (III) Clinicaltrials.gov # NCT01245335	BM-MNCs	Placebo	IM	210	Ongoing
MESENDO (II) Clinicaltrials.gov # NCT00721006	‘Stem cell mixture’	Placebo	IM	35	Completed; pending publication
ACT34-CLI (II) Clinicaltrials.gov # NCT00616980	CD34-positive cells	Placebo	IM	75	Completed; pending publication

^aIndicates the defined primary or secondary end points.

ABI, ankle–brachial index; BM-MNCs, bone marrow-mononuclear cells; EPCs, endothelial progenitor cells; GM-CSF, granulocyte/macrophage-colony stimulating factor; IA, intra-arterial; IM, intramuscular; MNC, mononuclear cells; MSC, mesenchymal stem cells; N/A, not applicable; PFWT, pain-free walking time; PWT, peak walking time; QoL, quality of life; SQ, subcutaneous; TBI, toe–brachial index; TcPO2, transcutaneous oxygen pressure; TWD, total walking distance.

A. Endothelial progenitor cells (EPCs)

Asahara et al. were the first to demonstrate that vasculogenesis is feasible through peripheral blood EPCs.⁹ Further investigation established that EPCs are immature ECs that express CD34 receptors as well as VEGFR-2 (KDR/Flk-1); EPCs are known to mobilize the bone marrow and initiate post-natal vasculogenesis. At least four origins of EPCs have been identified: hematopeiotic stem cells, myeloid cells (which differentiate into ECs), circulating progenitor cells (side population cells), and circulating mature ECs.⁷⁷

Generally, EPCs are referred to as peripheral blood or bone marrow-derived mononuclear cells that adhere to the matrix molecules, such as fibronectin, and express dual acetylated low-density lipoprotein (LDL) and ulex europaeus agglutinin-1 lectin (UEA-1 lectin).⁹ However, there are still controversies as to the exact definition of EPCs, given that hematopeiotic stem cells express both CD34 and CD133 receptors and other mixed populations of cells. This confusion stems from the fact that there has been no definitive assay to determine the exact molecules characteristic of mature versus immature EPCs. Nonetheless, there are recent suggestions that true EPCs do not express CD133 receptors, hence providing a potential way to differentiate them from mature ECs.^78,79 There are two other distinguishing factors between EPCs and mature ECs: EPCs have the ability to proliferate in vivo and form new vessels, while ECs are in the circulation, are committed, are non-dividing and are depleted with time – and their potential for angiogenesis is still under investigation. In hind-limb ischemia models, transplantation of EPCs shows some benefit in neovascularization.^80,81 Because circulating EPCs are rare in number, some studies have attempted to stimulate their release from the bone marrow using granulocyte/macrophage-colony stimulating factor (GM-CSF). The study START (STimulation of ARTeriogenesis using subcutaneous application of GM-CSF as a new treatment for peripheral vascular disease), for example, investigated the utility of GM-CSF administered subcutaneously in patients with PAD. The study was not able to show benefits in the group treated with GM-CSF versus placebo. In fact, there was a placebo effect in the control group: after days 14 and 90, there was no difference in walking distance between the two groups. As well, there was no difference in walking time and ABI between the groups. Microcirculation as measured by laser Doppler perfusion, however, was decreased in the control group, with no difference observed within the groups treated with GM-CSF.⁷⁰

Recent investigations have focused on the use of EPCs believed to be CD34+ or CD133+ enriched. For example, in a phase I/II trial, Kawamoto et al. studied the safety of autologous IM injection of GM-CSF-mobilized cells in patients with CLI due to atherosclerosis or thromboangiitis obliterans (TAO).⁷¹ The CD34+ cells were first mobilized from the bone marrow then isolated from the peripheral circulation by apheresis, expanded, then transplanted in a dose-escalating manner into 17 patients with no placebo arm (10⁵/kg, n = 6, 5 × 10⁵/kg, n = 8, and 10⁶/kg, n = 3). At 12 weeks’ follow-up, there were benefits in both the primary end points (toe–brachial index (TBI), pain scale, and total walking distance (TWD)) and secondary end points (adverse events, pain scale, ulcer healing, and perfusion measures). However, this is not a randomized controlled trial. A larger phase III trial is needed to offset the possibility of a placebo effect.

