Engineering vascularized tissues using natural and synthetic small molecules

Abstract

Vascular growth and remodeling are complex processes that depend on the proper spatial and temporal regulation of many different signaling molecules to form functional vascular networks. The ability to understand and regulate these signals is an important clinical need with the potential to treat a wide variety of disease pathologies. Current approaches have focused largely on the delivery of proteins to promote neovascularization of ischemic tissues, most notably VEGF and FGF. Although great progress has been made in this area, results from clinical trials are disappointing and safer and more effective approaches are required. To this end, biological agents used for therapeutic neovascularization must be explored beyond the current well-investigated classes. This review focuses on potential pathways for novel drug discovery, utilizing small molecule approaches to induce and enhance neovascularization. Specifically, four classes of new and existing molecules are discussed, including transcriptional activators, receptor selective agonists and antagonists, natural product-derived small molecules, and novel synthetic small molecules.

Introduction

The vasculature is the first organ to develop and vigorous growth of new blood vessels is essential for the formation of all other organs. Advances in embryology and experiments conducted in a variety of developmental models, including avian, murine and fish embryos, have helped to identify many of the molecular regulators of vascular development and to aid in the understanding of vascular remodeling in adult tissues. While the study of blood vessel growth was once more narrowly focused on the effects of a few known diffusible signals such as vascular endothelial growth factor-A (VEGF-A), researchers are now increasingly aware of the complexity of vascular growth and remodeling, as well as the vast array of both divergent and cooperative molecular signals governing these processes, including numerous soluble effectors and their receptors, mediators of cell-cell and cell-matrix interactions, and matrix degrading enzymes.1 Proper spatial and temporal regulation of these signals is critical to the formation of functional vascular networks, spanning the various arterial, venous, capillary and collateral vessel systems. In addition, impaired or abnormal function of the vasculature is associated with a wide range of disease pathologies, such as coronary artery disease, peripheral vascular disease and osteonecrosis. Animal models such as the dorsal skin and cranial windows, hind limb ischemia and corneal micro pocket have provided insight into the effective coordination of vascular remodeling during injury and adaptive responses. But, despite enormous expenditures supporting basic science research and increasing numbers of patients suffering from ischemia-related vascular diseases, available treatments to address such vascular deficiencies have shown limited effectiveness.

The growth and remodeling of vascular networks in the adult has been described as a continuum of processes spanning angiogenesis, arteriogenesis and vessel maintenance.2 Whereas early studies focused on inducing angiogenesis (formation of new vessels from pre-existing vessels) in areas of ischemia to increase capillary density, it is increasingly recognized that arteriogenesis (remodeling of collateral vessels to accommodate increased flow) is the critical adaptive response to the occlusion of a major arteriole.3 The targeted remodeling of larger collateral vessels via arteriogenesis may also facilitate superior network resistance reduction and a more homogeneous blood flow distribution to sites downstream of an arterial occlusion. Similarly, there is an emerging appreciation of the importance of maintenance via recruitment of pericytes and smooth muscle cells. Numerous studies suggest that without recruitment of such mural support cells, vessels are prone to regression. Thus, the coordination of the processes of angiogenesis, arteriogenesis and vessel maintenance to form an organized network containing arterioles and venules (in addition to capillaries), is absolutely critical to sustaining long-term patency and functionality of microvascular networks. In this review, the term “neovascularization” will be used as an umbrella terminology to describe the stimulation of vascular network growth particularly in circumstances in which it is uncertain whether an intervention is eliciting angiogenesis, arteriogenesis, maintenance or some combination of the three.

Therapeutic Neovascularization

Therapeutic neovascularization is the delivery of growth factors, small molecules, genes or cells to stimulate the growth of new vascular networks in injured, hypoxic or ischemic tissues for the purpose of restoring physiological tissue function. Such strategies seek to enhance the natural process of tissue repair and remodeling, with primary emphasis on ischemic tissue diseases. To date, there have been several randomized, double-blind, controlled phase II/III clinical trials evaluating both protein and gene delivery strategies (outlined in Table 1). These studies evaluated the efficacy of FGF-2 or VEGF₁₆₅ protein, adenoviral vectors encoding VEGF₁₂₁ or FGF-4, and VEGF₁₆₄ plasmid in cases of coronary artery disease (CAD) or peripheral vascular disease (PVD). Although the results point to possible efficacy, there are several areas for improvement, including the mode and method of delivery. For example, intracoronary or intravenous administration of large peptide growth factors may have resulted in ineffective levels of protein in the tissue space. Furthermore, these proteins are known to have a very short half-life.4 Conversely, gene-based approaches yielded longer-term expression of angiogenic factors, but the results were still not definitive. So although there have been several clinical trials to evaluate growth-factor induced therapeutic neovascularization, the trials thus far have failed to show clear evidence of efficacy and no factors to date have been approved for clinical use.

These disappointing outcomes exist despite high expectations of earlier animal studies that show clear revascularization potential in ischemic tissues. VEGF remains the most well-studied angiogenic growth factor for therapeutic neovascularization, but concerns persist regarding its efficacy due to observations of poorly disorganized, leaky and/or hemorrhagic blood vessel formation.2,6 More recent studies now highlight the necessity to stabilize newly formed capillaries via proliferation and recruitment of mural cells to form functional microvascular networks.7 Increasing research emphasis is being directed toward the therapeutic induction of arteriogenesis through the delivery of one or more well-established polypeptide growth factors, such as PDGF-BB (platelet-derived growth factor), Ang1 (angiopoiten 1) and TGFβ (transforming growth factor).8 Temporal delivery of growth factors, such as VEGF₁₆₄ and Ang1 or VEGF₁₆₅ and PDGF-BB, can be utilized to successfully arteriolize nascent vascular networks in adult tissues.7,9 But, while growth factors are powerful and promising tools for developing effective strategies in regenerative medicine, they also possess several potential drawbacks, such as high costs associated with recovery and purification of recombinant proteins and susceptibility to aggregation and degradation.10,11 The central theme of this review is that biological agents used for therapeutic neovascularization must be explored beyond the current well-investigated classes.

