Book Chapter (2) - 2007 - Lecht Et Al

Transworld Research Network 37/661 (2), Fort P.O., Trivandrum-695 023, Kerala, India

Angiogenesis: Basic Science and Clinical Applications, 2007: 99-113 ISBN: 978-81-7895-302-1 Editors: M. E. Maragoudakis and E. Papadimitriou

6

Nerve growth factor - a neurotrophin with angiogenic activity Shimon Lecht1, Ilaria Puxeddu1, Francesca Levi-Schaffer1, Reuven Reich1 Ben Davidson2, Erik Schaefer3, Cezary Marcinkiewicz4, Peter I. Lelkes5 and Philip Lazarovici1 1 Dept. of Pharmacology and Experimental Therapeutics, School of Pharmacy, Faculty of Medicine, Hebrew University, Jerusalem, Israel; 2Dept. of Pathology, The Norwegian Radium Hospital, Montebello N-0310, Oslo, Norway; 3Signal Transduction, Invitrogen Hopkinton, MA; 4Dept. of Neuroscience, Temple University, 5School of Biomedical Engineering, Science and Health Systems, Drexel University, Philadelphia, PA, USA

Abstract Nerve growth factor (NGF) is a neurotrophin promoting survival, proliferation, differentiation and neuroprotective effects in the embryonal and adult nervous system. In blood vessels of serous ovarian carcinoma tumors, co-expression of NGF and its receptors with bFGF and with VEGF was found. This finding motivated us to elucidate the possibility that NGF is a neurotrophin with angiogenic activity. Correspondence/Reprint request: Prof. Philip Lazarovici, Department of Pharmacology and Experimental Therapeutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem 91120, Israel E-mail: [email protected]

100

Shimon Lecht et al.

The high-affinity NGF receptor, TrkA, was detected by RT-PCR, western blotting, and immunofluorescence in numerous cultured endothelial cell, including primary human umbilical cord and dermal microvascular endothelial cells, as well as in clones of mouse brain (bEnd.3), rat heart (RHE-A) and adrenal medulla (PC12EN) endothelial cells. NGF (10-100 ng/ml) induced proliferation of these endothelial cells in vitro as measured by thymidine incorporation, as well as MTT and Alamar blue proliferation assays. NGF stimulated TrkA phosphorylation resulting in transient Erk phosphorylation, both measured with phospho-antibodies. The NGF stimulatory effect was inhibited by K252a, a selective antagonist of TrkA receptor. To clarify the effects of NGF on the migration of cultured endothelial cells, we optimized an omnidirectional migration assay, validated with human recombinant bFGF and VEGF, using human aortic endothelial cells (HAEC). The potency of NGF to stimulate HAEC migration was similar to that of VEGF and bFGF (EC50=0.5 ng/ml). NGF-mediated stimulation of HAEC migration was completely blocked by K252a, but not by the VEGF/Flk receptor antagonist SU-5416, indicating a direct effect of NGF on HAEC migration via TrkA receptor activation. The angiogenic effects of NGF were also quantitatively measured in the ex vivo rat aorta ring explant and quail embryonic chorioallantoic membrane (CAM) models. The angiogenic effect was obtained at concentrations of 200-300 ng/ml (aorta ring sprouting) and 0.5-5 µg/ml (quail CAM) NGF. In the quail CAM assay, the NGF effect was selectively blocked by 1 µM K252a, but not by SU-5416 (1 µM) indicating a direct, selective angiogenic effect of NGF via trkA receptor activation. These results cumulatively support the hypothesis that NGF, besides its well known neurotrophic effects in the nervous system, plays a direct, TrkA-mediated angiogenic effect in the cardiovascular system. The angiogenic properties of NGF may be beneficial in engineering new blood vessels and for developing novel anti-angiogenesis therapies for cancer.

Introduction Vascularization is a fundamental survival response that helps in the healing of damaged tissue. Postnatal neovascularization involves capillary sprouting of endothelial cells via a process termed angiogenesis, proliferation of arteriolar connections by a process called arteriogenesis and vasculogenesis, a process involving the proliferation of endothelial progenitors cells in situ resulting in capillary formation [1]. Angiogenesis, is a complex, tightly regulated process involving multiple cells, growth factors and extracellular matrix proteins. It plays a critical role in the morphogenesis of many normal tissues during embryonic life. The nervous system plays a fundamental role in the maturation of the embryonic cardiovascular system as evident from findings that mutations disrupting peripheral sensory nerves prevented normal arteriogenesis, while those that disorganize the nerves maintain the alignment of arteries with misrouted axons [2]. During embryonal development peripheral nerves provide a paracrine template that determines blood vessel angiogenesis [3]. Over the last decade numerous studies indicate that blood vessels interact with neurons via neurotrophic (survival), neurotropic (neurite outgrowth, differentiation) and neurogenic (proliferation and migration of progenitors) effects.