Similarly, in a phase I trial, Burt et al. evaluated the feasibility of autologous IM transplantation of CD133+ enriched EPCs in nine patients with CLI.⁷² The primary end point was amputation-free survival at 12 months; the secondary end points were relief of pain, new collateral vessel formation, and QoL. At 12 months’ follow-up, stem cell injection prevented leg amputation in seven out of nine subjects. Additionally, the remaining seven patients showed improvement in QoL and pain-free walking distance (PFWD). However, there was improvement at 3 and 6 months, but not at 12 months, on the Summary Performance Score. Again, large randomized controlled trials are needed, especially those that compare the various EPC-enriched populations so as to better understand the efficacy of these modalities in the treatment of PAD.

B. Bone marrow cells (BMCs)

As the name suggests, BMCs are derived from the bone marrow and can be either crude, unfractionated or mononuclear cells separated by density centrifugation.⁸² The BMCs are reservoirs of cells that act as precursors at several levels of maturity and multipotency. Most importantly, BMCs contain stem cells as well as progenitor cells in addition to HSCs, MSCs, hemangioblasts, and EPCs.⁸³ Animal studies indicate that transplantation of BMCs improves neovascularization in hind-limb ischemia. In 2002, the first set of human clinical trials using BMC transplantation in the treatment of PAD was reported.^8,84

In a recent phase II, randomized, double-blinded, placebo-controlled trial named PROVASA (Intraarterial Progenitor Cell Transplantation of Bone Marrow Mononuclear Cells for Induction of Neovascularization in Patients With Peripheral Artery Occlusive Disease), Walter et al. hypothesized that intra-aterial infusion of bone marrow mononuclear cells (BM-MNCs) is associated with improved limb perfusion as measured by ABI (primary end point) and with reduction of ischemic rest pain and improved ulcer healing (secondary end point).⁷⁴ This study utilized a unique method called randomized-start clinical trial design: 40 patients were randomized in a 1:1 manner to receive either intra-arterial infusion of autologous BM-MNCs or placebo, and at the end of 3 months, the placebo group were crossed over to active treatment and received the first dose of BM-MNCs, and the active group received a second dose of BM-MNCs. At 6 months’ follow-up, a comparison was made between the two groups, and patients who still had significant ulcers (Rutherford class 5) received extra therapy (third dose for the original active group and second for the original placebo group). This design assessed whether repeated treatments with autologous BMCs may be of benefit as compared to a single therapy, as had been done in previous studies. This trial was found to be safe and feasible and resulted in significant improvement in the secondary end point but no difference in the primary end point.

An interim analysis for the study ‘Use of Tissue Repair Cells (TRCs Autologous Bone Marrow Cells) in patients with PAD to Treat Critical Limb Ischemia’ (RESTORE-CLI) was recently published.⁷⁵ This is a phase II, prospective, randomized, double-blinded trial comparing the safety of IM injection of expanded autologous BMCs (tissue repair cells or TRCs) in patients with CLI as compared to placebo. A total of 46 patients were unblinded and considered for the interim analysis. Of the 46 subjects, 33 completed the full 12 months’ follow-up and 13 completed at least 6 months of follow-up. They were then randomized to receive TRCs or placebo (TRCs = 32, control = 14). At follow-up (days 0, 3, and 7; and months 3, 6, 9, and 12), there was no significant difference in the primary end point (safety) between the two groups. However, there was benefit in the secondary end points (major amputation-free survival and time to first occurrence of treatment failure – defined as major amputation, death, de novo gangrene, or doubling of wound size) in the group treated with TRCs. This study is the first reported placebo-controlled autologous stem cell trial utilizing enriched and expanded BM-MNCs for the treatment of CLI. This study also applies hard clinical end points such as amputation-free survival rather than surrogate perfusion end points (ABI, TBI, transcutaneous oxygen pressure (TcPO₂), etc.). A larger phase III trial is still needed to validate these findings.