Recent advances in medicinal chemistry utilizing computer-assisted combinatorial chemistry and traditional high-throughput screening of synthetic and natural product libraries have resulted in a variety of non-peptide small molecules that are capable of regulating cell and tissue function. The use of small molecules to promote neovascularization is an advantageous strategy to overcome the cost and classical limitations of protein-based therapies. In this review, specific emphasis is placed on potential pathways for novel drug discovery and utilization of small molecule approaches to enhance the native biological revascularization response in ischemic tissues. Four classes of new and existing molecules are reviewed: (1) transcriptional activators with multiple gene targets, including HIF-1α stabilizers; (2) receptor selective compounds that permit greater control over the manipulation of cell fate responses, specifically agonists to G protein-coupled receptors; (3) natural product-derived small molecules, including common dietary supplements; and (4) novel synthetic small molecules resulting from drug discovery programs.

Transcriptional Activators

HIF-1 (hypoxia inducible factor-1) is a heterodimeric transcription factor that is composed of two basic-helix-loop-helix (bHLH) subunits, HIF-1α and HIF-1β. Although the β subunit is constitutively expressed, the α subunit is regulated by oxygen levels; it is degraded rapidly under normoxic conditions and stabilized by hypoxic conditions. In normoxia, the HIF-1α subunit is a target for prolyl hydroxylation of two proline residues by HIF prolyl-4-hydroxylases (PHDs), a process that requires oxygen, iron (II), vitamin C and α-ketogluterate.12 The prolyl-hydroxylated HIF-1α promotes binding of the E3 ubiquitin ligase, von Hippel-Lindau protein (vHLP), targeting HIF-1α for quick degradation by the 26S proteasome. The oxygen-dependent hydroxylation of the asparagine residue in the transactivation domain of HIF-1α also blocks the binding of the coactivator proteins CREB and p300, which are required for transcription.13 Taken together, activation of HIF-1α depends on both proline hydroxylation and asparagine hydroxylation.

In contrast, during hypoxic conditions, the activity of HIF prolylhydroxylase is inhibited, since it utilizes oxygen as a cosubstrate. HIF α and β subunits dimerize in the presence of coactivators CREB and p300, leading to the transcription of over 100 genes that promote survival in low-oxygen conditions. These genes encode proteins that are important in the various stages of angiogenesis, including but not limited to, VEGF, placental growth factor (PLGF), angiopoietin-1 and -2 (Ang1, Ang2) and PDGF, as well as their respective receptors.14 Thus, HIF-1 is sometimes referred to as a “master regulator” because of its ability to either directly or indirectly regulate the cell type-specific expression of multiple angiogenic growth factors or cytokines. Because the limited efficacy of previous clinical trials may in part be attributed to the delivery of only a single factor, an emerging consensus among researchers in this field suggests that a combination of two or more stimulatory molecules, or the use of transcription factors, might yield a more robust induction of new blood vessel growth.15 To this end, strategies designed to increase HIF-1 activity acknowledge the complexity of neovascularization and aim to enhance the expression of multiple factors in order to elicit a more natural and coordinated response.

The potential for a new pharmacological strategy to treat ischemic tissue diseases based on the regulation of HIF pathways has gained momentum over the past decade. The importance of HIF-1 in vascular development was initially realized when HIF-1α^-/- mice were found to be embryonic lethal due to dramatic vascular regression resulting from extensive endothelial cell death.16 Transgenic mice constitutively expressing HIF-1α demonstrated hypervascularity of intact vessels without the associated edema, inflammation or vessel leakage as characterized by VEGF-overexpressing transgenic mice.17 Gene transfer approaches of constitutively active HIF-1 have been shown to be effective in reducing infarct size and enhancing neovascularization in an acute ischemic myocardium18 and in improving perfusion in a rabbit hind-limb ischemic model.19 These results underscore the importance of multiple signaling cues involved in vascular network formation and maintenance of newly formed vessels and highlight a potential niche for targeting transcription factors instead of individual genes. Although gene therapy approaches can be powerful tools to understanding the role of HIF signaling in ischemic tissue diseases, pharmacological approaches may provide cheaper more imminently translatable therapies to induce HIF-1 activation.

Because HIF-1 levels in vivo are determined by the rate of hydroxylation-dependent degradation, the prolyl hydroxylase domain proteins (PHD1, PHD2 and PHD3) hold promise as good targets for therapeutic neovascularization. Prolyl hydroxylases require iron as a cofactor to hydroxylate the two proline residues on the HIF-1α subunit, and thus iron-chelators have been proposed as HIF-1 activators. Recent studies have investigated the role of iron-chelation in inducing angiogenesis in a sponge model.20 Both L-mimosine and ethyl 3,4 dihydroxybenzoate (3,4 DHB) were repeatedly injected into polyurethane sponges and implanted subcutaenously in adult rats. After sponge excision and histological analysis, the cross-sections revealed significant differences in the extent of vascular infiltration into the center of the sponges between those treated with L-mimosine or 3,4 DHB and the vehicle control. Additionally, the induction of HIF-1 target genes within 16 hours of treatment demonstrated the ability of these iron chelators to mediate acute protection from hypoxia, which is clinically relevant in cases of acute ischemia or organ transplants. Dibenzoylmethane (DBM) is a nontoxic bioactive chemical compound found in low levels in licorice plants which also acts as iron chelator and may be a carcinogen antagonist.21 DBM dramatically increased HIF-1α protein levels in both cancerous and non-cancerous cells by preventing proteosomal degradation at a step prior to ubiquitination, resulting in a concomitant increase in VEGF secretion under normoxic conditions.22 Another iron chelator, ciclopirox olamine, has been found to also stimulate HIF-induced transcription of VEGF, leading to enhanced angiogenesis in vivo in both a murine skin wound model and the chicken chorioallantoic membrance model.23

However, the most-established iron chelators to date with the ability to activate HIF-1 are desferoxamine (DFO) and cobalt chloride (CoCl₂). Wang and Semenza and then Gleadle et al. first demonstrated the connection between DFO and HIF-1α stabilization under normoxia. DFO induced high-level expression of both VEGF and erythropoietin mRNA in a dose-dependent manner with kinetics similar to the induction of HIF-1 by hypoxia or CoCl₂. These results were attenuated in the presence iron.24,25 A decade later, the efficacy of DFO in enhancing the expression of HIF-1 regulated genes is still under investigation. Local application of DFO to mobilized latissimus dorsi muscles of adult sheep enhanced neovascularization, specifically capillary density, and recovery of skeletal muscle ischemia.26 Most recently, in vivo administration of DFO into a distraction gap model of bone repair resulted in increased angiogenesis and markedly improved bone regeneration.27 The authors first showed that DFO (and to a lesser extent L-mimosine) increased the formation of tube-like structures in a standard Matrigel tube formation assay with HUVECs and promoted endothelial outgrowth from fetal mouse metatarsels in explant culture in vitro. Subsequently, DFO delivery to the distraction gap every other day for a period of 10 days significantly increased both vessel number and vessel connectivity as well as bone volume compared with controls, all with a lack of overt toxicity. DFO is currently the only iron chelator available for clinical use in the US and is being used for the treatment of iron overload, including acute iron poisoning and chronic iron overload from routine blood transfusions.28 However, new studies such as the ones by Wan et al. and Chekanov et al. also highlight DFO as a potential angiogenic agent to improve blood supply to ischemic tissues.