Angiogenic effects of NGF

101

Therefore, we postulate the existence of a physiological crosstalk between the cardiovascular and nervous systems [4]. Recognizing the importance of this crosstalk for proper neovascularization and innervation of the damaged tissue, we hypothesized that upon injury the concerted action of angiogenic factors such as vascular endothelial growth factors (VEGFs) and neurotrophins such as nerve growth factor (NGF) might contribute to the repair of both nerves and blood vessels. The angiogenic effects of neurotrophins represent a novel emerging concept in pharmacology. Therefore this review will mainly focus on recent evidence that NGF, the prototype neurotrophin responsible for the maintenance of the sympathetic and sensory nervous systems and most often associated with the promotion of neuronal survival and differentiation [5] also stimulates angiogenesis.

NGF: Properties, receptors, signaling and functions NGF is an evolutionary conserved, polypeptide neurotrophin, which plays a crucial role in the sympathetic and sensory nervous systems [6]. A large number of NGFs have been discovered from different species, characterized by a known consensus sequence [7] which differs in 40 out of 118 amino acids of their β-chain, the chain that is solely responsible for NGF’s neurotropic and neurotrophic activities [8]. In mice, NGF is produced by the submaxillary gland as a precursor complex of about 130 kD (also named 7S). This complex is composed of three subunits: α, β, and γ. Recent studies indicate that peripheral tissues synthesize their own NGF precursor [9]. The chemical structure of these NGF isoforms is not yet known. The presence of multiple forms of NGFs in peripheral tissues may suggest also the occurrence of multiple receptor subtypes, an issue which has not yet been investigated. NGF, crucial for the survival and maintenance of sensory and sympathetic neurons, also affects septal cholinergic neurons in the brain [10] and neurosecretory chromaffin cells in the adrenal medulla [11]. The neurological losses observed in knockout mice lacking either NGF or its TrkA receptor provided strong evidence in support of the concept that NGF is responsible for the survival of nociceptive and thermoceptive sensory neurons as well as ganglionic and certain brain neurons [12]. The critical role of NGF in neuronal development has been extensively characterized. Recent findings point to an unexpected diversity of biological actions of NGF throughout adulthood and aging, in addition to its role during embryonic development [13]. Over the last decade, a growing body of evidence suggests that NGF signaling provides neuroprotective and repair functions in physiological and pathological conditions of the nervous system [13]. Recently, a series of studies have demonstrated that NGF exerts a variety of non neuronal effects. For example, NGF is an important modulator and mediator of the immune response [14]. In addition, NGF plays a role in inflammation and in the pathophysiology of tissue remodeling during allergic disorders [15], may enhance the healing process of wounded skin [16] and be beneficial in healing of injured corneas [17]. The biological functions of NGF are mediated through two classes of cell surface receptors: the tropomyosin kinase related receptor (TrkA), belonging to the tyrosine kinase-neurotrophin receptor family, and the p75 neurotrophin receptor (p75 NTR), common to all members of the neurotrophin family [18]. p75 NTR is a member of the super-family of tumor necrosis factor/cell death receptors (comprehensively reviewed by Miller and Kaplan [19]).

102

Shimon Lecht et al.

TrkA was originally discovered as a colon carcinoma oncogene, resulting from the gene fusion of the genes for tropomyosin and TrkA , which yields a constitutively active receptor, composed of the extracellular tropomyosin domain and intracellular TrkA tyrosine kinase domain [20]. Later, the TrkA proto-oncogene was identified as the major, biologically relevant, receptor of NGF [21, 22]. The TrkA receptor was cloned from a human erythroleukemic cell line, K562 [23], and from the rat pheochromocytoma cell line, known as PC12 [24]. The latter represents the major cellular model to investigate NGF’s mechanism of action [25]. The trkA gene encodes a membrane- bound receptor molecule comprised of extracellular, transmembrane and intracellular domains. The extracellular domain contains two cysteine-rich regions, interrupted by a leucine-rich domain, and two immunoglobulin-like regions [26] involved in the generation of the NGF binding domain. Upon NGF binding to the extracellular domain, the intracellular tyrosine kinase region is autophosphorylated in the wake of TrkA receptor occupancy and dimerization [27]. TrkA receptor-mediated signal transduction involves Ras-MAPK, phosphatidyl inositol 3 kinase (PI3K)/Akt, phospholipase Cγ, and other signaling molecules. Coordinated activation of all these signaling pathways constitutes the complex phosphorylation cascade induced by NGF and is cumulatively responsible for the majority of NGF’s “neuronal effects” [13].