In a phase I trial, Idei et al. evaluated the long-term clinical outcomes after autologous BM-MNC transplantation in subjects with CLI.⁷⁶ Specifically, they sought to compare outcomes in patients with atherosclerosis as the cause of CLI (n = 25) versus TAO (n = 26). Forty-six patients were used as control, with 30 having atherosclerosis and 16 having TAO. At the 4-year follow-up, there was benefit in the primary end points (major amputation-free survival and cumulative mortality rates) in the treatment groups. The major amputation-free survival rate was 48% in the treatment atherosclerosis group and 95% in the treatment TAO group as compared to 0% in the control atherosclerosis group and 6% in the control TAO group. The cumulative survival rate was 76% in the treatment atherosclerosis group and 100% in the treatment TAO group, and 67% in the control atherosclerosis group and 100% in the control TAO group. Additionally, there was a sustained improvement in the secondary end points (ABI and TcPO₂) 1 month to 3 years post-treatment in the TAO group. The opposite was observed in the atherosclerosis group, with improvement in ABI and TcPO₂ lasting for only 1 month. This study offers interesting insights in comparing response to BM-MNC therapy among different disease states of CLI. However, it is a non-randomized study and is subject to selection bias; subjects with TAO in the treatment group were instructed to discontinue smoking before treatment could be administered, while the control group was not prohibited from smoking. It also seems that subjects in the treatment group had less severe disease as compared to the control group (the majority being considered for major amputation). Moreover, patients with TAO are often young (< 50 years old) with less atherosclerosis risk factors besides smoking, may have better progenitor and bone marrow function, and generally have a better prognosis with smoking cessation. These issues render the interpretation of these findings challenging.

Another study worth mentioning is the ongoing MESENDO trial. This is a phase II safety and efficacy study that is evaluating a ‘stem cell mixture’ from BMCs in the treatment of CLI (clinicaltrials.gov # NCT00721006). In this study, subjects with bilateral lower extremity claudication are non-randomly assigned to receive subcutaneous injections of either a combination of autologous BMCs in the treatment limb or receive placebo in the opposite control limb. The primary outcome measures the improvement in new vessel formation after 4 months as imaged by a nuclear perfusion scan, and the secondary outcome measures patient safety and the improvement in resting leg pain, as identified by a visual analog scale.

Although the application of BMCs in the treatment of PAD is promising, the mechanism of action of BMCs in promoting vascular regeneration still remains controversial. The engraftment within the resident endothelium versus paracrine hypothesis continues to be debated. Some investigators studying myocardial regeneration support the engraftment hypothesis.^85,86 However, this mechanism has been conflicted in some instances.⁸⁷ Hence, paracrine activity has been proposed to be necessary for the completion of myocardial repair.⁸⁸ Further investigations are needed to remove these controversies.

The majority of stem cell therapy trials for PAD show a high level of placebo effect. Perhaps there could be reasons to explain this effect, including the possibility that: (1) we do not fully understand the process of therapeutic angiogenesis as a whole; (2) there is a need to improve the techniques of characterizing and identifying specific cells and cytokines relevant for neovascularization; (3) there is a need to standardize the optimum dose and route of administration; and (4) there is a need to perfect methods of efficacy evaluation in clinical trials. Moreover, the majority of these trials have been phase I or II trials, and more randomized-controlled phase III trials are needed to validate the degree of efficacy of the various cellular modalities. Additionally, phase III trials will need to include combination therapies along with the use of less invasive delivery methods such as IM injection as opposed to the more invasive intravascular routes. Methods of reporting end points for efficacy need to be standardized. For example, the improvement in hemodynamic and performance parameters as primary end points in efficacy measurement remains controversial. This is because hemodynamic parameters such as ABI, TcPO₂, nuclear perfusion, and angiography can be unreliable, are operator dependent, and can influence placebo response. For instance, ABI can be unreliable in diabetes, renal failure, and old age because the arteries maybe calcified and become non-compressible. Moreover, the test can be operator dependent given cuff size and body habitus; skin thickness, ambient temperature and systemic blood pressure may also affect TcPO₂ measurements; interpreting angiographic findings sometimes requires more than one reader and is operator dependent, relying on the timing of contrast infusion; and nuclear perfusion measurements can be unreliable depending on the amount of isotopes delivered at different times. Performance parameters such as walking time can also be unreliable; walking time is limited by its subjectivity and the patient’s own report of claudication onset. In addition, evaluating bilateral claudication can be challenging because it is sometimes difficult to tell which leg is producing symptoms. Nonetheless, recent trials are applying primary end points of major amputation, amputation-free survival, and death, which we believe are more robust and less likely to result in placebo effect.

