This Article Abstract Full Text (PDF) Correction (v96,pe7) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Frantz, S. Articles by Kelly, R. A. Search for Related Content PubMed PubMed Citation Articles by Frantz, S. Articles by Kelly, R. A. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Related Collections Receptor pharmacology Angiogenesis Cell signalling/signal transduction Gene regulation Growth factors/cytokines Ischemic biology - basic studies Physiological and pathological control of gene expression

(Circulation Research. 2005;96:15.)
© 2005 American Heart Association, Inc.

Reviews

Innate Immunity and Angiogenesis

Stefan Frantz, Karen A. Vincent, Olivier Feron, Ralph A. Kelly

From the Genzyme Corporation (K.A.V., R.A.K.), Cambridge Mass; Medizinische Universitätsklinik Würzburg (S.F.), Würzburg, Germany; University of Louvain (O.F.), Brussels, Belgium.

Correspondence to Ralph A. Kelly, MD, Genzyme Corporation, 15 Pleasant Street Connector, PO Box 9322, Framingham, MA 01701-9322. E-mail Ralph.Kelly{at}genzyme.com

This Review is part of a thematic series on Angiogenesis, which includes the following articles:

Endothelial Progenitor Cells: Characterization and Role in Vascular Biology

Bone Marrow–Derived Cells for Enhancing Collateral Development: Mechanisms, Animal Data, and Initial Clinical Experiences

Influence of Mechanical, Cellular, and Molecular Factors on Collateral Artery Growth (Arteriogenesis)

Innate Immunity and Angiogenesis

Syndecans

Growth Factors and Blood Vessels: Differentiation and Maturation
Ralph Kelly Guest Editor

	Abstract

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
Activation of an innate immune response is among the first lines of defense after tissue injury. Restoring blood flow to the site of injured tissue is often a necessary prerequisite for mounting an initial immune response to pathogens and for subsequent initiation of a successful repair of wounded tissue. The multiple links among pathogen recognition and suppression, increased angiogenesis, and tissue repair are the topics of this review, which examines of the roles of antimicrobial peptides, mammalian toll-like receptors (TLRs), inflammatory cytokines, and putative "danger" signals, among other signaling pathways, in triggering, sustaining, and then terminating an angiogenic response.

Key Words: angiogenesis • cytokines • HIF • inflammation

	Introduction

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
One of the earliest physiologic responses to tissue injury or infection is an increase in vascular permeability and blood flow to the affected area, initiated by regional vasodilation and followed by enhanced angiogenesis to facilitate the wound healing process. Bacterial-derived lipopolysaccharide (LPS), for example, can activate endothelial sprouting directly by signaling through mammalian toll-like receptors (TLRs) and TRAF6, even in the absence of additional cytokines.¹

The concomitant increase in vasodilation is attributable in part to increases in local tissue concentrations of adenosine and adenine nucleotides, which increase rapidly after cellular injury or ischemia. Animals with targeted disruption of adenosine A_2A receptors exhibit less inflammation but also less wound-related angiogenesis after injury.² Several mammalian TLRs (specifically TLRs 2, 4, 7, and 9), in the presence of activating ligands such as LPS or unmethylated CpG motifs, have been shown to synergize with adenosine, acting at A_2A receptors, to result in increased synthesis and release of vascular endothelial growth factor (VEGF)-A isoforms. This results in vasodilation, recruitment of CD31⁺ endothelial progenitor cells, and an acceleration of the induction and extent of inflammation-induced angiogenesis while concurrently suppressing any further release of inflammatory cytokines (eg, tumor necrosis factor [TNF]).^3,4 In contrast, in the absence of adenosine release, the activation of these TLRs increased TNF expression, but not VEGF-A (at least in murine macrophages).

This recent recognition of a role for adenosine in the regulation of the innate immune response highlights the debate concerning the mechanisms that control of the activation and suppression of this arm of the immune system. Janeway was among the first to highlight the importance of a contextual alarm signal indicating the presence of infection or "microbial nonself" >15 years ago when he pointed out that in the absence of pathogen-related "adjuvants" (ie, essentially admixtures of bacterial cell wall components and/or other microbial debris), little or no immune response was usually elicited to many foreign pathogens.⁵ He postulated the presence of "pathogen-associated molecular patterns," highly conserved motifs that are often essential for microbial survival and virulence that are detected by "pattern recognition receptors" on host immune cells. This theory also accommodated the fact that cells not expressing major histocompatibility complex class I proteins on their plasma membranes or cells expressing "markers of abnormal self," such as those infected with viral pathogens, trigger apoptosis and engulfment by NK cells and other phagocytes.⁶

Eight years after Janeway first posited the existence of a pathogen recognition system that could provide the appropriate immunologic context, or "signal two," after infection by a specific pathogen, Medzhitov, Preston-Hurlburt, and Janeway identified the first of the vertebrate toll receptors, now designated TLR4. Drosophila toll was well recognized to play an essential role in dorsal/ventral patterning in the fly during development by binding to its ligand, spätzle. After development, toll becomes an important signaling pathway of the innate immune system of the fly.⁷ Although no vertebrate homologue of the fly ligand spätzle has been identified, to date, 13 mammalian paralogues of Drosophila toll have been identified, as well as a number of their ligands⁸ (Figure 1). In addition, although invertebrates lack an adaptive immune response, this does not mean that rapid and appropriate immune responses do not occur.⁹ Specific pathogen recognition by the appropriate pattern recognition receptor then triggers an innate immune response appropriate for the identified pathogen.

View larger version (79K): [in this window] [in a new window]   Figure 1. Toll-like receptor (TLR) signaling pathways. The TLRs are among the best characterized of the innate immunity "pattern recognition" proteins. TLRs are named for Drosophila toll, a protein essential for normal dorsal–ventral patterning during development in the fly, but which then assumes a central role in innate immunity. The figure illustrates the known vertebrate tolls and their intermediate signaling cascades. Most mammalian TLR responses signal via the adapter protein MyD88 or by the TIR domain adapter protein TIRAP, among others. Although TLR signaling is known to be required for initiating an innate immune response to pathogens and for triggering an appropriate adaptive immune response, low-level TLR activation has recently been shown also to play an essential role in preserving homeostasis between host cells and commensal organisms in the gut and presumably other tissues. Figure modified from Beutler B. Innate immunity: an overview. Mol Immunol. 2004;40:845–859, with permission from Elsevier.