Expression of NGF and TRKA activity in endothelial cells Nerve growth factor is a protein produced in the peripheral tissues to allow proper sympathetic innervation by the autonomic nervous system during embryonal development as well as for reinnervation upon tissue injury. It is well known that arteries synthesize nanogram amounts of NGF [28]. Elevated levels of NGF have been reported in the arteries of spontaneously hypertensive rats [29]. In blood vessels, NGF was believed to be mainly produced by the smooth muscle cells, which increase its expression upon injury [30]. During the last years, however, it has become evident that also blood vessels endothelial cells synthesize NGF and release it upon injury. For example, dermal endothelial cells produce NGF for nerve fiber maintenance and regeneration [31], human umbilical vein endothelial cells express NGF which is upregulated upon starvation [32], mouse aortic endothelial cells produce bioactive NGF [33], brain capillary endothelial cells secrete NGF after inflammation [34, 35], human placental blood vessels [36] and periodontal [37] endothelial cells express NGF implicated in sympathetic neurons regeneration. Furthermore, in addition to NGF both human brain and umbilical cord-derived endothelial cells produce also brain-derived neurotrophic factor (BDNF) supporting a general role of neurotrophins in the dynamic maintenance of the endothelial barrier function. [38, 39] and may suggest that capillary microvasculature act to support neuronal recruitment and survival by secreting neurotrophins. TrkA and p75 NGF receptors are present and active in choroidal and retinal [40], aorta [33, 41], brain capillary [34], umbilical vein [32] and adrenal medullae [42] endothelial cells. Figure 1 presents a typical experiment in which NGF in a dose dependent fashion induced TrkA receptor autophosphorylation in adrenal medulla endothelial cells, a process blocked by K252a, a TrkA antagonist. To date, very few studies have addressed the signaling pathways activated by NGF in endothelial cells. A novel approach to characterize multiple signaling pathways activated

Angiogenic effects of NGF

103

by NGF in endothelial cells was initiated in our laboratories by measuring NGF effects on the phosphorylation of some specific signaling molecules in dermal capillary endothelial cells using the Invitrogene® human Mercator™ PhosphoArray system, shown in Figure 2. The system is composed of 4 microscope slides (Figure 2A) each containing 16 pads (Figure 2B). Each pad is precoated with ten different monoclonal antibodies, each directed towards a specific signaling molecule, spotted in triplicate (Figure 2C).

Figure 1. NGF induced TrkA phosphorylation and the inhibitory effect of K252a, TrkA antagonist in rat adrenal medulla endothelial cells. Semiconfluent cultures of cells were treated with mouse β-NGF at the concentration mentioned for a period of 30 min in the absence (A) or presence of 1 µM K252a (B).The dose-dependent increase in TrkA phosphorylation was measured by western blotting of the immunoprecipitated TrkA receptor and visualized with phosphotyrosine antibodies. Note the strong inhibitory effect of K252a.

Figure 2. Invitrogen® Mercator™ phospho array system and working protocol. (A) – Mercator™ phospho array system holder. Left side: empty, middle: two slides each containing 16 pads, in each pad layout of 36 spots containing antibodies for signaling molecules; right side: slide covered with red rubber frame to generate wells on the top of the pads to allow antibody reactions. (B) – schematic of 16 pad slides. (C) – the layout of triplicate antibodies for signaling molecules in a single pad. The negative controls measure non specific association of the lysate proteins to the pad (yellow spots). (D) – working protocol with the Mercator™ phospho array system.

104

Shimon Lecht et al.

Figure 3. NGF-stimulated phosphorylation of signaling molecules in human dermal capillary endothelial cells measured by Invitrogen® Mercator™ phospho array system. Cultures of endothelial cells were treated with NGF (10 ng/ml) or left untreated, lysed and incubated with the pads using the protocol described in Figure 2D. The fluorescence was measured and the activity calculated from calibration curves is presented at the different time points.

Upon incubation of the slides with lysates of endothelial cell treated with NGF or untreated (control) for different periods of times, the signaling molecules are “fished-out” by the specific monoclonal antibodies immobilized on the pad (Figure 2D). Thereafter, the glass slides are incubated with phosphospecific antibodies against the captured antigens and a secondary antibody fluorescent conjugate to detect the level of phosphorylation of the intracellular signaling molecules (Figure 2D). The level of phosphorylation of signaling molecules is measured with a DNA scanner and is transformed to phosphorylation values using dose-response curves for each individual signaling molecule as established though calibration experiments. The measurements are highly sensitive, specific and reproducible, and comparable to those obtained by western blotting. Figure 3 indicates that upon stimulation with NGF focal adhesion kinase (FAK), Src, paxillin, Akt, p38 and HSP27 are rapidly stimulated by NGF, while EGF receptor and ATF transcription factor were not affected. NGF-induced stimulation of FAK on Tyr397, and Src on Tyr418 and their substrate paxillin on Tyr118, as well as Akt on Ser473 is most probably related to NGF-induced migration of endothelial cells. This conclusion supports recent findings on apparent relationship between NGF induced migration of pig aortic endothelial cells and PI3K/Akt stimulation [41]. Our data suggest that by activating TrkA receptors in dermal endothelial cells NGF initiated a cascade of signaling molecules similar to its effects in neuronal cells [43]. A comparison of NGF-induced, TrkA mediated signaling in endothelial cells and in neurons is currently under investigation in our laboratory.