The iron chelator CoCl₂ was the first transition metal shown to activate HIF-1.29 However, hypoxia-like effects resulting from cobalt treatment were realized much earlier (1950s) and exploited on the US market as a drug for treating anemia in adults and children. It was later pulled from the market due to relative ineffectiveness and high renal toxicity, most likely due to increased oxidative stress in cells.30 Since then, much work has been done to understand the role of cobalt in activation of HIF-1 transcription and its effects on angiogenesis. Tanaka et al. investigated the effects of cobalt on tubulointerstitial hypoxia, a hurdle in progressive renal diseases. The authors found that administration of cobalt to rats that had undergone subtotal nephrectomy significantly improved the tubulointerstitial hypoxia by stimulating endothelial cell proliferation and resultant capillary network formation, while maintaining endothelial barrier integrity.31 CoCl₂ was also found to increase VEGF generation in human microvascular endothelial cells in a HIF-1 dependent manner.32 Furthermore, chemical preconditioning with CoCl₂ induced a delayed preconditioning-like cardioprotection against ischemia-reperfusion injury.33 Specifically, CoCl₂ mimicked acute systemic hypoxia, a current cardioprotective treatment modality,34 reducing myocardial infarct size in a HIF-1α-dependent mechanism. Similarly, chronic treatment with CoCl₂ increased both arteriolar and capillary supply in the rat myocardium.35 Various other small molecule PHD inhibitors with pro-angiogenic potential have also been reported that act through iron chelation, including FG-0041, S956711 and baicalein.20,36,37 Taken together, these studies promote CoCl₂, a low-cost water-soluble iron chelator, as a potential therapeutic in ischemic tissues as it may be an effective inducer of an ischemia-resistant phenotype in various organs and tissues.

Although the aforementioned iron chelators show promise as HIF-1 activators, it is important to note that not all iron chelators stimulate the HIF-1 pathway, two of which include curcumin22 and HBSer.38 Furthermore, iron is a necessary cofactor for many cellular functions, including oxidative phosphorylation and arachidonic acid signaling, so the relative efficacy of iron chelators may be hindered by potential side effects.39 Additional exploration into the mechanisms and promiscuity of iron chelators is warranted.

In addition to iron (II), the HIF prolyl and asparaginyl hydroxylases are also dependent on 2-oxoglutarate (2OG) as a cofactor for the hydroxylation of the HIF-1α subunit. 2OG analogues have been identified that structurally mimic 2OG (a methylene group is replaced by an amine group at position 3), but are unable to stimulate hydroxylation of the primary substrate.40 N-oxaloglycine (NOG) does not permeate cell membranes and has been shown to inhibit enzymes other than the 2OG oxygenases, posing a problem of specificity.41 The ester form of NOG, dimethyloxalylglycine or DMOG, has an increased membrane permeability and is a potent PHD inhibitor. Recent work has shown that inhibiting PHD with DMOG resulted in enhanced HIF-1α activity and subsequent VEGF expression in both human adult lung microvascular endothelial cells and human alveolar epithelial cells.42 Furthermore, DMOG stabilized HIF-1α in human microvascular endothelial cells in vitro and attenuated cytokine-induced neutrophil infiltration into the myocardium, leading to a significant reduction in myocardial infarct size in vivo.43 In a mouse model of hindlimb ischemia, DMOG inhibited HIF PHDs, significantly increasing HIF-1α protein levels, as well as the expression of VEGF and its receptor Flk-1, and stimulated neovascularization in the hindlimb muscles.44 Similarly, in a rat model of ischemia, increased VEGF levels alone did not enhance angiogenesis, but were instead dependent on enhanced levels of Flk-1 as well.45 Recently, a synthetic PHD inhibitor (TM6008) has been discovered that stimulates HIF-1 activity without a co-dependence on iron chelation.39 FibroGen, Inc. (South San Francisco, CA) is interested in the development of HIF prolyl hydroxylase inhibitors related to erythropoiesis, cytoprotection, metabolism and vascular biology and has filed a patent on several compounds that stabilize HIF-1α by inhibiting 2OG dioxygenase enzyme activity.46 Taken together, these studies suggest that stabilization and activation of HIF-1 is a promising strategy for the treatment of ischemic diseases. Potential HIF-1α stabilizers are outlined in Table 2.

Receptor Selective Agonists and Antagonists

G protein-coupled receptors (GPCRs) have been identified as molecular targets because they represent one of the largest gene families in the human genome.47 As of 2006, GPCRs were the target of more than 30% of currently marketed drugs (>200).48 These drugs target only a small fraction of the total members in the GPCR superfamily, so there is tremendous potential within the pharmaceutical industry to develop small molecule therapeutics that target the remaining family members, including more than 100 orphan receptors for which their ligands have yet to be identified.49 GPCRs are activated by various stimuli, such as hormones, nucleosides, neurotransmitters, and lipid mediators to name a few, and initiate second messenger cascades that result in diverse cellular functions. Receptor selective compounds permit greater control over the manipulation of cell fate responses, several of which are discussed here and outlined in Table 3.

Sphingosine 1-Phosphate

Sphingosine 1-phosphate (S1P) is a bioactive sphingolipid that is now in the spotlight for its role in many critical cellular processes, including cell proliferation, migration and survival,50,51 cytoskeletal arrangements and cell motility,52–54 angiogenesis and vascular maturation,55,56 and trafficking of immune cells,57 among others. S1P has been shown to play such diverse roles because it not only functions inside of the cell, but can be secreted extracellularly and act as a ligand of the S1P receptors, a family of GPCRs designated as S1P_1–5.58 The importance of S1P in vascular development was elucidated using global S1P₁^-/- mice. The fate of these mice was embryonic lethality at E12.5–14.5 due to massive hemorrhaging as a result of aberrant smooth muscle cell (SMC) recruitment to nascent endothelial tubes.56 These results suggest that S1P₁ is critical for vascular SMC migration in vivo and that stimulating S1P signaling may prove to be an effective strategy for promoting therapeutic neovascularization in regions of ischemia or tissue insult.