C. Mesenchymal stem cells (MSCs)

MSCs originate from the bone marrow and are not hematopoietic in nature. MSCs can differentiate into other cells, namely osteoblasts, myoblasts, chondrocytes, and adipocytes.⁸⁹ There are no specific markers for MSCs. However, in general, it is recognized that MSCs express mainly CD90, CD105, CD44, and Stro-1 cells and cellular adhesion molecules VCAM-1 and ICAM-1, but not hematopoietic markers such as CD34, CD14, CD45 or CD11.⁹⁰ MSCs have generated great interest as a source of cell therapy for the treatment of several ischemic diseases.^91-93 MSCs have several advantages: (1) they are easy to isolate from the bone marrow and expand ex vivo to adequate numbers; (2) MSCs demonstrate low immunogenicity and have the potential to be modified for allogeneic transplantation without the need for immunosuppression; (3) they can be modified by a viral vector to carry specific genes that participate in neovascularization; and (4) they can be delivered systemically or locally without significant adverse effects.

Similarly, the mechanism of action of MSCs toward neovascularization still remains controversial. It is largely thought to occur through the release of cytokines that may have paracrine or distance tissue effects, resulting in the stimulation of endogenous ECs to migrate, differentiate, and proliferate in situ, thus replenishing depleted resident ECs. There are numerous animal studies showing benefits with transplantation of MSCs in ischemic conditions. The first trial with MSCs was done in 1995, when autologous MSCs were infused in humans; and 2002 saw the first infusion of allogeneic MSCs in humans.^94,95 Limited data exist on the use of MSCs as a sole cellular agent in the treatment of PAD. Current studies have used combination cell therapy: EPCs plus MSCs or MNCs plus MSCs approaches. For instance, in a phase I trial, Lasala et al. studied the utility of an intra-arterial infusion of an autologous mixture of MSCs plus MNCs in the treatment of CLI in 10 diabetic patients (Fontaine stages 2B to 4 of CLI).⁷³ At 10 months’ follow-up, there was improvement in the primary end point (safety and feasibility) and secondary end points (rest pain, exercise tolerance, ABI, TcPO₂, and collateralization). Obviously, this is a non-randomized trial and no significant conclusions can be made from the findings, and further large trials are needed. However, some technical aspects of cell delivery and preparation will need to be refined to affect the efficacy of transplanted cells. This might be due to a poor rate of homing, engraftment, and survival of transplanted cells to the site of ischemia. With continued advancement in the techniques of cell selection and modification, the future of using MSCs as a therapeutic modality for PAD looks even brighter.

D. Adipose-derived stem cells (ADSCs)

Autologous adipose tissue also represents an emerging source of stem cells for therapeutic utility in ischemic disease. Adipose tissue is composed mainly of two categories of cells: mature adipocytes (MA), which form the majority of the adipose tissue volume, and stromal cells, often termed stromal vascular fraction (SVF).⁹⁶ It has been demonstrated that SVF contains multipotent mesenchymal stem/progenitor cells that are capable of differentiating to various lineages such as myocytes, adipocytes, fibroblasts, osteoblasts, and chondroblasts.^97,98 The MSCs contained in adipose tissue are often called adipose-derived stem/progenitor cells (ADSCs or ASCs) or adipose-derived regenerative cells (ADRCs), and have the capability to regenerate injured tissue.⁹⁶