NGF induced proliferation of endothelial cells is TRKA and ERK mediated Over the last years it becomes clear that NGF is a mitogen for endothelial cells from different vascular beds and species [4]. Thus, NGF induced in vitro proliferation of

Angiogenic effects of NGF

105

Figure 4. NGF induced transient Erk phosphorylation in mouse brain microcapillary bEnd.3 endothelial cell line. Semiconfluent cultures of mouse brain endothelial cells, serum starved overnight, were incubated with 10 ng/ml mouse β-NGF for the periods of time indicated. The lysates of the cells were submitted to SDS-PAGE followed by western blotting with phospho-Erk1 followed by pan-Erk1/2 (insert). The quantitation of Erk phosphorylation activity is presented as a percent of control (0 min) and the kinetics is shown for a period of 60 min.

endothelial cells isolated from human umbilical vein [32], immortalized rat brain [34], human eye choroidal [40] and human dermal microvasculature [44] endothelial cells. The involvement of the TrkA receptor in NGF-induced proliferation of human umbilical vein endothelial cells was clearly demonstrated by the inhibitory effect of K252a [32], a well known antagonist of NGF action [45]. Using a unique clone of endothelial cells from the rat adrenal medulla lacking TrkA receptors [46], we provided evidence that TrkA receptors are required for the NGF-induced proliferative response. Furthermore, overexpression of human TrkA receptor in this endothelial line resulted in NGF-induced signal transduction and DNA synthesis, thus reconstituting the proliferation response of the cells in response to NGF [42]. Other in vitro studies have implicated MAPK signaling in the NGF-induced vascular effects [32, 47] and suggested that NGF may play an autocrine role in the endothelium [32, 33, 34]. Indeed, as shown in Figure 4, NGF induced transient Erk phosphorylation, with the typical kinetics required for proliferation [48]. Further cellular studies are required to identify the cell cycle regulatory proteins (cyclins, cdc kinases, etc) that are modulated by NGF in endothelial cells.

NGF – induced migration of endothelial cells is TRKA mediated To measure NGF effects on the migration of endothelial cells, we utilized an novel omnidirectional migration assay using human aortic endothelial cells (HAEC) and validated the assay with known angiogenic mitogens, such as human recombinant basic fibroblast growth factor (bFGF, Figure 5) and human recombinant vascular endothelial growth factor (VEGF, [49]).

106

Shimon Lecht et al.

The potency of nerve growth factor purified from various species (viper, mouse, and human recombinant) to stimulate HAEC migration was similar to that of VEGF and bFGF (EC50 of 0.5 ng/ml). Recombinant human bFGF was significantly more efficacious than either NGF or VEGF, both of which stimulated HAEC migration by 30% over basal spontaneous migration [49]. NGF-mediated stimulation of HAEC migration was completely blocked by the NGF/TrkA receptor antagonist K252a (30 nM) (Figure 5C) but not by the VEGF/Flk receptor antagonist SU-5416 (250 nM) [49], indicating a direct effect of NGF via TrkA receptor activation on HAEC migration. NGF stimulation of HAEC migration was additively increased by either VEGF or bFGF, suggesting a potentiating interaction between their tyrosine kinase receptor signaling pathways [49]. Other investigators using human eye choroidal [40] and pig aortic endothelial cells [41] have also demonstrated NGF-induced migration using invasion through porous membrane as their assay. The ability of NGF to stimulate migration of endothelial cells in vitro strongly implies that this factor may play an important angiogenic role in the cardiovascular system, in addition to its well known effects in the nervous system.

Figure 5. NGF-induced migration of human aortic endothelial cells in the omnidirectional ring migration assay. The organization of the endothelial cell monolayer at 0 (A) and 72 (B) hours after plating. Before plating, the bottom side of the well was eatched to generate a reference line with the radius of 2 mm (dark lines) representing the distance from the center of the ring to the periphery. The front of the monolayer is represented by the green curve. For example, by comparing the front of the cells at 0 time and 72 hours after plating we measure the increase in the migration of the cell monolayer. To estimate precisely the distance of migration, black lines were superimposed on the images and 60 measurements of the radii from the center of the ring to the front were made. (C) The effect of NGF on endothelial cell migration and the inhibitory effect of K252a. Endothelial cells were plated at a density of 20,000 cells/ring and were treated with either NGF (50 ng/ml) in the absence or presence of a nontoxic concentration of 30 nM K252a or bFGF (10 ng/ml) or K252a alone or left untreated (control). Cell migration was measured in microns and is presented as percentage of control (Mean ± SD). * p < 0.01 (compared to control); ** p < 0.01 (compared to NGF alone).