Exogenous delivery of S1P in vitro has been shown to enhance both proliferation and migration of vascular endothelial cells (ECs) and vascular SMCs,59–61 as well as reduce oxygen- and nutrient-deprived cell death.62 Furthermore, S1P has been shown to promote vascular stabilization by recruitment of pericytes and SMCs to newly-formed vessels63 and to stimulate SMC differentiation into a more contractile phenotype.61 Although there are numerous studies performed in vitro that highlight the potential therapy of S1P, few studies have been published that evaluate the efficacy of S1P delivery on neovascularization or wound healing in vivo. Wacker et al. were the first group to develop biomaterials for the controlled delivery of S1P.64 The authors showed that sustained delivery of S1P from poly(ethylene glycol) (PEG) hydrogels enhanced EC migration in vitro and stimulated angiogenesis in a chick chorioallantoic membrance (CAM) assay in vivo, demonstrating that S1P maintained its bioactivity after encapsulation in the hydrogels. A year later, the effects of S1P in wound healing were reported. Daily injection of S1P into full-thickness dorsal wounds of diabetic mice enhanced the rate and extent of wound closure and stimulated neovascularization in the wound area.65 Then in early 2008, the angiogenic actions of S1P in a model of hindlimb ischemia were published. Administration of S1P intramuscularly into the mouse hindlimb dose-dependently stimulated blood flow recovery and increased intramuscular capillary density without increasing vascular permeability.66 Additionally, SPHK1-transgenic mice that overexpress the S1P-synthesizing enzyme sphingosine kinase-1 (SPHK1) showed accelerated blood flow recovery and increased capillary density, compared to controls, due to a 1.8-fold increase in intramuscular S1P levels. To this end, local increases in available S1P, regardless of delivery mode, stimulated therapeutic neovascularization in an ischemic zone.

Although previous studies support the therapeutic use of S1P, the results may have been enhanced using a sustained delivery system that obviates the need for multiple exogenous injections. Most recently, our group evaluated the effects of sustained release of S1P from biodegradable poly(lactic-co-glycolic acid) (PLAGA) films on microvascular remodeling and bone regeneration in vivo.67 Release of S1P from PLAGA significantly enhanced lumenal diameter expansion of arterioles after 3 days, as well as the number of proliferating SMCs on smooth muscle α-actin-stained microvessels, compared to unloaded PLAGA controls. Furthermore, sustained release of S1P in a rat cranial defect model resulted in significantly greater amounts of new bone formation after 2 and 6 weeks of healing that was accompanied by significant increases in the numbers of blood vessels located within the defect area, compared to controls. These experiments, as well as those by Wacker et al. demonstrate that S1P maintains its bioactivity following encapsulation and release from polymeric biodegradable biomaterials and can be utilized as a therapeutic to improve healing outcomes.

There is strong evidence that vascular SMCs are highly specialized cells that retain remarkable plasticity and can undergo profound phenotypic changes during development and neovascularization, adult microvacular remodeling and wound repair.68 However, in contrast to the desirable pro-proliferative effects of S1P on vSMCs in the aforementioned examples, inhibition of vSMC proliferation may also be advantageous in certain disease states. A classic example of SMC phenotypic modulation in adult blood vessels occurs during macrovessel injury. Balloon angioplasty causes an acute vascular injury, resulting in phenotypic switching of medial SMCs.69 These SMCs proliferate and migrate towards the lumen, ultimately leading to a narrowing of the vessel wall and neointimal hyperplasia. S1P has many diverse effects on vascular cells, including SMCs, and these actions are tailored by activation of specific S1P receptors.70 Analysis of S1P signaling in vascular SMCs suggests that the S1P₁ receptor plays a role in SMC proliferation and migration,60,61 whereas the S1P₂ receptor inhibits SMC migration by inhibiting small GTPase Rac71 and stimulates SMC differentiation to a more contractile phenotype.61 Wamhoff et al. recently evaluated the role of S1P receptors in acute vascular injury using specific receptor antagonists.72 Daily injection of S1P₁/S1P₃ antagonist VPC44116 significantly decreased neointimal hyperplasia by approximately 50% following an acute balloon injury of the rat carotid artery compared to vehicle control. Conversely, inhibiting S1P₂ with antagonist JTE013 further increased S1P-induced proliferation of rat aortic SMCs in vitro, supporting the belief of S1P₁/S1P₃-mediated proliferation. Meanwhile, Reidy and colleagues demonstrated that the expression levels of specific S1P receptors vary within inbred mouse strains, leading to significantly different responses to acute vascular injury. S1P₂-null mice exhibited larger neointimal lesions after ligation of the carotid artery compared to wild-type mice, suggesting that S1P₂ normally acts to suppress SMC proliferation.73,74 Taken together, these studies highlight the potential therapeutic benefits of S1P receptor agonists and antagonists to modulate SMC phenotype to a pro- or anti-proliferative state depending on the disease state and desired outcome. For example, activating the S1P₁ and S1P₃ receptors on vSMCs may be an effective strategy for promoting arteriogenesis in ischemic tissues by inducing vSMC proliferation and increasing the diameter of arterioles and blood flow to the area. Conversely, inhibiting the S1P₁ and S1P₃ receptors with a pharmacological antagonist may be advantageous following coronary balloon angioplasty where attenuating vSMC proliferation and neointimal hyperplasia are desirable. Table 4 highlights potential S1P receptor agonists and antagonists and their effects on the vasculature.

Lysophosphatidic Acid

Similar to S1P, lysophosphatidic acid (LPA) is also a bioactive lipid with important biological roles, including the stimulation of cell proliferation, migration, differentiation and survival.75 LPA activates four GPCRs, designated as LPA₁–LPA₄; unlike S1P, however, the physiological roles of individual LPA receptors are not well understood and there have been far fewer studies reporting on the angiogenic potential of LPA. LPA dose-dependently increased EC proliferation and, utilizing an in vitro wound healing model (scratch assay), LPA significantly enhanced EC migration and wound closure compared to control.76 Wu and colleagues demonstrated that EC migration may be enhanced with LPA treatment by regulating the expression of matrix metalloproteinase-2 (MMP-2), an enzyme involved in degradation of the extracellular matrix during angiogenesis.77 Furthermore, LPA was found to stabilize barrier function of ECs in culture, an important event in late stage angiogenesis.78 LPA has also been implicated to play important roles in vascular development, as described using transgenic mouse models. LPA₁-null mice are not embryonic lethal, in contrast to S1P₁ knockouts, but are characterized as semi-lethal resulting from impaired suckling behavior, craniofacial dysmorphism, and decreased postnatal growth rate.79 LPA₂^-/- mice are viable with no obvious phenotypical differences, but exhibit reduced second messenger signaling cascades in response to LPA. Furthermore, LPA₁^-/-–LPA₂^-/- double knockout mice phenocopy LPA₁-null mice, as they are semi-lethal with increased rates of frontal hematomas.