The transplantation of ADSCs increases angiogenesis in animal models of hind-limb ischemia.^99,100 There are no current human clinical trials for neovascularization targeting PAD, but there have been trials involving adipocytes and cardiac progenitor cells in the treatment of cardiac diseases.^101,102 Like most cell-based therapies, the mechanism of action of ADSCs in neovascularization is not well understood. However, recent data indicate that the ADSCs do not have the capability to differentiate into ECs or EPCs as was originally thought, but that ADSCs can differentiate into smooth muscle cells and pericytes.¹⁰³ Not surprising, transplantation of ADSCs influences therapeutic angiogenesis through the secretion of pro-angiogenic chemokines and cytokines such as SDF-1, VEGF and HGF.^99,104 In particular, the chemokine SDF-1 is now thought to play a crucial role in ADSC-mediated angiogenesis.⁹⁹

One of the advantages of ADSCs is that they can be isolated through minimally invasive procedures such as liposuction or excision of the subcutaneous fat pad, which can be then expanded ex vivo. Even though ADSCs presents a cheap and easily accessible source for stem cells for therapeutic angiogenesis, before they become routinely used further studies will need to be done. It is important to determine the amount of ADSCs to be transplanted, as well as the potential attenuated response of transplanted autologous ADSCs in patients with co-morbid conditions such as diabetes, hypertension, advanced age, and tobacco abuse, which are all known to affect the number and function of circulating progenitor cells.^105,106

E. Human embryonic stem cells (hESCs)

Human ESCs are pluripotent cells originating from the inner cell mass of blastocyts of an embryo and are capable of self-renewal and differentiation into any cell lineage composed of the three germ layers: ectoderm, mesoderm, and endoderm. These unique characteristics make hESCs an excellent source of cells for therapeutic angiogenesis. It has been demonstrated that hESCs can be directed to differentiate into cardiomyocytes, endothelial cells, and vascular smooth muscle cells ex vivo.^107-109 Indeed, hESCs can also be directed to express the endothelial marker CD31, express adhesion molecules such as platelet and endothelial cell adhesion molecule (PECAM) and von Willbrand factor (vWF), as well as exhibit endothelial functions such as uptake of acetylated LDL and formation of capillary tubules in matrigel.¹¹⁰

In an animal study, there was successful limb salvage in mice transplanted with hESCs post femoral artery ligation.¹¹¹ The use of hESCs in therapeutic angiogenesis will have to overcome some controversies before they become routinely used in humans. First, is the ethical issue surrounding the use of human embryos, which led the FDA to ban the widespread use of hESCs. Additionally, there is limited knowledge as to the long-term effects of transplantation of hESCs in humans. There is concern among experts that administration of hESCs may have adverse effects such as differentiation into teratomas.¹¹² The technical challenges of allogeneicity related to the use of hESCs will also have to be overcome.

F. Induced pluripotent stem cells (iPSCs)

As mentioned above, embryonic stem cells are the only cells that traditionally are capable of omnipotent plasticity and differentiation into other distinct cell lines. However, this has been redefined by the development of iPSCs; the notion that fully differentiated and committed adult cells can be reprogrammed in vitro through the introduction of transcription factors that induce somatic cells to express genetic profiles characteristic of an embryonic stem cell has revolutionized the prospect of stem cells as a therapeutic modality in the treatment of ischemic diseases.

Takahashi et al. reported this groundbreaking experiment in 2006 when they showed that the transduction of four key transcription factors (Oct3/4, Sox2, c-Myc, and Klf4) into the nucleus of adult mouse fibroblasts, generated stem cells capable of differentiating into cells with the three germ layers.¹¹³ Since then, iPSCs have also been derived from human fibroblasts utilizing various transcription factors¹¹⁴ with disease-specific profiles.¹¹⁵

The development provided by iPSCs presents promising alternatives, especially for revascularization therapy in PAD. To date, iPSCs have been reprogrammed in vitro into cardiomyocytes, vascular mural cells, and vascular endothelial cells.^116,117 Using a mouse model, Suzuki et al. showed that iPSC-derived Flk-1 cells can stimulate angiogenesis in ischemic limb tissue by temporarily elevating the concentration of available VEGF and by directly incorporating endothelial cells into ischemic tissue.¹¹⁸ While the first human trial for embryonic stem cell therapy was approved in 2010 for use in patients with spinal cord injuries, owing to safety concerns there are no current human trials of ischemic diseases done using iPSCs as delivery tools for therapeutic angiogenesis.