An alternative hypothesis to Janeway’s "recognition of microbial nonself" hypothesis for activation of an immune response to infection or injury has been proposed by Matzinger in the form of the "danger" model, which proposes alternative mechanisms for the induction of an appropriate immune response. In this construct, the presence of potentially infectious "nonself" does not necessarily trigger an immune response unless there is evidence of tissue injury, which she termed "alarm" signals. One example of the contextual importance of the alarm signal in the "danger" construct is LPS, which, when "buried" within the lipid membrane of bacteria in the gut, is harmless unless unmasked in the context of injury, which then results in a "danger" signal and a robust immune response.¹⁰

The importance of adenosine release in the synergistic activation of the innate immune response, as discussed, is relevant here. Heath and Carbone,¹¹ for example, have proposed a number of possible "danger" signals. Extracellular adenosine, for example, accumulates rapidly in ischemic tissue because of the metabolism of ATP and hypoxia-enhanced activity of Ecto-5'-Nucleotidase, not only inducing vasodilation but also possibly acting as a "danger" signal in the context of infection and/or injury.¹² Thus, the release of adenosine and the activation of specific TLRs may provide a contextual "injury" signal, as well as a trigger for appropriate tissue repair processes, such as the induction of angiogenesis.

Additional evidence in favor or the danger hypothesis has recently been reported. Shi, Evans, and Rock¹³ identified a chromatographic fraction obtained from BALB/c3T3 cells exposed to UV light that acted as a low-molecular-weight "endogenous adjuvant" for priming CD8⁺ T cells in mice injected with HIV gp120 antigen. Using several complementary assays, Shi et al¹³ identified a moiety from cellular lysates as uric acid that acted as an effective immunologic adjuvant when in its nonsoluble, crystalline form. As noted by Heath and Carbone,¹¹ in addition to the adjuvant role noted, crystalline uric acid may also play a role in directly triggering an immune response—data that are obviously relevant to the pathophysiology of gouty arthritis.

There is also mounting evidence that heat shock factors in vertebrates can act as a contextual "danger" signal, in addition to "conventional" pathogen-associated molecular patterns such as a LPS or lipotechoic acid. As shown by Prohászka and Fst (Table 4¹⁴), there is mounting evidence for heat shock protein (HSP) recognition and signaling via several TLRs (vide infra). However, as noted by Reis e Sousa,¹⁵ with several possible exceptions, TLRs 2 and 4 appear to be the most common "receptors" for these endogenous ligands, raising at least the question of contamination by endotoxins in the assays used in several of these reports.

TLR signaling is not limited to pathogen recognition and triggering of appropriate innate and adaptive antimicrobial immune responses. Components of the innate immune system appear to play an essential role in maintaining normal mucosal cell homeostasis in the gut, among other organs.¹⁶ Rakoff-Nahoum et al¹⁷ observed that mice with targeted deletion of MyD88, an obligate intracellular adapter molecule in several TLR signaling pathways, exhibited markedly increased bowel inflammation after exposure to the colonic toxin, dextran sulfate sodium, despite the absence of this critical TLR adapter protein. Both TLR4 and TLR2 knockout mice also exhibited increases in bowel inflammation, leading to colonic bleeding and ulceration. These data suggest that specific TLRs, in the presence of normal colonic microflora, trigger signaling via MyD88 that suppresses inflammation in normal colonic mucosa. TLR signaling is known to induce in normal colonic mucosa increased expression of several HSPs (eg, hsp25 and hsp72¹⁸).¹⁷ This HSP-mediated cytoprotective response was markedly suppressed in MyD88^–/– mice.¹⁷

	Wound Repair and Angiogenesis

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
Repair of wounded tissue involves a well-orchestrated sequence of events, initiated by platelet aggregation and fibrin polymerization, followed by infiltration of leukocytes.¹⁹ Among the earliest cues for initiation of the wound repair response is the release of "growth factor"-related cytokines (eg, platelet-derived growth factor [PDGF], a-fibroblast growth factor [aFGF], bFGF, transforming growth factor (TGF)-, and VEGF-A isoforms, among others) by keratinocytes, resident mast cells, degranulating platelets, and other damaged stromal cells. This is followed in short order by infiltrating neutrophils, macrophages, and other inflammatory cells that are recruited to the area of damaged tissue. Leukocyte migration is also essential for removal of necrotic cellular debris, whereas many cell types may participate in the removal of apoptotic cells. Wounded keratinocytes and other cells transiently become leaky and release calcium, which may act as a transcriptional activator.²⁰ Fibrin synthesis and polymerization also help stabilize the wounded tissue and provide a scaffold into which infiltrating cells can migrate.

Cytokine signaling also plays an important role in the angiogenic response to wound repair and infection. High local levels of FGFs, released from extracellular matrix and from damaged parenchymal cells, contribute to the earliest phase of a local angiogenic response.²¹ As wound maturation progresses, additional angiogenesis is induced by increasing tissue levels of VEGFs, which peak several days after injury.²² The rate of endothelial cell proliferation in the presence of a large and virulent microbial burden is very high. In C3H mice inoculated with Mycoplasma pulmonis, for example, the diameter of arterioles, capillaries, and venules had doubled within 1 week of inoculation, with BrdUrd labeling increasing by 20-fold,²³ peaking at day 5, and then followed by a slow, yet steady, decrease.

Many cell types contribute to the source of VEGF in maturing epithelial wounds, including keratinocytes, macrophages, fibroblasts, endothelial cells, and T cells,²⁴ among others. Activation of endothelial cells in the context of wounding, which in vivo nearly always implies some degree of infection, results in initiation of coagulation cascades and impaired fibrinolysis.²⁵ Tissue factor on endothelial cells is induced by inflammatory cytokines and by pentraxins, a class of innate immunity proteins including C-reactive protein and pentraxin 3, among others, whose expression and/or activation are enhanced by pathogens and tissue injury.

Microvascular hyperpermeability induced by one or more VEGFs and the introduction of a provisional plasma-derived matrix precede the onset of endothelial cell proliferation and migration and accompany new blood vessel formation (many tumors have co-opted this mechanism that developed in multicellular organisms for purposes of tissue defense, renewal, and repair). Interestingly, small molecule autacoids that regulate the inflammatory response (eg, eicosanoids, adenosine, among others), as well as many cytokines (eg, bFGF, TGFß, TNF), which do not have an intrinsic capacity to increase vascular hyperpermeability, do so indirectly by inducing expression and release of VEGFs.