Nearly half a decade ago, researchers showed that autotaxin, a LPA-producing enzyme, is angiogenic itself. ATX incorporation into Matrigel plugs enhanced endothelial tube formation, compared to controls, in a similar fashion to VEGF delivery.80 ATX is a secreted lysophospholipase D that generates LPA and promotes stabilization of nascent vessels; ATX-null mice are embryonic lethal at day E9.5 because properly formed vessels failed to mature.81,82 ATX heterozygous mice were found to have only half the lysophospholipase D activity and LPA levels of wild type mice, demonstrating the importance of ATX in producing LPA.82 Although all of these studies suggest that LPA signaling is critical for normal vascular development, none have directly demonstrated that the lysophospholipid is angiogenic in vivo.

Lynch and colleagues recently evaluated the direct angiogenic actions of LPA, the first study of its kind.83 Utilizing the CAM assay, the authors showed that LPA induced a robust angiogenic response similar to VEGF or S1P treatment. Furthermore, LPA-treated vessels appeared larger than VEGF-treated vessels, suggesting an arteriogenic effect of LPA. LPA-induced angiogenesis was found to be mediated by the LPA₁ and LPA₃ receptors, since delivery of LPA₁/LPA₃ antagonist VPC32183 blocked the response. The angiogenic actions of LPA were direct and were not a result of a non-specific inflammatory response; the anti-inflammatory corticosteroid hydrocortisone was unable to block LPA-induced angiogenesis. Although it is now promising that LPA is angiogenic in vivo, more experiments should be performed in rodents using selective pharmacological compounds to target specific LPA receptors to further elucide the role of LPA in microvascular remodeling. Table 5 outlines LPA receptor agonists and antagonists that are currently available.

Adenosine

Adenosine is a potent vasodilator produced in response to tissue hypoxia or ischemia that acts to maintain oxygen delivery within a physiological range. Adenosine signals through a family of GPCRs, of which four receptor subtypes have been identified: A₁, A_2A, A_2B and A₃.84 Although well-known for its vasodilatory effects, adenosine also maintains oxygen levels in ischemic tissues by promotoing the formation of new blood vessels. Over two decades ago, adenosine was shown to stimulate angiogenesis in a CAM assay.85 Since then, many investigators have shown that adenosine or adenosine analogues stimulate EC migration and proliferation in vitro in a concentration-dependent manner.86 Similarly, adenosine promotes vascular growth in vivo in several tissues, including skeletal muscle, myocardium and wound beds.87,88 For example, Ziada et al. showed that intravenous delivery of adenosine in rabbits resulted in approximately a 27% increase in capillary density in the myocardium and in approximately a 26% increase capillary-to-muscle fiber ratio in skeletal muscles.89

The main pro-angiogenic effects of adenosine have so far been attributed to adenosine-induced secretion of pro-angiogenic growth factors, such as VEGF and bFGF in an A_2A receptor-mediated fashion.90 Importantly, Lebovich and colleagues identified a unique mechanism whereby activation of A_2A converts macrophages from a pro-inflammatory state that generates tumor necrosis factor-α (TNFα) and interleukin-6 (IL-6) to a pro-angiogenic phenotype that primarily secretes VEGF.91 The A₁ receptor was also shown to play a role in angiogenesis as A₁ receptor agonist CPA enhanced neovascularization by 40% compared to controls in a mechanism that appeared to involve A₁-mediated VEGF secretion from monocytes.92 Taken together, adenosine receptor agonists could play a role as therapeutic targets to enhance angiogenesis either directly or indirectly through production of pro-angiogenic factors. To this end, small molecule agonists and antagonists to adenosine receptors have been cloned and developed and their efficacy in promoting therapeutic neovascularization is under investigation.93

Natural Product-Derived Small Molecules

Natural product-derived small molecules are an attractive option to treat various diseases. In general, they have wide acceptability, and are well tolerated and economical. Historically, natural compounds have provided structural platforms upon which significant volume of current drug products are derived. Between 1981 and 2002, 10% of new products introduced to the drug market were natural products and 61% were natural product-derived.94,95 Some of the most commonly prescribed antibiotics are derived from natural products including â-lactams (cf. penicillins and cephalosporin C), macrolides (i.e., erythromycin), and aminoglycosides (i.e., streptomycin).96–100 More recently, epidemiologists have uncovered new plant-based therapies by identifying dietary habits of various ethnic groups and correlating them to diseases such as coronary artery disease or cancer.94 Others have taken the strategy to perform large scale screening of traditional Chinese medicinal herbs to determine possible angiogenic effects. For example, 47 crude plant extracts were recently tested in a CAM assay and in vitro bovine aortic endothelial cell (BAEC) proliferation assays. Of these, 24 demonstrated angiogenic properties with Epimedium sagittatum, Trichosanthes kirilowii and Dalbergia odorifera performing the strongest in both CAM and BAEC models.101

Resveratrol

The “French paradox” named by epidemiologists describes the unique observation that, despite a diet known to be high in both cholesterol and saturated fats, people of southern France are less prone to ischemic heart disease. Subsequently, a link between moderate consumption of red wine and low incidence of coronary artery disease was hypothesized.102,103 The medicinal component of red grapes, resveratrol, was first identified in 1940 and recent emerging evidence has supported the diverse biochemical and physiological effects of this compound including estrogenic, anti-platelet, anti-inflammatory, anti-apoptosis, ROS scavenging and angiogenic properties.104 Specifically, in terms of angiogenic activity, endothelial progenitor cells exposed to resveratrol showed enhanced proliferation, migration and adhesion. Enhanced tube formation was accompanied by increased VEGF expression. Cell cycle analysis suggested mitogenic effects of resveratrol are achieved via a shift to the G₁/S transition.105 The therapeutic effects of resveratrol in improving cardiac function following infarction are most potent when used as a preventative treatment to exert such pre-conditioning effects increased capillary density, improved left ventricular function, and increased expression of anti-apoptotic and pro-angiogenic factors NFκB and Sp-1.106 Oral administration of resveratrol showed increased VEGF and tyrosine kinase receptor Flk-1 expression three weeks following myocardial infarction in a rat model. Exogenous incubation with resveratrol leads to an increase in both eNOS protein expression and eNOS-derived nitric oxide production in human umbilical vein endothelial cells (HUVEC).107 Three-day incubation with exogenous resveratrol lead to increased iNOS expression in cultured bovine pulmonary artery endothelial cells.108 Furthermore, increased expression of iNOS, VEGF, Flk-1 and eNOS occurred in a time-specific manner to resveratrol. Taken together, these results suggest resveratrol may be a promising candidate for therapeutic neovascularization strategies.