There are several potential benefits associated with using iPSCs in vascular regeneration. The use of iPSCs, while having the same characteristics as hESCs, is non-controversial because the generation of iPSCs does not involve the harvesting of cellular components from the developing embryo. Additionally, the supply of iPSCs is ample in every individual, and consequently could easily be harvested non-invasively. This makes it likely that the iPSCs would be autologous to the recipient, eliminating the need for immunosuppression during treatment. iPSC technology could also be tailored to be patient-specific, further energizing the field of personalized medicine.

However, it is still unclear whether iPSCs would provide the magic bullet for individuals seeking stem cell therapy for ischemic diseases. First, it is possible that portions of the iPSC DNA may accumulate genetic errors in the reprogramming process. For instance, if retroviral integration occurs causing mutations in loci responsible for regulating cell-cycle or tumor suppression, oncogenes could be provoked, leading to carcinogenesis. Another potential problem could arise if cells fail to differentiate appropriately following transplantation – teratomas or teratocarcinomas may develop from the uncontrolled activation of these cells at unpredicted rates. It is also possible that unhealthy donors may have dysfunctional iPSCs, which could cause further damage to recipients requiring stem cell stimulation. Consequently, safely administering iPSCs to humans would first require the development of a thorough screening process to determine if cells are safe for implantation. Overall, more research is required on how to directly induce iPSCs specifically into endothelial cells and how to ensure that these cells have been fully reprogrammed and are safe for implantation before iPSCs are considered a viable form of therapeutic angiogenesis. Figure 1 is a depiction of the various sources of iPSCs.

Figure 1.

The three cell lines currently under investigation for vascular regeneration. Endothelial stem cells are extracted from vascular endothelium. Adult stem cells are harvested from progenitor cells in the myocardium or the bone marrow stroma. Induced pluripotent stem cells are somatic cells extracted from either adipocytes or dermal fibroblasts and are reprogrammed via retroviral transfection.

Select potential targeted delivery methods

Delivery methods for genes encoding growth factors, bioactive endothelial proteins and cells for therapeutic angiogenesis in vascular diseases have generally been suboptimal. This is in part because of several factors: (1) angiogenesis is a complex process that may require more than one growth factor or stem cell type to affect therapeutic angiogenesis; (2) there are host barriers, including the immune system, that neutralize viral-based gene delivery, endogenous nucleases that degrade naked plasmid DNA, and the ability of the host to integrate and express the cloned gene; (3) there is a lack of precise and standardized dosage of genes, growth factors or stem cells to be delivered; (4) most growth factors have a short half-life; and (5) there are technical difficulties in delivering genes, growth factors and cells non-invasively to specific target sites without systemic effects. Specifically, these barriers have been met through the traditional delivery methods that have relied on viral vectors, such as lentivirus and adeno-associated virus, as vehicles or naked plasmid DNA plus or minus a carrier molecule. These vehicles are delivered either intra-arterially, intravenously, or intramuscularly. Below is a brief discussion of four potential delivery methods for growth factor-encoded genes and cell-based therapies dedicated to therapeutic angiogenesis in PAD.