In addition, resident mast cells and infiltrating neutrophils and macrophages are the likely tissue source of numerous cytokines traditionally associated with a localized inflammatory response, such as TNF- and IL-1, among others, many of which have been associated with increased vessel permeability and angiogenesis.²⁶ Animal models of impaired wound healing, such as the leptin-deficient (db/db) diabetic mouse, are characterized by profound defects in wound-related angiogenesis.²⁷ Administration of a topical formulation of rhVEGF-A₁₆₅ also has been shown to accelerate cutaneous wound healing of diabetic (db/db) mice, in part by inducing expression of PDGF and FGF-2, and by increasing both circulating and wound-related levels of AC133⁺/VEGF-R²⁺/CD11b^– endothelial progenitor cells.²⁸

The impaired wound-healing phenotype of the db/db mouse also can be rescued by localized (ie, to the wound) transfection of the homeobox gene HoxD3, a transcription factor essential for normal embryogenesis, but is also associated with angiogenesis and extracellular matrix remodeling after development.²⁷ Another transcriptional regulator only recently associated with angiogenesis is Net, a member of the ternary complex factor known primarily as a transcriptional repressor, but which has also been shown to be a positive inducer of transcription after phosphorylation by Ras or Src. Zheng et al²⁹ have shown that mice homozygous for a mutant version of Net (Net^/) exhibit delayed and diminished wound healing, which could be largely attributed to a profound decrease in wound-related angiogenesis. VEGF-A expression in keratinocytes and mononuclear cells in the wounds of Net^/ mice was markedly reduced as well. Interestingly, the 10% of Net^/d mice that survive development (and were used for these experiments) exhibited normal development of the vascular tree but had abnormalities in lymphatic development.²⁹

Despite the central role of VEGFs in cutaneous wound repair, Jacobi et al³⁰ noted that a single intravenous injection of an adenoviral-delivered transgene, encoding a soluble form of VEGF-R² (ie, Flk-1), acting as a "VEGF-trap," reduced skin wound-related angiogenesis but did not affect the rate of wound closure in either db/db diabetic or wild-type animals (although, as expected, wound closure in diabetic animals was slower than in wild-type animals, with or without the VEGF-trap construct). This result was also observed in other in vivo wound repair models in which angiogenesis was selectively impaired.^31,32 Nevertheless, these investigators also observed that the tensile strength of the resulting scar was not affected, suggesting that other VEGF-independent mechanisms of wound repair continue to be operative in this experimental model. Also, as shown by Salvucci et al in 2004,³³ inflammatory cytokines such as TNF- and interferon- may also induce expression of TIMP-1 and other proteins that disrupt extracellular matrix-dependent tube formation, emphasizing the importance of local contextual signaling in determining the extent of wound-related angiogenesis.

Tissue macrophages are responsible for much of the regulated induction of new blood vessel growth during wound repair, fostering regression of some vessels while inducing growth and remodeling of others.³⁴ Circulating monocytes and tissue macrophages have also been shown to play a central role in the remodeling of existing resistance and small conduit vessels into larger, muscular arterioles, particularly in the setting of tissue ischemia (see review by Scholtz and Schaper³⁵ in this series of reviews). The rabbit hind limb occlusion model³⁶ has shown, for example, that although infusion of unstimulated autologous (CD14⁺) monocytes did not accelerate resumption of blood flow, infusion of allogeneic monocytes did accelerate restoration of perfusion, as did infusion with autologous monocytes after ex vivo adenoviral-mediated transfection with a granulocyte macrophage colony-stimulating factor (GM-CSF) transgene. Tissue macrophage release of matrix metalloproteinases appears also to be required for the extracellular matrix remodeling essential for effective angiogenesis to occur.³⁷ Schaper et al recently documented that CCR2-deficient mice fail to undergo adequate vascular remodeling and enlargement in response to limb ischemia.³⁸ Thus, activation of specific "proinflammatory" pathways appears to be essential in restoring blood flow even in the absence of significant tissue injury or loss.

TGFß also plays a pivotal role in the initiation and maintenance of wound repair, activating fibroblasts for the synthesis of new extracellular matrix proteins, while slowing the rate of re-epithelialization.³⁹ Wound repair modeling in Drosophila has identified a number of genes involved in this process including, perhaps as expected, Rho GTPases, JNKs, and AP-1, but also decapentaplegic (Dpp) the Drosophila homologue of BMP4, and the phosphatase puckered (puc). Signaling by puc eventually suppresses the activity of JNKs, thereby slowing the activity of the wound repair process.²⁰ TGFß-enhanced, wound repair-related angiogenesis results in initial inhibition of endothelial cell proliferation, increased production of and remodeling by extracellular matrix proteins, and the recruitment and transformation of smooth muscle cell/pericyte precursors, a process that involves SMADs 2/3 and 3/4, and members of the activin receptor-like kinases 1 and 5.⁴⁰

Another important characteristic of many experimental animal—typically cutaneous—wounds is hypoxia.⁴¹ Skin wounds are usually characterized by an interruption in local tissue perfusion or infection, or both, resulting in a characteristic decline in tissue oxygen tension content. The expression of the inducible isoform of nitric oxide synthase (iNOS), a "high-output," calcium-independent NOS isoform, is known to increase in response to hypoxia and is associated with the early innate immune response in many species, although the extent of translation of iNOS mRNA to active enzyme varies among species and on the type of infection.⁴²

Albina and Reichner⁴³ have recently demonstrated that the iNOS-mediated flux of arginine to citrulline—an essential step in the production of bacteriostatic levels of nitric oxides (NO_x)—is diminished at very low oxygen tensions because of substrate limitations, despite increases in iNOS expression and protein levels. Importantly, the increase in iNOS levels was attributable not only to increased amounts of inflammatory cytokines in wounded infected tissue but also to HIF-1–dependent increases in iNOS gene expression. Activation of HIF-1, as is described later, has now been shown to be responsive not only to the intracellular oxygen concentration but also to the activation of several innate immunity signaling pathways (see Figure 2B). With reference to wound repair, HIF-1 also activates PAI-1, which impacts fibrin deposition and wound closure.⁴⁴

View larger version (54K): [in this window] [in a new window]   Figure 2. Regulation of HIF-1 activity by inflammation and growth factor-related cytokines. The figure depicts two mechanisms that result in activation of HIF-1. In addition to its role in the regulation of many genes involved in the response to intracellular hypoxia (B), HIF-1 is also activated by a number of other intracellular signaling pathways, even in the absence of hypoxia. These include "inflammation-related" signaling pathways triggered by activation of TLRs, tumor necrosis factor (TNF), and IL-1ß, among others, as well as signaling pathways associated with survival/growth factors (eg, IGF-1, IGF-2, FGF-2, TGFß1, among others). Heat shock proteins, after activation of a PI3K/Akt pathway, have also been shown to delay degradation of HIF-1 by directly interacting with the Per-ARNT-Sim (PAS) domain.

	Antimicrobial Peptides and Angiogenesis

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
Among the best characterized of the antimicrobial peptides of the innate immune response are the defensin and cathelicidin families of proteins (see recent comprehensive review by Bulet et al⁴⁵). These germline-encoded antimicrobial peptides act as innate immunity "pattern recognition receptors" and as effectors of the innate immune response. In addition, several also appear to initiate the process of tissue repair, largely by inducing an angiogenic response localized to the site of injury.