Angiogenic effects of NGF

107

NGF-induced rat aortic ring sprouting The aortic ring assay is a relatively simple ex vivo model of angiogenesis, in which angiogenic sprouting results from complex interactions between endothelial cells, vascular smooth muscle, pericytes and the extracellular matrix components. Rat aortic ring explant cultures were prepared by a modification of a standard method [50]. The ability of NGF to induce sprouting in rat aortic rings is shown in Figure 6. Quantitative evaluation of micrographs of aortic rings treated with NGF for 3 days showed a significant dose–dependent enhancement of vascular sprouting (Figure 6B, C), as compared to the control aortic rings incubated with medium alone (Figure 6A). Based on these results we suggest a role of NGF in aortic ring angiogenesis by the concomitant promotion of matrix remodeling, endothelial cell migration, and subsequent sprout formation, i.e. through a concerted induction of diverse processes. This notion of the complex action of NGF in angiogenesis is supported in part by its previously described in vitro effects.

Figure 6. NGF-induced rat aortic ring sprouting. Rat aortic rings grown in collagen gel were incubated with NGF at different concentration or VEGF or with medium alone (control) for 3 days. Microvessel formation was photographed (A – control; B – NGF 300 ng/ml) and quantified by direct counting (C) – data represent Mean ± SEM of triplicate experiments. * p < 0.01 (compared to control).

NGF promotes angiogenesis in the quail chorioallantoic membrane To confirm the in vivo relevance of the in vitro and ex vivo angiogenic effects of NGF, we attempted a quantitative analysis of the angiogenic properties of NGF in the quail chorioallantoic membrane (CAM) model of angiogenesis and compared the

108

Shimon Lecht et al.

angiogenic effects of NGF from different species to those of human recombinant VEGF165 and bFGF [51]. Using the quail CAM assay, we measured NGF effects on the natural vascularization focusing on the morphology of small blood vessels of the quail CAM arterial tree (Figure 7A) as previously described [52, 53]. For quantitative morphometry, grayscale images of the blood vessels in the CAM arterial tree where binarized, skeletonized and then quantitated and expressesd as fractal dimension (Df). In comparison to phosphate buffered saline-treated controls, NGFs from different species (mouse, viper and cobra) increased the rate of angiogenesis in a dose-response fashion from 0.5 to 5 µg. The maximal angiogenic effect in mouse NGF treated embryos was Df = 1.275±0.059 compared to Df = 1.148±0.018 for control embryos (Figure 7B and [51]). This stimulatory effect was similar to that of 0.5 µg rhVEGF (Df =1.248±0.029, p<0.001) and 1.5 µg rhbFGF (Df = 1.264±0.02, p<0.001). The mouse NGF - induced angiogenic effect was blocked by

Figure 7. NGF promotes angiogenesis in the quail chorioallantoic membrane. (A) – Assay of angiogenesis. Fertilized eggs at embryonic day 0 (E0) were cleaned and maintained for 3 days in an incubator. At E3 the eggs were opened and the embryo transferred to a 6 well tissue culture dish. At E7 growth factors and drugs were applied in drops on the surface of each CAM. At E8 the embryos were fixed for two days, thereafter dissected, and mounted on a glass slide. Images of distal arterial branches located in the central region of the CAM were acquired and analyzed by NIH ImageJ software. The images were binarized and skeletonized to calculate the fractal coefficient Df [52, 52]. (B) – the angiogenic effect of NGF on the morphometry of the quail CAM. Upper pictures represent the arterial vasculature; quantitative image analysis expressed as change in Df in response to growth factors and drugs treatment. * p < 0.05 compared to control; ** p < 0.01 compared to NGF.

Angiogenic effects of NGF

109

1 µM K252a (Df = 1.149±0.018, p<0.001, Figure 7B). As discussed above, K252a is a selective antagonist of the NGF/TrkA receptor. These data again indicate a selective angiogenic effect of NGF via activation of the quail embryo TrkA receptor and confirm our previous findings on NGF induction of neovascularization in the chick embryo CAM [32]. In the latter study, gelatin sponges embedded with NGF were surrounded by allantoic vessels that developed radially towards the implant after 4 days of treatment [32]. The angiogenic effects in avian models are also supported by the vasculogenic effects of NGF after induction of limb ischemia produced in rats by surgical excision of left femoral artery [54] and upon NGF application on human pressure ulcers [55]. Taken together, these findings imply that NGF, in addition to its well-known effects in the nervous system, may also play angiogenic, physiological role in the cardiovascular system.