Interestingly, emerging evidence from literature suggests that resveratrol also plays an effective role in anti-cancer therapies. This compound has been shown to effectively kill cancer cells by inhibiting cell survival and anti-apoptotic signaling pathways. These contrasting physiological effects of resveratrol can be correlated to differences in dosing concentration. At low dosage (2 ng/mL), resveratrol induces angiogenic actions, whereas at higher dosages (1.2 mg/L), it can inhibit cell survival and angiogenic pathways. Furthermore, pharmacokinetics, bioavailability and metabolism likely play a critical role in appropriate delivery strategies to produce appropriate therapeutic outcomes. Specifically, oral administration leads to rapid absorption to plasma, liver, kidney and heart.109 Therefore, tissue-engineered controlled release systems may be particular advantageous for sustained low dosage delivery of resveratrol to areas of ischemia insult.

Ginseng

Ginseng, named by botanist Carl Meyer, was derived from Greek to mean all-healing (pan akos). Ginseng has historically been a key ingredient to Chinese medicine and is also the most extensively used herbal medicine in the western world.110 Annual sales exceed $200 million in the US alone.94,111 To date, more than 30 ginsenosides, a class of steroid-like compounds that are the principle active components in ginseng, have been identified. Interestingly, differences in vascular response to various extracts of ginseng have recently been correlated with different ratios of ginsenosides. HPLC studies coupled with neovascularization assays revealed different ratios in ginsenoside subgroups based on geographic location of the plant resulting in altered physiologic effect. Specifically, Panax and American ginseng have a predominance of Rb1, effective in preventing cancer in multiple animal models.112,113 In contrast, the predominant ginsenoside in Sanqi ginseng is Rg1, demonstrating proangiogenic activity. A standardized avascular polyether polyurethane scaffold implant was used to assess morphological and functional neovascularization in which strong immunostaining for von Willebrand Factor (vWF), a blood glycoprotein required for normal hemostasis, was observed in Rg1 treated groups compared with controls. Variations in Rg1:Rb1 dosing ratios demonstrated that larger percentages of Rg1 resulted in significant induction of angiogenesis within the implant compared with controls. In contrast, co-delivery of large amounts of Rb1 inhibited Rg1-induced neovascularization.110 This study confirms that ginseng subtype Rb1 is a promising candidate for wound healing and tissue regeneration applications.

VEGF-mediated angiogenesis depends on increased NO expression. In a scratch wound assay, Rg1 exerted a concentration dependent effect on HUVEC proliferation compared with controls.110 Rg1-induced HUVEC proliferation was abolished in the presence of L-NAME, a non-selective inhibitor of NOS, thus suggesting an NO-dependent proliferative response. Others have demonstrated Rg1 treatment activates eNOS in HUVEC thus enhancing production of NO following treatment.114 Increased VEGF and NO expression in HUVECs following Rg1 treatment is mediated via the glucocorticoid receptor (GR).114 Others have also confirmed enhanced HUVEC proliferation in response to exogenous Rg1 in addition to demonstrating enhanced tube formation, microvascular outgrowth from aortic rings, and neovascularization in a subdermal implant in response to Rg1 delivery. Furthermore, gene expression profiling in Rg1-treated HUVEC suggest upregulation of genes involved in cell adhesion, migration and cytoskeleton restructuring (RhoA, RhoB, IQGAP1, CALM2, Vav2 and LAMA4).115 The ginsenoside, Re, has also shown many similar pro-angiogenic properties to that of Rg1.116,117

Traditionally, steroids such as dexamethasone are principally thought to disrupt capillary function and abate endothelial cell proliferation.118 However, as discussed above, Rg1 ginsenoside treatment promotes neovascularization. This discrepancy between steroid classes may be related to dosage. It is thought that high steroid dosage may lead to inhibition of neovascularization whereas low steroid dosage may promote angiogenic activities.119 Interestingly, comparisons of binding affinity for GRs show that Rg1 is 10-fold less likely to bind to its GR than dexamethasone and 100-fold less likely to activate gene transcription when in direct competition with dexamethasone.114,120 Furthermore, differences in the processing of raw ginseng extracts can also alter the structural fingerprint, and thus phenotypic outcome.110 To this end, observations of diverse physiological effects resulting from differences in species, geography location, processing technique, and dosage will require further investigation and greater government regulation if the successful implementation of ginseng derivatives as angiogenic therapeutics is to occur.

Curcumin

Active compounds in turmeric, a yellow-colored spice prevalent in both Asian and Indian cooking, are curcumoids. These are divided into three subgroups; curcumin (curcumin I), demethoxycurcumin (curcumin II) and bisdemethoxycurcumin (curcumin III).121 Both oral and topically treatment of curcumin have demonstrated a variety of pharmalogical effects including anti-inflammatory,122,123 anti-cancer,124 anti-oxidant125,126 and anti-microbial127 properties. Several studies have highlighted the pro-angiogenic properties of curcumin by both oral and topical application. Specifically, serum-free curcumin application showed significant upregulation of CD31,128 E-selectin,129 and both VEGF and VEGFR-2.121 Additionally, Kiran and colleagues showed significant capillary network formation in HUVEC and RAEC cultures, vessel sprouting in aortic ring assays, and increased vascular density in CAM assays in response to curcumin treatment.121 In contrast, HUVEC cultured in serum-containing media with curcumin application demonstrated anti-angiogenic properties including a decreased expression of CD31, E-selectin, Ang-1, VEGF and VEGFR-2 as well as a lack of capillary network formation.121 Therefore, knowledge of the local microenvironment and possible interactions with co-delivered compounds will be critical to elicit an appropriate biological response. Additionally, appropriate dosing for therapeutic application is particular challenging given than turmeric may exist in the diet and is known to produce effects when administered both orally and systemically. Furthermore, both oral consumption and i.p. injection of curcumin resulted in approximately 75% excretion suggesting low absorption within the GI tract, highlighting a potential opportunity for a tissue-engineering controlled release system.130,131