A. In vivo electroporation (EP)

Electric pulse delivery, often called electroporation (EP) or electropermeabilization, for gene transfer and gene therapy is one of the biotechnological and biomedical applications of cell EP.¹¹⁹ In essence, EP is an electrical means of transfecting living cells with nucleic acids. It involves a short, but intense electrical pulse applied in cells either pre-mixed with the DNA of interest or applied in vivo in the muscle or skin that has been injected IM with the DNA of interest. The electrical current polarizes the membrane potential across the cells which creates conductive pores in the cell membrane and, in turn, allows nucleic acids to be transported to the target cells or organs. Neumann et al. developed this technique in 1982 by demonstrating that DNA could be introduced into viable mouse lyoma cells by means of electric pulses.¹²⁰ Since then, this technique has evolved to include in vivo applications in both experimental animal and human trials, dedicated mainly for vaccine development against HIV and cancer therapy.^121-123

However, similar applications utilizing EP as a delivery modality for therapeutic angiogenesis have only been done in experimental animal models of hind-limb ischemia, all of which showed improved limb salvage.^124,125 So far, the application of EP in cancer treatment and vaccination against HIV has been found to be effective and safe. Nonetheless, further studies need to be done before this novel method could be applied as a physical delivery method of IM injected plasmid DNA to treat PAD.

B. Contrast-enhanced ultrasound (CEU)

Additionally, CEU has great potential to be applied as a delivery modality. Ultrasound has been used for decades in medical diagnostic purposes, but its application as in vivo experimental nucleic acid delivery is relatively recent.¹²⁶ CEU applies the principle of sonoporation, which consists of the application of ultrasound to permeabilize cell membranes in order to improve internalization of large molecules.¹²⁷ The mechanism of sonoporation is not well understood but is thought to be via acoustic cavitation, coupled with acoustic pressure.¹²⁸ Ultrasound contrast agents can be used to improve the efficacy of this modality. Typically, a microbubble is a gas-filled molecule consisting of a carrier molecule plus the DNA of interest, which is then infused intravenously. The contrast is then followed by ultrasound to the site of interest (e.g. lower extremity vasculature). The destruction of the microbubble by high ultrasound energy results in the release of the DNA, thus enhancing a site-specific targeted delivery.

Ultrasound waves have been used in vivo to deliver naked DNA into the muscle of mice¹²⁹ and, most notably, in the delivery of VEGF-165-encoded plasmid DNA in a mouse model of hind-limb ischemia.¹³⁰ Kobulnik et al. showed that CEU had lower rates of DNA transfection but resulted in more angiogenesis.¹³⁰ CEU is currently limited to cardiac diagnostic imaging. However, there are still concerns of safety in its application in DNA delivery in the treatment of vascular diseases, and further studies need to be done.

C. Nanoparticle (NP)-based delivery systems

The concept of the NP drug delivery system is borrowed from the term nanotechnology, which means the design and construction of structures from their submicroscopic molecular components (< 100 nm in diameter). Consequently, nanomedicine is an application of nanotechnology and provides opportunities for the diagnosis and treatment of various diseases. Therefore, the design of a NP platform for drug, gene, or growth factor delivery relies on disease-specific cell receptors targeted by the NP. Hence, NP can be designed from molecular subparts that target disease at subcellular and molecular levels. These capabilities not only allow NPs to be used for imaging modalities, but also as a delivery method for therapeutic compounds, the so-called theragnostics.

Thirty years ago, liposomes were used in an experimental model as nanocarriers conjugated to ligands targeting specific cell receptors.¹³¹ Since then several NP platforms have been developed for drug delivery and imaging application, mostly dedicated to cancer.¹³² These developments have provided potential strategies that could be used to design NPs for diagnostic and therapeutic uses specific to vascular diseases. The use of NP as a delivery method of genes, cells, and growth factors to augment neovascularization in the treatment of ischemic disease is especially appealing. The major advantages associated with NPs as carriers for pro-angiogenic factors in ischemic disease include: (1) the ability to achieve targeted administration to specific disease territories without systemic effect; (2) the potential for a sustained exposure and bioavailability; and (3) the ability of an enhanced permeability and retention (EPR) on a larger vascular bed.