NGF expression and TRKA activation are common events in solid ovarian carcinoma tumor and associated blood vessels As part of our efforts to define the role of NGF and its receptors in ovarian cancer evolution and metastasis, we have recently reported on the expression of NGF and its receptors TrkA and p75 in effusions, primary and metastatic tumors of serous ovarian

Figure 8. Expression of NGF and TrkA compared to angiogenic molecules in human serous ovarian carcinoma tumors. Immunohistochemistry of tumor sections stained with rabbit NGF polyclonal antibody, monoclonal anti phospho TrkA antibody and 203-rabbit polyclonal antibody panTrkA. Angiogenic molecules VEGF and bFGF were visualized by in situ hybridization using probes against human VEGF165 and human bFGF. (A) – NGF expression (brown color) in cancer cells in a serous ovarian carcinoma. (B) – weak diffuse VEGF mRNA expression in both the tumor and the stromal cell compartments. (C) – intense expression of the bFGF mRNA (dark blue) in both tumor and stromal cell compartments. (D) – TrkA protein membrane expression (dark brown) in a primary serous ovarian carcinoma. (E) – phosphoTrkA membranal expression (brown) in peritumor vessels in the vicinity (left) of a primary carcinoma. (F) – vascular phosphoTrkA expression (brown) at a peritumoral site.

110

Shimon Lecht et al.

carcinoma patients [56, 57, 58]. The expression and activity of NGF and its receptors were compared to the expression of other angiogenic factors, such as bFGF and VEGF (Figure 8), of conventional markers of cell cycle and apoptosis, as well as of other signaling molecules. Using an antibody against the phosphorylated form of the TrkA receptor, we found intense expression of activated TrkA receptor on the membranes of intratumoral and peritumoral blood vessels (Figure 8). NGF expression was predominant in about 80% of the tumors. NGF and TrkA expression correlated with that of bFGF protein as well as bFGF mRNA, as well as with the expression of VEGF mRNA in tumors and stromal cells. Coexpression of NGF with angiogenic molecules in blood vessel endothelial cells (Figure 8) suggests that the pro angiogenic role attributed to NGF in in vitro and in vivo models maybe clinically relevant markers in assessing cancer progression and prognosis. The expression of NGF and activated TrkA receptor correlated with poor outcome in advanced stage serous ovarian carcinoma patients as evident from Kaplan-Meier survival curves [58]. The expression and activation of TrkA receptor was reduced in effusion (metastatic cells) compared with solid ovarian carcinoma and appeared to be independent of cell cycle progression in the tumor. Taken together these findings indicate that NGF/TrkA may serve as a novel prognostic marker in ovarian carcinoma and suggest a pro-angiogenic role for the NGF/TrkA autocrine loop, which may become an attractive target for anti cancer drug development.

Conclusions NGF, the prototype growth factor of the neurotrophin family, induces proliferation and migration of different cultured endothelial cells, as well as sprouting of capillaries from aortic rings and enhanced formation of small arteries in the quail CAM assay. In addition, NGF and its receptors are also highly co-expressed in serous ovarian carcinoma tumoral blood vessels concomitant with angiogenic growth factors, bFGF and VEGF. Cumulatively, these findings provide new evidences for a role of NGF as a pleiotropic pro-angiogenic growth factor in the embryonic cardiovascular system as well as during inflammation and tumor growth in adult organisms. Therefore, NGF may be regarded as the prototypic member of major growth factors families with dual activity on both cardiovascular and nervous systems due to the presence of TrkA receptors in both systems (Figure 9). Angiogenesis and neurogenesis are distinct yet, as we know now, related biological processes that primarily occur during embryonal development but continue to play an important physiological role in the adult. We propose that upon different insults (ischemia, injury, etc) the combined action of angiogenic factors (VEGF, bFGF) and neurotrophins, such as NGF, contribute to the regeneration of both nerves and blood vessels. The angiogenic properties of NGF, as well as the recently described neuronal effects of classical “angiogenic” factors, such as bFGF and VEGF [4], strongly support the novel pharmacological concept that these growth factors are the main interlocutors in the crosstalk between the cardiovascular and nervous system, which is instrumental for adequate neovascularization and innervation of regenerating tissue. There is growing evidence in tumor biology to suggest that both NGF and bFGF/VEGF may contribute to tumor angiogenesis and cancer cell growth. The increased

Angiogenic effects of NGF

111

Figure 9. Schematic of the dual angiogenic and neurotropic activity of NGF mediated by TrkA receptors in the nervous (left) and cardiovascular (right) systems.

knowledge about the role of these growth factors tyrosine kinase receptors in the etiology of and in the pathological pathways involved in tumor progression has lead to the recent implementation of novel, more specific therapeutic approaches based on either neutralization of the growth factor with selective monoclonal antibody (e.g. Avastin® {bevacizumab} for VEGF) or inhibition of the tyrosine kinase receptors using small molecules (e.g. SU-11248 for VEGF/Flk receptor). It is tempting to propose that in a similar fashion anti-NGF antibodies and/or K252a-like derivatives, or other antagonists of NGF/TrkA receptors, will find their way into the clinic, as antiangiogenic, anticancer drugs.