Cathelicidins comprise one class of antimicrobial peptides, which are released by neutrophils and macrophages, as well as by some gonadal, gastrointestinal, and respiratory epithelial cells in response to evidence of infection. Their direct antimicrobial activity against viruses, bacteria, and parasites is mediated principally by disruption of microbial membranes via several mechanisms. In addition, these peptides also act as chemoattractants and activators for inflammatory cells.¹⁶

Only one human cathelicidin has been identified to date: LL-37/hCAP-18, which is closely related to the murine cathelin-related antimicrobial peptide. LL-37 also has been shown to exhibit angiogenic activity by signaling through the G-protein–coupled formyl peptide receptor-like 1.⁴⁶ Formyl peptide receptor-like 1 is present in macrophages, neutrophils, lymphocytes (primarily NK cells, B cells, and T lymphocytes), keratinocytes, as well as in epithelial and endothelial cells. It is the activation of endothelial cells by LL-37, as well as recruitment of neutrophils and macrophages, which induces the angiogenic response. LL-37/hCAP-18 has been documented to play a role in re-epithelialization of skin wounds.⁴⁷ LL-37/hCAP-18 also activates a number of cytokine pathways in monocytes and tissue macrophages, including CCL2 (formerly termed MCP-1;⁴⁸) and its receptor CCR2, among others.

Interestingly, the porcine cathelicidin PR-39, a proline-rich, 39-amino acid protein, for which no obvious homologue exists in humans, has been shown by Simons et al to induce a robust angiogenic response in in vivo and in vitro experimental models.⁴⁹ Unlike other cathelicidins, however, the mechanism of action of PR-39 is unique in that it prevents proteosomal degradation of HIF-1, even when oxygen is not limiting. This is attributable to the reversible and relatively selective effect of PR-39 on the 20S proteosome.⁵⁰ PR39 also blocks activation of NFB, again by relatively selective inhibition of proteosomal degradation of IB. This may account, in part, for the observation that transgenic animals overexpressing PR39 in the heart lack significant inflammation despite sustained high cardiac myocyte expression levels of HIF-1.⁵⁰

The antibiotic activity of each of the families of antimicrobial peptides discussed was noted long before their pro-angiogenic effects were discovered.⁵¹ Angiogenins are an exception, however, having initially been associated with induction of vasculogenesis in in vitro and in vivo assays.⁵² The angiogenins are members of a family of ribonucleases that exhibit antimicrobial activity. The high rate of sequence divergence among the murine angiogenin isoforms suggested a role in host defense, however, and recent work has verified that these innate immunity peptides possess both antibacterial and antifungal activities, in addition to promoting angiogenesis and wound repair.⁵²

	Angiogenic Cytokines and Inflammation

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
A central role for angiogenesis in sustaining an inflammatory response has only recently become evident. In the context of acute infection and wound repair, angiogenesis is largely beneficial as discussed. Terminating the angiogenic response may be more problematic, however, in part because of the fact that angiogenic cytokines may act to sustain an inflammatory response, including increased tissue edema. After the cessation of the original angiogenic stimulus, the newly formed network undergoes extensive hemodynamic remodeling, which is governed in part by other cytokines acting to stabilize or "normalize" the newly formed vessel.

One angiogenic cytokine implicated in this "vascular remodeling" paradigm is placental growth factor (PlGF). Carmeliet, Persico, et al were the first to note that PlGF, which, unlike VEGF-A, is not required for normal development, appeared to be required for postnatal angiogenesis, particularly in the context of "arteriogenesis" (ie, vascular remodeling leading to new collateral vessel formation after ischemic injury and wound healing).⁵³ In subsequent reports from these investigators,^54,55 the importance of VEGF-R1 (flt-1) in PlGF signaling, and the importance of PlGF in amplifying and sustaining an inflammatory response, became apparent. PlGF was shown to amplify the angiogenic response to VEGF-A in part by inducing VEGF production in parenchymal cells, by initiating an arteriogenic response, and by recruiting smooth muscle cell precursors to the nascent arteriolar vasculature. The amplification step was mediated in part by PlGF-dependent activation of VEGFR-1, with subsequent phosphorylation of VEGF-R² (flk-1). Moreover, the VEGF-A isoforms alone could not substitute for the loss of PlGF, and the combination of PlGF and VEGF-A resulted in a superior angiogenic response to either cytokine alone. Similar data were generated by Pipp et al,⁵⁶ who in addition provided evidence for a role for monocytes and tissue macrophages in restoring blood flow, because animals treated with PlGF that had been depleted of monocytes demonstrated no additional benefit on blood flow compared with a VEGF-A agonist alone.

As reviewed by Nagy,⁵⁷ PlGF signaling appears to be essential for the vascular remodeling necessary to form large arteriolar resistance vessels in some tissue. This is supported by experiments by Scholz et al,³⁵ which documented accelerated induction of collateral blood flow (ie, arteriogenesis) in PlGF^–/– mice reconstituted with bone marrow from PlGF^+/+ animals, compared with animals reconstituted with PlGF^–/– bone marrow.³⁵

Although our understanding of the role that angiogenic cytokines play in normal development and physiology has increased rapidly since the first description of vascular permeability factor by Senger, Dvorak, and Dvorak et al nearly 20 years ago,⁵⁸ the association of VEGFs with pathologic angiogenesis, aside from tumor vasculature, has only recently been appreciated. Ozawa et al⁵⁹ have shown that release of high levels of VEGF-A₁₆₄ from retrovirally transfected skeletal myoblast-induced hemangioma formation, whereas myoblast clones releasing lower levels of VEGF-A₁₆₄ did not, pointing to a "threshold" relationship between dose and abnormal physiologic outcomes. PlGF in particular has been associated with enhanced inflammation and edema formation, whereas PlGF deficiency (in PlGF^–/– animals) correlated with a diminished and abbreviated inflammatory response, despite similarly upregulated levels of VEGF-A isoforms in both phenotypes.⁶⁰ Also, patients with severe sickle cell anemia (as defined by 3 or more vaso-occlusive episodes per year) were found to have elevated PlGF levels compared with controls or less ill sickle cell patients, and PlGF significantly increased levels of TNF, IL-1ß, MCP-1, and IL-8, among other cytokines, in vitro.^61,62 In an allergic encephalomyelitis model that mimics multiple sclerosis, VEGF-A immunoreactivity increased in parallel with the development of the typical histologic changes characteristic of encephalomyelitis, implicating this angiogenic cytokine in facilitating the development of long-term neural inflammation and eventually demyelination.⁶³ VEGF-A infusions directly into the neocortex of normal rats increased vascular density progressively over a 7-day period and were accompanied by an early (<24 hours) and marked dose-dependent extravasation of IgG and leukocytes.⁶⁴ Staining for ED-1 antigen indicated that most of the inflammatory cells were of monocyte/macrophage origin. The marked inflammatory response to VEGF-A was not apparent with infusion of angiopoietin-1, in which no effect on leukocyte extravasation was observed.