Curcumin is also an effective treatment for accelerating skin wound closure in both diabetic rats and genetically diabetic mice. Specifically, both oral and topical treatment with curcumin showed enhanced neovascularization, increased migration of macrophages and dermal myofibroblasts to the wound zone, and higher levels of collagen and TGFβ expression within the wound zone.132 Others have also demonstrated enhanced production of TGFβ and extracellular matrix protein and increased formation of granulation tissue in rat and guinea pig models.133 In mice receiving whole-body γ-radiation, pretreatment with oral doses of curcumin accelerated cutaneous wound closure and significantly increased vascular density, collagen production, nitric oxide expression and DNA synthesis following within the wound zone following irradiation.134 Others have developed slow release systems using collagen gels were developed to provide controlled delivery of curcumin to the wound site. Results showed application of curcumin-loaded gels accelerated wound closure, enhanced cell proliferation, and was effective in scavenging free radicals compared with untreated and collagen-treated controls.135

Sokotrasterol Sulfate

Currently, there are more than a dozen natural products isolated from marine sponges undergoing clinical or advanced pre-clinical assessment. Sponges are chemically diverse and physiological activity is well-documented. Therefore, Karsan et al. screened a library of crude marine extracts to target possible candidates with the ability to stimulate vessel remodeling for therapeutic angiogenesis. Screening led to identification of a promising target, sokotrasterol, a sulfated steroid. Preliminary experiments demonstrated that sokotrasterol induced sprouting of HUVECs with a bell curve dosedependency.136 Maximal sprouting occurred at 5 µg/ml loading. Murphy and colleagues demonstrated sokotrasterol-loaded gelatin sponges stimulated new vessel growth in a CAM assay similar to levels of FGF-2. Furthermore, local delivery of sokotrasterol to a hindlimb ischemia model accelerated reperfusion and increased vascular density compared with vehicle controls. Vessel sprouting via sokotrasterol stimulation was determined to be dependent on cycloxygenase-2 (COX-2) activation, VEGF expression, and the integrin αV-β3. Furthermore, fibroblasts did not respond to sokotrasterol stimulation suggesting that the angiogenic effects of this compound are selective. The synthetic analog of sokotrasterol, cholestanetrisulfate, was equally affective in restoring reperfusion to occluded hindlimbs. Additionally, this analog can be easily synthesized in a six-step process with 90% yield.137 These results suggest marine sponge extracts may be promising natural-based candidates for therapeutic vascularization applications.

Synthetic Small Molecules

Recent advances in computer assisted drug discovery and high-throughput phenotype screens of synthetic and natural product libraries have resulted in a variety of non-peptide small molecules that are capable of regulating cell and tissue function. The discovery of new small molecule therapies that specifically promote neovascularization may allow for the creation of functional vascular networks without the cost and classical limitations of protein-based therapies. Our colleagues have previously reported on the discovery of novel modulators of angiogenesis for treatment of cancers.138,139 From this program emerged phthalimide neovascularization factor (PNF1), a novel small molecule inducer of angiogenesis (Fig. 1). PNF1 is a structural analog of thalidomide with a cyclopentanone substitution on the imide group. PNF1 possesses a molecular weight similar to that of known small molecule therapeutics. In contrast to endogenous peptide-based growth factors, including VEGF, PNF1 is a much smaller molecule (38.2 kDa and 229 amu, respectively). In addition, this small organic molecule might better withstand processing methods commonly used for incorporation into and sterilization of polymeric biomaterials that may compromise peptide-based growth factors.140–142

In vitro evaluation of the compound demonstrated efficacy in enhancing proliferation, survivability, and capillary network formation by human microvascular endothelial cells (HMVEC). Preliminary in vivo studies also indicated an early stage angiogenic response induced by the drug in rat mesenteric windows.143 Subsequently, network analysis tools were utilized to identify genetic networks of the global biological processes involved in PNF1 stimulation, and to describe known molecular and cellular functions that were most highly regulated by the drug. The most significantly perturbed networks identified gene products associated with transforming growth factor-β (TGFβ) after 24 hours, a protein which has many known implications on angiogenesis.144 Further interrogation of PNF1's mechanism using a novel compendium method identified both TNFα and TGFβ as signaling pathways associated with drug function over 1–48 hours.145 In vivo delivery of PNF1 at 5% loading was examined in the dorsal skinfold window chamber model. Results show that structural microvascular network remodeling induced by PNF1 is both angiogenic and arteriogenic. Specifically, controlled release of this drug led to spatial and temporal increases in location-dependent vessel density and microvessel diameter in treated tissues, with maintenance of functional vascular patterns. Furthermore previously observed increase in microvessel length density induced by PNF1 was abolished in CCR₂^-/- mouse chimeras, suggesting a critical role for monocyte recruitment and promotion of arteriogenic processes on PNF1 function. Collectively, these experiments suggest that PNF1 is effective for therapeutic induction of neovascularization for possible treatment of ischemic tissue disorders and vascularization of engineered tissues.146

Conclusions

Many significant age-associated diseases, such as myocardial infarction, peripheral limb ischemia or osteonecrosis, arise from impaired or abnormal function of the microvasculature, the complex network of small vessels transporting blood and nutrients to and from various tissues in the body. As a consequence, the development of effective strategies to promote the growth of mature vascular networks is an important clinical need. The small molecule approaches reviewed here hold promise as successful inducers of therapeutic neovascularization for the treatment of ischemic tissue diseases and provide clear alternatives to protein- and gene-based approaches. Further investigation into the basic science of how natural product-derived small molecules elicit cellular responses, as well as cultivating novel drug discovery programs for synthetic small molecules, is needed to realize the full promise of new pharmacological approaches for induction of neovascularization.