Acknowledgements The authors would like to acknowledge the help of Mrs. Zehava Cohen and Mr. Tigran Haroutiunian with the preparation of the figures. This work was supported in part by a grant-in-aid form the Stein Family Foundations (PL and PIL) and a grant-in-aid from the National Space and Aeronautics Agency (NASA) to PIL.

References 1. 2. 3. 4. 5. 6.

Carmeliet, P. 2000, Nat Med, 6, 389. Mukouyama, Y.S., Shin, D., Britsch, S., Taniguchi, M., and Anderson, D.J. 2002, Cell, 109, 693. Miller, G. 2002, Science, 296, 2121. Lazarovici, P., Marcinkiewicz, C., and Lelkes, P.I. 2006, Curr Pharm Des, 12, 2609. Chao, M.V. 2003, Nat Rev Neurosci, 4, 299. Levi-Montalcini, R. 1987, Embo J, 6, 1145.

112

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

Shimon Lecht et al.

Jaaro, H., Beck, G., Conticello, S.G., and Fainzilber, M. 2001, Trends Neurosci, 24, 79. Ullrich, A., Gray, A., Berman, C., and Dull, T.J. 1983, Nature, 303, 821. Bierl, M.A., Jones, E.E., Crutcher, K.A., and Isaacson, L.G. 2005, Neurosci Lett, 380, 133. Gnahn, H., Hefti, F., Heumann, R., Schwab, M.E., and Thoenen, H. 1983, Brain Res, 285, 45. Unsicker, K., Krisch, B., Otten, U., and Thoenen, H. 1978, Proc Natl Acad Sci U S A, 75, 3498. Barbacid, M. 1995, Ann N Y Acad Sci, 766, 442. Sofroniew, M.V., Howe, C.L., and Mobley, W.C. 2001, Annu Rev Neurosci, 24, 1217. Tabakman, R., Lecht, S., Sephanova, S., Arien-Zakay, H., and Lazarovici, P. 2004, Prog Brain Res, 146, 387. Puxeddu, I., Ribatti, D., Crivellato, E., and Levi-Schaffer, F. 2005, J Allergy Clin Immunol, 116, 531. Kawamoto, K., and Matsuda, H. 2004, Prog Brain Res, 146, 369. Micera, A., Lambiase, A., Puxeddu, I., Aloe, L., Stampachiacchiere, B., Levi-Schaffer, F., and Bonini, S. 2006, Exp Eye Res, 83, 747. Kaplan, D.R., and Miller, F.D. 1997, Curr Opin Cell Biol, 9, 213. Miller, F.D., and Kaplan, D.R. 2001, Cell Mol Life Sci, 58, 1045. Barbacid, M., Lamballe, F., Pulido, D., and Klein, R. 1991, Biochim Biophys Acta, 1072, 115. Hempstead, B.L., Martin-Zanca, D., Kaplan, D.R., Parada, L.F., and Chao, M.V. 1991, Nature, 350, 678. Klein, R., Nanduri, V., Jing, S.A., Lamballe, F., Tapley, P., Bryant, S., Cordon-Cardo, C., Jones, K.R., Reichardt, L.F., and Barbacid, M. 1991, Cell, 66, 395. Martin-Zanca, D., Oskam, R., Mitra, G., Copeland, T., and Barbacid, M. 1989, Mol Cell Biol, 9, 24. Meakin, S.O., Suter, U., Drinkwater, C.C., Welcher, A.A., and Shooter, E.M. 1992, Proc Natl Acad Sci U S A, 89, 2374. Fujita, K., Lazarovici, P., and Guroff, G. 1989, Environ Health Perspect, 80, 127. Arevalo, J.C., Conde, B., Hempstead, B.L., Chao, M.V., Martin-Zanca, D., and Perez, P. 2000, Mol Cell Biol, 20, 5908. Greene, L.A., and Kaplan, D.R. 1995, Curr Opin Neurobiol, 5, 579-87. Zettler, C., Bridges, D.C., Zhou, X.F., and Rush, R.A. 1996, J Neurosci Res, 46, 581. Clemow, D.B., Spitsbergen, J.M., McCarty, R., Steers, W.D., and Tuttle, J.B. 1998, Exp Cell Res, 244, 196. Donovan, M.J., Miranda, R.C., Kraemer, R., McCaffrey, T.A., Tessarollo, L., Mahadeo, D., Sharif, S., Kaplan, D.R., Tsoulfas, P., Parada, L., and et al. 1995, Am J Pathol, 147, 309. Gibran, N.S., Tamura, R., Tsou, R., and Isik, F.F. 2003, Shock, 19, 127. Cantarella, G., Lempereur, L., Presta, M., Ribatti, D., Lombardo, G., Lazarovici, P., Zappala, G., Pafumi, C., and Bernardini, R. 2002, Faseb J, 16, 1307. Tanaka, A., Wakita, U., Kambe, N., Iwasaki, T., and Matsuda, H. 2004, Biochem Biophys Res Commun, 313, 1009. Moser, K.V., Reindl, M., Blasig, I., and Humpel, C. 2004, Brain Res, 1017, 53. Biernacki, K., Antel, J.P., Blain, M., Narayanan, S., Arnold, D.L., and Prat, A. 2005, Arch Neurol, 62, 563. Toti, P., Ciarmela, P., Florio, P., Volpi, N., Occhini, R., and Petraglia, F. 2006, J Endocrinol Invest, 29, 337. Xu, W.P., Mizuno, N., Shiba, H., Takeda, K., Hasegawa, N., Yoshimatsu, S., Inui, T., Ozeki, Y., Niitani, M., Kawaguchi, H., Tsuji, K., Kato, Y., and Kurihara, H. 2006, J Periodontol, 77, 800. Kim, H., Li, Q., Hempstead, B.L., and Madri, J.A. 2004, J Biol Chem, 279, 33538. Leventhal, C., Rafii, S., Rafii, D., Shahar, A., and Goldman, S.A. 1999, Mol Cell Neurosci, 13, 450.