Even in the absence of exogenously administered PlGF or other inflammatory cytokines, cutaneous overexpression of VEGF-A₁₆₄ driven by a keratin-14 (K14) promoter in mice resulted in lesions very similar to those observed in psoriatic humans.⁶⁵ The lesions exhibited not only the hypervascularity and abnormal epidermal proliferation typical of psoriatic skin in humans but also epidermal microabscesses and inflammatory infiltrates characteristic of the human disease, all of which could be reversed using another variation of a "VEGF-trap," consisting of the Ig1 domain of VEGF-R1, the Ig3 domain of VEGF-R², and the Fc domain of human IgG1.⁶⁶ Interestingly, the characteristics of the synovial inflammation associated with psoriatic arthritis also may be unique to this disease. VEGF-A and TGFß1 levels were higher in psoriatic synovial fluid than levels obtained from aspirations of rheumatoid arthritic joints, and biopsy results showed synovial vessels that were significantly larger and more tortuous than those obtained from rheumatic joints.⁶⁷ Patients who received intravenous infusions of infliximab, a monoclonal anti-TNF antibody, exhibited reductions in VEGF-A, VEGFR-2, CD-31, and several indices of neovascularization on post-treatment synovial biopsy results.⁶⁸

Pro-angiogenic cytokines other than VEGF-A are also likely to play a role in the pathophysiology of arthritis, because mice with targeted disruption of VEGF-B exhibited significantly less joint swelling and inflammation than wild-type animals.⁶⁹ Also, in a model of rheumatoid arthritis induced by collagen, the angiopoietin receptor Tie2 and its ligand, angiopoietin-1 increased rapidly in affected joints, an increase that was associated temporally with an increase in synovial fluid TNF- levels.⁷⁰ In patients with osteoarthritis, measures of endothelial cell proliferation (ie, endothelial cell fractional area and proliferative index on histologic analysis) all have been shown to increase in proportion to the extent of macrophage infiltration.⁷¹ Members of the ETS family of transcription factors, which among other functions regulate the expression of several inflammatory cytokines, also induce expression of a number of angiogenic genes, including angiopoietin-1, among others.⁷² In addition, a dominant-negative ETS-1 inhibited neuropilin-1 and angiopoeitin-2 expression in in vitro assays and suppressed development of retinal angiogenesis in a murine model of proliferative retinopathy.⁷³

In a model of retinal neovascularization induced by prolonged exposure of neonatal rats to 80% oxygen to induce an avascular retina, followed by a return to room air to induce retinal blood vessel growth, there was a hypoxia-mediated, and relatively isoform-selective, 2- to 3-fold increase in VEGF-A₁₆₄ levels.⁷⁴ Unlike physiologic neovascularization, pathophysiologic retinal neovascularization was characterized by the presence of adherent leukocytes and the development of leukocytic fronds, which, within 1 week, had begun to extend into the vitreous. A VEGFR-1 chimeric protein was used to selectively inhibit signaling through the VEGF-A₁₆₄ isoform, which diminished both leukocyte (largely monocytic) adherence and suppressed the development of retinopathy. Conversely, forced expression of IL-1ß selectively in the retina (using a A-crystalline promoter) resulted in an inflammatory infiltrate with marked increases in retinal VEGF levels.⁷⁵

The ultimate extent of the pro-inflammatory activity of VEGFs will depend obviously on the physiologic and cellular context, as noted earlier; for example, VEGF also upregulates heme oxygenase-1 in human umbilical vein endothelial cells and other endothelial cell phenotypes. Activation of heme oxygenase-1 markedly increased endothelial cell proliferation while decreasing leukocytic infiltration in an LPS-induced angiogenesis model. Conversely, suppression of heme oxygenase-1 activity with metalloporphyrins resulted in a decreased angiogenic response despite an increase in inflammatory infiltrates.⁷⁶

In contrast to VEGFs and PlGF, angiopoietin signaling via Tie2 often suppresses NFB activation, decreasing leukocytic cell migration and endothelial cell permeability, and, particularly in the case of Tie2 activation, reducing intercellular adhesion molecule and vascular cell adhesion molecule expression.⁷⁷ Using a yeast two-hybrid screen, Hughes et al⁷⁷ identified ABIN-2 (A20 binding inhibitor of NFB activation-2) as a required signaling intermediate for suppression of NFB in endothelial cells. Interestingly, platelet endothelial cell adhesion molecule knockout mice developed normally but had a marked deficit in the accumulation of inflammatory cells—largely lacking neutrophilic infiltrates—in a foreign body accumulation assay.⁷⁸

Unconventional angiogenesis-promoting autacoids, such as angiotensins, which have long been implicated in both physiologic and pathophysiologic angiogenesis, stimulate endothelial cell proliferation and migration by effects that are both indirect (eg, via FGF-2 release) and direct (eg, by inducing changes in matrix metalloproteinase activities, resulting in alterations in extracellular matrix composition).⁷⁹ In an in vivo subcutaneous Matrigel plug assay, angiotensin II induced not only endothelial nitric oxide synthase and VEGF-A expression, with recruitment of endothelial and smooth muscle cells, but also macrophage infiltration.⁸⁰ These in vitro observations were supported by experiments by Sasaki et al,⁷⁹ who documented that the infiltration of mononuclear cell infiltrates (principally macrophages and T cells) into the ischemic hind limb of wild-type mice was largely absent in angiotensin receptor AT-1a^–/–-deficient animals.

	HIF-1, Angiogenesis, and Innate Immunity

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 
The central role of HIF-1 in the regulation of hypoxia-driven angiogenesis has been extensively described elsewhere.^81,82 The discovery of the HIF-1 prolyl and asparaginyl hydroxylases provided a link between intracellular O₂ levels and the rate of HIF-1 degradation and transactivation, respectively (Figure 2B). The arylhydrocarbon receptor nuclear translocator (ARNT or HIF-1ß) is known to regulate the expression of a number genes, the products of which are involved in angiogenesis, including VEGF, PDGF, PlGF, Flt-1, adrenomedullin, angiopoeitin-2, and angiopoeitin-4, among others,^83,84 and also directly suppresses transcription of prolylhydroxylases when oxygen is limiting.⁸⁵ A decrease in pH appears to sustain HIF signaling even as the intracellular O₂ content increase, because of acidosis-induced sequestration of the von Hippel-Lindau protein.⁸⁶ Genomic analyses of selected cell types transduced with constitutively active HIF-1 transgenes have also identified unconventional pro-angiogenic peptides that are regulated by HIF-1, such as adrenomedullin and the natriuretic peptides ANP and BNP, for example.⁸⁷ HIF-1 activation, as in many other cell types, also regulates the expression of a number of proteins involved in apoptosis.⁸⁸ In endothelial cells (eg, human umbilical vein endothelial cells), HIF-1 decreased apoptosis in response to anoxia reperfusion injury, a response that depends in large part on the cellular context (eg, intracellular levels of metabolic substrate needed for respiration⁸⁹). Laderoute et al⁹⁰ have recently shown that activation of intracellular extracellular signal–regulated kinases, c-Jun amino-terminal kinases (JNKs), and stress-activated protein kinases, and phosphorylation of c-Jun, require HIF-dependent triggering of glucose utilization.