Although most in vivo models using NP platforms for the delivery of pro-angiogenic compounds in the treatment of ischemic disease has been limited to growth factors, especially in the field of tissue regenerative medicine, methods utilizing nucleic acid such as miRNAs are also under development.¹³³ Indeed, this technology has been explored to design growth factors that would stimulate therapeutic angiogenesis in vascular disease. To date, a bFGF conjugated to a heparin nanosphere has been shown to have the capacity to function as a controlled, long-term nanocarrier in the delivery of bFGF to efficiently augment angiogenesis.¹³⁴ The interaction between bFGF and heparin allows the nanocarrier to release bFGF in a sustained fashion over 3 weeks. This novel NP was tested in a mouse model of hind-limb ischemia by subcutaneous injection; after 4 weeks the results demonstrated that mice receiving a single injection of the NP had a higher number of capillary growths as compared to mice that received daily injections of free bFGF.¹³⁴ In a similar study, Chappell et al., while utilizing ultrasound-guided microbubble destruction of NP-bearing FGF-2, reported therapeutic angiogenesis in a mouse model of hind-limb ischemia.¹³⁵ Several other in vitro NP growth factor delivery systems have been developed, including VEGF, insulin growth factor-1 (IGF-1), and tumor growth factor-β3 (TGF-β3).¹³³ For example, in a mouse model of limb ischemia, Kim et al. reported improved limb perfusion (1.7-fold) in mice receiving IM injection of VEGF conjugated to fluorescent PEGylated silica NPs (R-SiNPs) as compared to control mice.¹³⁶ Follow-up imaging studies also confirmed that the delivery of the NPs was done in a targeted fashion to the entire vasculature of the ischemic limb (increased accumulation of NPs in days 1 and 3, and gradual decrease in days 7 and 14 post femoral artery ligation), rather than to the entire non-ischemic vascular beds, further strengthening the potential use of NPs for theragnostic purposes in vascular disease.

D. Genetically engineered stem cells

Finally, genetically engineered stem cells have the potential to provide a sustained source of endothelial growth factors and other relevant cytokines to maintain adequate stimulus for stable therapeutic angiogenesis in PAD. Indeed, it has been demonstrated that genetically modified stem cells express tissue-specific paracrine factors that enhance cell survival, migration, and persistence at the injury site.^137,138 There is also an encouraging report demonstrating the efficacy of genetically modified stem cells in ischemic limb.¹³⁹ Typically, the process of stem cell genetic modification involves isolation of autologous cells, which are then transduced with a plasmid DNA encoding a specific endothelial growth factor. The cells are expanded ex vivo and then transplanted back to the patient. This approach provides the capability of not only delivering more than one endothelial growth factor, but also the added benefit of delivering cell therapy in a targeted manner. However, there are also some potential shortfalls. This approach may be time consuming and involves extra steps that may delay therapy, including tissue culture, which is also prone to contamination. Nonetheless, this approach is bound to revolutionize gene- and cell-based delivery methods once further studies are validated.

Conclusion

Therapeutic modalities to affect angiogenesis in the treatment of PAD continue to evolve despite the fact that early clinical trials with growth factors such as VEGF, FGF, and HGF have shown dismal results. This may be explained by the fact that there is lack of a better selection process for genes encoding specific growth factors, the correct dosing, as well as targeted delivery methods that selectively stimulate proliferation, migration, engraftment, and differentiation of resident endothelial cells. Moreover, there is the realization that, perhaps, a single growth factor-based therapy could be suboptimal in stimulating angiogenesis, especially in large territories of ischemia. Nonetheless, there is a sustained effort within the scientific community to develop new strategies that stimulate stable and long-lasting therapeutic angiogenesis. More upstream transcription factors such as HIF-1α, as well as the stimulation or inhibition of other unconventional pathways through small molecules or miRNAs, are also promising. Cell-based therapies continue to improve, especially with the advent of hESCs, iPSCs, and others, which present the opportunity to provide personalized therapies to select patients; engineered stem cells have been developed and this modality has the potential of being utilized to deliver a combination of stem cells as well as specific growth factors in a targeted fashion. Effective future strategies are likely to continue to be developed, including novel delivery methods such as in vivo EP, CEU, and NPs that will complement the current therapeutic angiogenesis modalities in the treatment of PAD.