Angiogenic effects of NGF

113

40. Steinle, J.J., and Granger, H.J. 2003, Auton Neurosci, 108, 57. 41. Rahbek, U.L., Dissing, S., Thomassen, C., Hansen, A.J., and Tritsaris, K. 2005, Pflugers Arch, 450, 355. 42. Jiang, H., Movsesyan, V., Fink, D.W., Jr., Fasler, M., Whalin, M., Katagiri, Y., Monshipouri, M., Dickens, G., Lelkes, P.I., Guroff, G., and Lazarovici, P. 1997, J Cell Biochem, 66, 229. 43. Lazarovici, P., Matsuda, Y., Kaplan, D.R., and Guroff, G. 1997, Toxins and signal transduction, Y. Gutman and P. Lazarovici (Ed.), Harwood Academic Publishers, Amsterdam, 69. 44. Raychaudhuri, S.K., Raychaudhuri, S.P., Weltman, H., and Farber, E.M. 2001, Arch Dermatol Res, 293, 291. 45. Koizumi, S., Contreras, M.L., Matsuda, Y., Hama, T., Lazarovici, P., and Guroff, G. 1988, J Neurosci, 8, 715. 46. Rasouly, D., Shavit, D., Zuniga, R., Elejalde, R.B., Unsworth, B.R., Yayon, A., Lazarovici, P., and Lelkes, P.I. 1996, J Cell Biochem, 62, 356. 47. Calza, L., Giardino, L., Giuliani, A., Aloe, L., and Levi-Montalcini, R. 2001, Proc Natl Acad Sci U S A, 98, 4160. 48. Marshall, C.J. 1995, Cell, 80, 179-85. 49. Dolle, J.P., Rezvan, A., Allen, F.D., Lazarovici, P., and Lelkes, P.I. 2005, J Pharmacol Exp Ther, 315, 1220. 50. Nicosia, R.F., and Ottinetti, A. 1990, Lab Invest, 63, 115. 51. Lazarovici, P., Gazit, A., Staniszewska, I., Marcinkiewicz, C., and Lelkes, P.I. 2006, Endothelium, 13, 51. 52. Parsons-Wingerter, P., Lwai, B., Yang, M.C., Elliott, K.E., Milaninia, A., Redlitz, A., Clark, J.I., and Sage, E.H. 1998, Microvasc Res, 55, 201. 53. Parsons-Wingerter, P., Elliott, K.E., Clark, J.I., and Farr, A.G. 2000, Arterioscler Thromb Vasc Biol, 20, 1250. 54. Emanueli, C., Schratzberger, P., Kirchmair, R., and Madeddu, P. 2003, Br J Pharmacol, 140, 614. 55. Bernabei, R., Landi, F., Bonini, S., Onder, G., Lambiase, A., Pola, R., and Aloe, L. 1999, Lancet, 354, 307. 56. Odegaard, E., Staff, A.C., Abeler, V.M., Kopolovic, J., Onsrud, M., Lazarovici, P., and Davidson, B. 2007, Hum Pathol, 38, 140. 57. Davidson, B., Lazarovici, P., Ezersky, A., Nesland, J.M., Berner, A., Risberg, B., Trope, C.G., Kristensen, G.B., Goscinski, M., van de Putte, G., and Reich, R. 2001, Clin Cancer Res, 7, 3457. 58. Davidson, B., Reich, R., Lazarovici, P., Nesland, J.M., Skrede, M., Risberg, B., Trope, C.G., and Florenes, V.A. 2003, Clin Cancer Res, 9, 2248.