In addition to these, by now, canonical pathways for the regulation of HIF activity, there are multiple additional hypoxia-independent signaling pathways that impact HIF-1 activity beyond protein stability and transactivation as discussed. Genomic analyses of selected cell types transfected with constitutively active HIF-1 transgenes have also identified unconventional pro-angiogenic peptides that are regulated by HIF-1, such as the natriuretic peptides ANP and BNP, for example.^84,87,91 In recent comprehensive reviews by Bilton and Booker⁹² and by Giacci et al,⁹³ these signaling pathways were noted to include survival/growth factors (eg, insulin-like growth factor-1 [IGF-1], IGF-2, FGF-2, TGFß1, insulin, epidermal growth factor), cytokines such as TNF and IL-1ß, as well as oncogenes such as HER2^NEU, among others. LPS alone, even in the absence of inflammatory cytokines, has also been shown to directly activate HIF-1 signaling to a greater extent than during hypoxia, at least in vitro in primary bone marrow-derived cells.⁹⁴ Receptor-mediated activation of HIF-1 is largely independent of the cellular O₂ content, but both mechanisms can act cooperatively to enhance HIF-1 protein stability and, consequently, activity.^95–97 The effects of specific signaling pathways are often cell-type selective as well because, for example, both TNF and IL-1ß were shown in HepG2 cells to increase DNA binding of HIF-1, yet only IL-1ß increased HIF-1 protein levels.⁹⁸

Activation of specific HIF- isoforms can be selectively regulated, because phorbol esters, as well as heat shock and hydrogen peroxide (and presumably other oxygen radicals) also can induce a HIF-1 splice variant lacking exon 11 (HIF-1⁷⁸⁵), the fifth human HIF-1 splice variant to be described, which also retains transcriptional regulatory activity.⁹⁹ Several HSPs (specifically HSP70 and HSP90), after activation of a PI3K/Akt pathway, have been shown to delay degradation of HIF-1 by directly interacting with the Per-ARNT-Sim domain, thus competing with HIF-1ß/ARNT for binding at this site on HIF-1.^100,101 HSP binding to HIF-1ß/ARNT acts as a chaperone, maintaining HIF-1 in a receptive state for signaling. PKC activation also has been shown to prolong HIF activity indirectly by inhibiting the transcription of the asparaginyl hydroxylase (factor-inhibiting HIF), which suppresses HIF association with p300 and synthesis of HIF-dependent genes.¹⁰²

NFB activation also has been shown to be required for both TNF-–mediated and IL-1–mediated activation of HIF-1 (see Figure 2B). Zhou et al¹⁰³ demonstrated that activation of NFB induced accumulation of both unmodified and ubiquitinated HIF-1, but no change in HIF-1 mRNA levels, implying that regulation of HIF expression is also mediated by a translational mechanism, and not just by attenuating proteosomal degradation. A similar observation was made by Jung et al,¹⁰⁴ who noted also that receptor-interacting protein-1 appears to be required for TNF-dependent induction of HIF-1. In a separate report, Jung et al¹⁰⁵ documented that IL-1ß also increases HIF-1 activation via NFB, an effect dependent in part on increased expression of cyclooxygenase 2. These data may explain the observation that IL-1ß–increased cyclooxygenase 2 activity also results in an increase in eicosanoid-dependent angiogenesis.¹⁰⁶

Despite the observation that several "inflammatory" cytokines could activate HIF-1–dependent pathways in a number of cell types, the link between HIF-1 activation and induction of innate immunity was not obvious. In contrast with earlier reports in which transgenic animals with enhanced VEGF-A expression exhibited anatomical and physiologic abnormalities of the microvasculature resembling, at least in part, those characteristics of inflammation,¹⁰⁷ Elson et al¹⁰⁸ found that in transgenic mice engineered to overexpress a constitutively active form of HIF-1 in the skin (ie, wild-type HIF-1 driven by a keratin 14 promoter), they did not observe evidence of increased edema or inflammation despite increases in both the number and size of vessels within the dermis.

However, Ceradini et al¹⁰⁹ have shown that increased expression of SDF-1 (CXCL12) is dependent on activation of HIF-1 signaling in endothelial cells, resulting in recruitment and engraftment of CXCR4-positive endothelial progenitor cells, suggesting that HIF-1 activation is involved in the regulation of specific cytokines associated with induction of inflammation. For a recent comprehensive review of the role of chemokine signaling in angiogenesis and wound healing, see Romagnani et al.¹¹⁰

A central role for HIF-1 in sustaining, if not initiating, inflammation was revealed in the report by Cramer et al,¹¹¹ who selectively deleted HIF in leukocytes by expressing cre recombinase under the control of the lysosyme M promoter (lysM-cre/HIF-1). This did not result in large changes in absolute or relative numbers of circulating subsets of white cells in the absence of an inflammatory stimulus. However, these cells selectively exhibited reductions in intracellular VEGF-A mRNA and protein, as well as genes involved in gluconeogenesis and glucose transport, consistent with the absence of HIF-1 activity.

As a consequence, peritoneal macrophages and neutrophils from these myeloid lineage-specific, HIF-1–depleted animals had marked reductions of ATP levels, resulting in diminished leukocyte aggregation and motility. Moreover, the phenotype of myeloid lineage-specific HIF-1–null animals was distinct from lysM-Cre/VEGF-A knockout mice, in that the former exhibited almost no inflammatory infiltrate and edema after a phorbol ester-induced inflammatory stimulus, whereas the latter had an extensive inflammatory infiltrate but only minimal edema formation.¹¹¹

To determine the phenotype of animals with leukocyte-specific HIF-1 expression unrestrained by prolyl or asparaginyl hydroxylation, and subsequent proteosomal degradation, Cramer et al¹¹¹ also studied myeloid-specific, von Hippel-Lindau–deficient (VHL-LysM-CRE) mice, which lack the ability to downregulate HIF-1 activity. These animals exhibited an exaggerated response to phorbol esters, with inflammation that was significantly greater than in wild-type controls, as determined by myeloperoxidase activity. These in vivo data were supported by in vitro observations of Moon et al¹¹² that desferrioxamine markedly increased not only HIF-1 protein levels in human mast cells but also inflammatory cytokines such as IL-6, IL-8, and TNF-. Nathan,¹¹³ in an editorial comment on the report by Cramer et al,¹¹¹ noted that mast cells, which act as sentinels for the innate immune response in many tissues, also express lysosyme, and suggested that the phenotype of the lysM-cre/HIF-1 animals might have also been attributable to a defect in mast cell function.