Abbreviations

VEGF	=	vascular endothelial growth factor
FGF	=	fibroblast growth factor
PDGF	=	platelet derived growth factor
Ang1	=	2, angiopoietin 1, 2
TGFβ	=	transforming growth factor-β
HIF-1	=	hypoxia inducible factor-1
PHD	=	prolyl-4-hydroxylases
vHLP	=	von Hippel-Lindau protein
PLGF	=	placental growth factor
3,4 DHB	=	3, 4 dihydroxybenzoate
DBM	=	dibenzoylmethane
DFO	=	desferoxamine
2OG	=	2-oxoglutarate
NOG	=	N-oxaloglycine
DMOG	=	dimethyloxalylglycine
GPCR	=	G protein-coupled receptor
S1P	=	sphingosine 1-phosphate
LPA	=	lysophosphatidic acid
SMC	=	smooth muscle cell
EC	=	endothelial cell
PEG	=	poly(ethylene glycol)
SPHK1	=	sphingosine kinase 1
PLAGA	=	poly(lactic-co-glycolic acid)
MMP-2	=	matrix metalloproteinase-2
ATX	=	autotaxin
TNFα	=	tumor necrosis factor-α
IL-6	=	interleukin-6
BAEC	=	bovine aortic endothelial cell
HUVEC	=	human umbilical vein endothelial cell
NO	=	nitric oxide
PNF1	=	phthalimide neovascular factor 1
HMVEC	=	human microvascular endothelial cell

Figures and Tables

Figure 1 Synthesis of phthalimide neovascular factor 1 (PNF1), a synthetic small molecule with angiogenic properties.143

Table 1 Summary of select clinical trials of therapeutic angiogenesis

Trial	Disease	Treatment	Conclusions
VIVA	CAD, chronic angina	VEGF165 protein; IC, IV	Significant improvement in angina class in high-dose VEGF group
FIRST	CAD, chronic angina	FGF2 protein; IC	Placebo effects; reduced angina in FGF2-reated patients; greater effects seen in highly symptomatic patients (>65 yrs)
TRAFFIC	PVD, atherosclerosis	FGF2 protein; IA	Significant improvement at day 90 for single-dose FGF2
AGENT	CAD, chronic angina	FGF4 gene, adenoviral vector; IC	No significant differences in exercise treadmill test or angina improvement; response better with low-dose than high-dose

* modified from ref. 5; VIVA, Vascular Endothelial Growth Factor in Ischaemia for Vascular Angiogenesis Trial; FIRST, Fibroblast Growth Factor Initiating Revascularization Trial; TRAFFIC, Therapeutic Angiogenesis with Recombinant Fibroblast Growth Factor-2 for Intermittent Claudication; AGENT, Angiogenic Gene Therapy Trial; CAD, coronary artery disease; PVD, peripheral vascular disease; IC, intracoronary; IV, intravenous; IA, intra-arterial; VEGF, vascular endothelial growth factor; FGF, fibroblast growth factor.

Table 2 Chemical structures of HIF-1αstabilizers

3,4 DHB, dihydroxybenzoate; DBM, dibenzoylmethane; DFO, desferoxamine; NOG, N-oxaloglycine; DMOG, dimethyloxalylglycine.

Table 3 Chemical structures of receptor selective compounds

S1P, sphingosine 1-phosphate; LPA, lysophosphatidic acid.

Table 4 Vascular-related effects of S1P receptor agonists and antagonists

Compound	Agonist Activity	Antagonist Activity	Commercial Availability	Vascular-related effects
FTY720	S1P1 S1P3 S1P4 S1P5		Cayman Chemical	Inhibits VEGF-induced vascular permeability, mediates EC adherens junction formation, increases EC survival [147]; enhances EC migration [148]; decreases EC, SMC proliferation at high doses (>1µM) [149]
SEW2871	S1P1		Tocris Biosciences	Prevents monocyte/EC interactions in diabetic mouse model [150]; inhibits VEGF-induced EC sprouting in spheroid assay [149]
VPC01091	S1P1 S1P4 S1P5	S1P3	None	Increases neointimal hyperplasia (SMC) proliferation) after acute balloon injury [72]
JTE-013		S1P2	Tocris Biosciences	Increases EC, SMC migration [151]; improves diabetic wound healing [65]; increases SMC proliferation, suppresses S1P-induced SMC differentiation [72]; enhances EC tube formation [152]
VPC23019		S1P1 S1P3	Avanti Polar Lipids, Inc.	Attenuates 11,12-EET-induced EC proliferation [153]; no effect on S1P-induced proliferation of mesodermal progenitor cells [154]; enhances S1P-induced vasoconstriction [155]
VPC44116		S1P1 S1P3	None	Attenuates neointimal hyperplasia (SMC proliferation) after acute balloon injury [72]
VPC25239		S1P1 S1P3	None	Decreases SMC proliferation, potentiates S1P-induced SMC differentiation [72]
W146		S1P1	Avanti Polar Lipids, Inc.	Dose-dependently increases capillary leakage in skin and lung [156]

VEGF, vascular endothetial growth factor; EC, endothelial cell; SMC, smooth muscle cell; S1P, sphingosine 1-phosphate; 11, 12 EET, cis-epoxyeicosatrienoic acids with angiongenic properties.

Table 5 Vascular-related effects of LPA receptor agonists and antagonists

Compound	Agonist Activity	Antagonist Activity	Commercial Availability	Vascular-related effects
2(S)-OMPT	LPA3		Avanti Polar Lipids, Inc.	Induces pericyte-like cell proliferation [157]; promotes hypertrophy in neonatal rat cardiomyocytes [158]; induces angiogenesis in CAM assay [83]
VPC31143	LPA1		Avanti Polar Lipids, Inc.	Increases mean arterial pressure, decreases heart rate after intravenous injection in rats [159]
VPC32183		LPA1 LPA3	Avanti Polar Lipids, Inc.	Attenuates 2(5)-OMPT-induced angiogenesis in CAM assay [83]; blocks LPA-induced intracellular [Ca2+] in skeletal muscle cells [160]
VPC12249		LPA1 LPA3	Avanti Polar Lipids, Inc.	Attenuates LPA-induced intracellular [Ca2+] in skeletal muscle cells [160]
Ki16425		LPA1 LPA3	Sigma-Aldrich	Blocks LPA-induced expression of SMC marker alpha-SMA [161]; inhibits low-dose LPA-induced monocyte migration, enhances monocyte migration in presence of high-dose LPA [162]
DGPP 8:0		LPA3	Avanti Polar Lipids, Inc.	Inhibits LPA-induced transcellular resistance in endothelium [163]; attenuates LPA-induced intracellular [Ca2+] in skeletal muscle cells [160];

LPA, lysophosphatidic acid; CAM, chick chorioallantoic membrane; SMC, smooth muscle cell; SMA, smooth muscle actin; DGPP, dioctylglycerol pyrophosphate.

Table 6 Chemical structures of natural small molecules

Acknowledgements

This work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases grant AR-052352-01A1 to Dr. Botchwey. Lauren Sefcik has been supported by predoctoral fellowships from the National Science Foundation and American Heart Association and Dr. Petrie Aronin by the University of Virginia Biotechnology Training Program grant 5T32GM008715-08.