Cramer et al also determined that wild-type animals demonstrated diminished limb function and delayed resolution of inflammation in a noninfectious passive model of arthritis when compared with lysM-cre/HIF-1 mice.¹¹¹ The relevance of these data for human disease is supported by Giatromanolaki et al,¹¹⁴ who performed immunohistochemical analyses of biopsy specimens of synovial tissue of patients with rheumatoid and osteoarthritic joint disease of the hip, knee, and/or wrist, using as controls specimens obtained from patients with no history of arthritis undergoing hip fracture surgery. Patients with active arthritis exhibited increased levels of HIF-1 and HIF-2, as well as increased VEGF-A, platelet-derived endothelial cell growth factor, and VEGF complexed with Flk-1. The percentage of cells that stained positively for HIF-1 and HIF-2 was significantly higher in both rheumatoid and osteoarthritic synovium compared with the hip fracture-only controls, and upregulation of both HIF isoforms was observed in blood vessels, macrophages, fibroblasts, and lymphoid cells. Interestingly, mechanical trauma alone has been shown to induce HIF-1 expression. Well-oxygenated bovine cartilage disks exposed to repetitive mechanical strain rapidly increased expression of HIF-1 and subsequently VEGF-A, whereas disks not exposed to impact stress did not.¹¹⁵

In addition to endothelial nitric oxide synthase, HIF-1 is also known to regulate the expression of iNOS, an important effector arm of innate immunity.¹¹⁶ Moncada et al and Brune et al have recently highlighted the complexity of the molecular pharmacology of NO during hypoxic stress and its role in regulating both oxidative phosphorylation and HIF-1 activity.¹¹⁷ Within cells, as NO levels increase to the high levels associated with iNOS induction (sufficient, for example, to achieve bacteriostatic levels), both groups document that the activities of the HIF-1 prolylhydroxylases decrease, thus permitting sustained HIF signaling, even after intracellular O₂ levels begin to increase. This would provide an obvious selective advantage to cells undergoing both hypoxic and infectious challenges. In addition, pharmacological doses of NO delivered by NO donors (and presumably high levels of endogenous NO_x after iNOS induction) also increased HIF-1 expression via a MAPK/PI3K pathway.¹¹⁸

In conclusion, there is new evidence for multiple links between pathogen recognition and induction of an initial innate immune response, with multiple "growth factor" signaling pathways associated with angiogenesis.

	Acknowledgments

 
This work was supported by the Deutsche Forschungsgemeinschaft (Fr 1377/4-3) (to S.F.). O.F. is a FNRS Research Associate.

	Footnotes

 
This manuscript was sent to Hans Donald Heistad, Consulting Editor, for review by expert referees, editorial decision, and final disposition.

Original received May 17, 2004; resubmission received October 21, 2004; revised resubmission received December 2, 2004; accepted December 2, 2004.

	References

Top Abstract Introduction Wound Repair and Angiogenesis Antimicrobial Peptides and... Angiogenic Cytokines and... HIF-1{alpha}, Angiogenesis, and... References

 

1. Takeda K, Akira S. TLR signaling pathways. Semin Immunol. 2004; 16: 3–9.[CrossRef][Medline] [Order article via Infotrieve]
2. Montesinos MC, Desai A, Chen JF, Yee H, Schwarzschild MA, Fink JS, Cronstein BN. Adenosine promotes wound healing and mediates angiogenesis in response to tissue injury via occupancy of A(2A) receptors. Am J Pathol. 2002; 160: 2009–2018.[Abstract/Free Full Text]
3. Pinhal-Enfield G, Ramanathan M, Hasko G, Vogel SN, Salzman AL, Boons GJ, Leibovich SJ. An angiogenic switch in macrophages involving synergy between Toll-like receptors 2, 4, 7, and 9 and adenosine A(2A) receptors. Am J Pathol. 2003; 163: 711–721.[Abstract/Free Full Text]
4. Montesinos MC, Shaw JP, Yee H, Shamamian P, Cronstein BN. Adenosine A(2A) receptor activation promotes wound neovascularization by stimulating angiogenesis and vasculogenesis. Am J Pathol. 2004; 164: 1887–1892.[Abstract/Free Full Text]
5. Janeway CA, Jr. Approaching the asymptote?Evolution and revolution in immunology. Cold Spring Harb Symp Quant Biol. 1989; 54: 1–13.
6. Yokoyama WM, Kim S, French AR. The dynamic life of natural killer cells. Annu Rev Immunol. 2004; 22: 405–429.[CrossRef][Medline] [Order article via Infotrieve]
7. Hoffmann JA. The immune response of Drosophila. Nature. 2003; 426: 33–38.[CrossRef][Medline] [Order article via Infotrieve]
8. Beutler B. Inferences, questions and possibilities in Toll-like receptor signalling. Nature. 2004; 430: 257–263.[CrossRef][Medline] [Order article via Infotrieve]
9. Loker ES, Adema CM, Zhang SM, Kepler TB. Invertebrate immune systems–not homogeneous, not simple, not well understood. Immunol Rev. 2004; 198: 10–24.[CrossRef][Medline] [Order article via Infotrieve]
10. Matzinger P. The danger model: a renewed sense of self. Science. 2002; 296: 301–305.[Abstract/Free Full Text]
11. Heath WR, Carbone FR. Immunology: dangerous liaisons. Nature. 2003; 425: 460–461.[CrossRef][Medline] [Order article via Infotrieve]
12. Ledoux S, Runembert I, Koumanov K, Michel JB, Trugnan G, Friedlander G. Hypoxia enhances Ecto-5'-Nucleotidase activity and cell surface expression in endothelial cells: role of membrane lipids. Circ Res. 2003; 92: 848–855.[Abstract/Free Full Text]
13. Shi Y, Evans JE, Rock KL. Molecular identification of a danger signal that alerts the immune system to dying cells. Nature. 2003; 425: 516–521.[CrossRef][Medline] [Order article via Infotrieve]
14. Prohaszka Z, Fust G. Immunological aspects of heat-shock proteins-the optimum stress of life. Mol Immunol. 2004; 41: 29–44.[CrossRef][Medline] [Order article via Infotrieve]
15. Reis e Sousa C. Toll-like receptors and dendritic cells: for whom the bug tolls. Semin Immunol. 2004; 16: 27–34.[CrossRef][Medline] [Order article via Infotrieve]
16. Yuan Q, Walker WA. Innate immunity of the gut: mucosal defense in health and disease. J Pediatr Gastroenterol Nutr. 2004; 38: 463–473.[CrossRef][Medline] [Order article via Infotrieve]
17. Rakoff-Nahoum S, Paglino J, Eslami-Varzaneh F, Edberg S, Medzhitov R. Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell. 2004; 118: 229–241.[CrossRef][Medline] [Order article via Infotrieve]
18. Malagoli D, Conte A, Ottaviani E. Protein kinases mediate nitric oxide-induced apoptosis in the insect cell line IPLB-LdFB. Cell Mol Life Sci. 2002; 59: 894–901.[CrossRef][Medline] [Order article via Infotrieve]
19. Svendsen MN, Werther K, Nielsen HJ, Kristjansen PE. VEGF and tumour angiogenesis. Impact of surgery, wound healing, inflammation and blood transfusion. Scand J Gastroenterol. 2002; 37: 373–379.[CrossRef][Medline] [Order article via Infotrieve]
20. Jacinto A, Martinez-Arias A, Martin P. Mechanisms of epithelial fusion and repair. Nat Cell Biol. 2001; 3: E117–E123.[CrossRef][Medline] [Order article via Infotrieve]
21. Risau W. Mechanisms of angiogenesis. Nature. 1997; 386: 671–674.[CrossRef][Medline] [Order article via Infotrieve]
22. Karayiannakis AJ, Zbar A, Polychronidis A, Simopoulos C. Serum and drainage fluid vascular endothelial growth factor levels in early surgical wounds. Eur Surg Res. 2003; 35: 492–496.[CrossRef][Medline] [Order article via Infotrieve]
23. Ezaki T, Baluk P, Thurston G, La Barbara A, Woo C, McDonald DM. Time course of endothelial cell proliferation and microvascular remodeling in chronic inflammation. Am J Pathol. 2001; 158: 2043–2055.[Abstract/Free Full Text]
24. Mor F, Quintana FJ, Cohen IR. Angiogenesis-inflammation cross-talk: vascular endothelial growth factor is secreted by activated T cells and induces Th1 polarization. J Immunol. 2004; 172: 4618–4623.[Abstract/Free Full Text]
25. Keller TT, Mairuhu AT, de Kruif MD, Klein SK, Gerdes VE, ten Cate H, Brandjes DP, Levi M, van Gorp EC. Infections and endothelial cells. Cardiovasc Res. 2003; 60: 40–48.[Abstract/Free Full Text]
26. Chen JX, Chen Y, DeBusk L, Lin W, Lin PC. Dual functional roles of Tie-2/angiopoietin in TNF-alpha-mediated angiogenesis. Am J Physiol Heart Circ Physiol. 2004; 287: H187–H195.[Abstract/Free Full Text]
27. Hansen SL, Myers CA, Charboneau A, Young DM, Boudreau N. HoxD3 accelerates wound healing in diabetic mice. Am J Pathol. 2003; 163: 2421–2431.[Abstract/Free Full Text]
28. Galiano RD, Tepper OM, Pelo CR, Bhatt KA, Callaghan M, Bastidas N, Bunting S, Steinmetz HG, Gurtner GC. Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow-derived cells. Am J Pathol. 2004; 164: 1935–1947.[Abstract/Free Full Text]
29. Zheng H, Wasylyk C, Ayadi A, Abecassis J, Schalken JA, Rogatsch H, Wernert N, Maira S-M, Multon M-C, Wasylyk B. The transcription factor Net regulates the angiogenic switch. Genes Devel. 2003; 17: 2283–2297.[Abstract/Free Full Text]
30. Jacobi J, Tam BY, Sundram U, Degenfeld Gv G, Blau HM, Kuo CJ, Cooke JP. Discordant effects of a soluble VEGF receptor on wound healing and angiogenesis. Gene Ther. 2004; 11: 302–309.[CrossRef][Medline] [Order article via Infotrieve]
31. Tsou R, Fathke C, Wilson L, Wallace K, Gibran N, Isik F. Retroviral delivery of dominant-negative vascular endothelial growth factor receptor type 2 to murine wounds inhibits wound angiogenesis. Wound Repair Regen. 2002; 10: 222–229.[CrossRef][Medline] [Order article via Infotrieve]
32. Roman CD, Choy H, Nanney L, Riordan C, Parman K, Johnson D, Beauchamp RD. Vascular endothelial growth factor-mediated angiogenesis inhibition and postoperative wound healing in rats. J Surg Res. 2002; 105: 43–47.[CrossRef][Medline] [Order article via Infotrieve]
33. Salvucci O, Basik M, Yao L, Bianchi R, Tosato G. Evidence for the involvement of SDF-1 and CXCR4 in the disruption of endothelial cell-branching morphogenesis and angiogenesis by TNF-alpha and IFN-gamma. J Leukoc Biol. 2004; 76: 217–226.[Abstract/Free Full Text]
34. Lingen MW. Role of leukocytes and endothelial cells in the development of angiogenesis in inflammation and wound healing. Arch Pathol Lab Med. 2001; 125: 67–71.[Medline] [Order article via Infotrieve]
35. Scholz D, Elsaesser H, Sauer A, Friedrich C, Luttun A, Carmeliet P, Schaper W. Bone marrow transplantation abolishes inhibition of arteriogenesis in placenta growth factor (PlGF)–/– mice. J Mol Cell Cardiol. 2003; 35: 177–184.[CrossRef][Medline] [Order article via Infotrieve]
36. Herold J, Pipp F, Fernandez B, Xing Z, Heil M, Tillmanns H, Braun-Dullaeus RC. Transplantation of monocytes: a novel strategy for in vivo augmentation of collateral vessel growth. Hum Gene Ther. 2004; 15: 1–12.[CrossRef][Medline] [Order article via Infotrieve]
37. Johnson C, Sung HJ, Lessner SM, Fini ME, Galis ZS. Matrix metalloproteinase-9 is required for adequate angiogenic revascularization of ischemic tissues: potential role in capillary branching. Circ Res. 2004; 94: 262–268.[Abstract/Free Full Text]
38. Heil M, Ziegelhoeffer T, Wagner S, Fernandez B, Helisch A, Martin S, Tribulova S, Kuziel WA, Bachmann G, Schaper W. Collateral artery growth (arteriogenesis) after experimental arterial occlusion is impaired in mice lacking CC-chemokine receptor-2. Circ Res. 2004; 94: 671–677.[Abstract/Free Full Text]
39. Goumans MJ, Lebrin F, Valdimarsdottir G. Controlling the angiogenic switch: a balance between two distinct TGF-b receptor signaling pathways. Trends Cardiovasc Med. 2003; 13: 301–307.[CrossRef][Medline] [Order article via Infotrieve]
40. Sanchez-Elsner T, Ramirez JR, Sanz-Rodriguez F, Varela E, Bernabeu C, Botella LM. A cross-talk between hypoxia and TGF-beta orchestrates erythropoietin gene regulation through SP1 and Smads. J Mol Biol. 2004; 336: 9–24.[CrossRef][Medline] [Order article via Infotrieve]
41. Albina JE, Reichner JS. Oxygen and the regulation of gene expression in wounds. Wound Repair Regen. 2003; 11: 445–451.[CrossRef][Medline] [Order article via Infotrieve]
42. Mahoney E, Reichner J, Bostom LR, Mastrofrancesco B, Henry W, Albina J. Bacterial colonization and the expression of inducible nitric oxide synthase in murine wounds. Am J Pathol. 2002; 161: 2143–2152.[Abstract/Free Full Text]
43. Albina JE, Reichner JS. Detection of reactive oxygen intermediate production by macrophages. Methods Mol Med. 2003; 78: 369–376.[