Potentiated Angiogenic Effect of Scatter Factor/Hepatocyte Growth Factor via Induction of Vascular Endothelial Growth Factor
The Case for Paracrine Amplification of Angiogenesis
Background—Scatter factor/hepatocyte growth factor (SF/HGF) is a pleiotropic growth factor that stimulates proliferation and migration of endothelial cells (ECs) via the c-Met receptor, present on ECs as well as other cell types, including smooth muscle cells (SMCs). We studied the effects of recombinant human (rh) SF/HGF in vitro and in vivo in a rabbit model of hindlimb ischemia. We further compared these effects with those of recombinant human vascular endothelial growth factor (rhVEGF165), an EC–specific mitogen.
Methods and Results—In vitro, rhSF/HGF and rhVEGF165 exhibited similar effects on proliferation and migration of ECs. When both cytokines were administered together, the result was an additive effect on EC proliferation and a synergistic effect on EC migration. Application of rhSF/HGF to cultures of human SMCs resulted in the induction of VEGF mRNA and protein. In vivo, administration of rhSF/HGF (500 μg×3) was associated with significant improvements in collateral formation (P<.001) and regional blood flow (P<.0005) and with a significant reduction in muscle atrophy (P<.0001). These effects were significantly more pronounced than those of rhVEGF165 administered according to the same protocol (P<.05). Neither remote angiogenesis nor other pathological sequelae were observed with either rhSF/HGF or rhVEGF165.
Conclusions—The pleiotropic effects of certain growth factors may potentiate angiogenesis via a combination of direct effects on EC proliferation and migration and indirect effects that result in the generation of other potent EC mitogens from non-EC populations. The synergistic effects demonstrated when SF/HGF and VEGF are administered together in vitro may be reproduced in vivo by SF/HGF-induced upregulation of VEGF in vascular SMCs.
Angiogenesis involves activation, migration, and proliferation of ECs1 and is regulated by certain growth factors.2 Exogenous administration of growth-regulating molecules that stimulate angiogenesis has potential utility for the treatment of tissue ischemia not amenable to conventional revascularization techniques. This strategy, referred to as “therapeutic angiogenesis,” has been validated in various animal models of limb or myocardial ischemia3 4 5 6 7 8 and more recently in patients treated with plasmid DNA encoding for VEGF.9 10
SF/HGF is a recently characterized growth factor11 12 that has a disulfide-linked heterodimer structure and an apparent molecular weight of 80 kD.11 Its receptor has been identified as the c-Met proto-oncogene product, a transmembrane tyrosine kinase.12 Synthesis of SF/HGF by mesenchymal cells (eg, fibroblasts and SMCs), coupled with demonstrated effects on epithelial cells and ECs, suggests a paracrine mode of action.13 SF/HGF is considered to be a principal mediator of mesenchymal-epithelial/endothelial interactions that contribute to embryogenesis, organ regeneration, wound healing, and angiogenesis.13 14 15 16
Indeed, previous investigations have established that SF/HGF directly stimulates proliferation and migration of cultured ECs,17 18 promotes development of capillary-like structures in vitro,18 19 and stimulates blood vessel formation in Matrigel plugs and normal cornea.16 18 Furthermore, SF/HGF has been shown to be induced in skeletal muscle after ischemic injury.14 More recently, SF/HGF has been implicated in capillary EC regeneration in the ischemically injured myocardium.20
Previous studies from our laboratory and others have established that certain cytokines, such as platelet-derived growth factor-BB and transforming growth factor-β1, can upregulate VEGF expression in vascular SMCs and thereby exert indirect angiogenic effects,21 even though they are nonmitogenic or even inhibitory for EC proliferation in vitro.22 23 Because the c-Met receptor for SF/HGF is expressed on both ECs and vascular SMCs,24 we considered the possibility that SF/HGF could also upregulate VEGF expression in SMCs. Thus, the angiogenic potency of SF/HGF in vivo might be enhanced through coordinately regulated direct and indirect effects.
The present series of experiments demonstrates that the direct effect of SF/HGF on EC proliferation and migration is similar to that of VEGF; the combination of SF/HGF and VEGF, however, is shown to produce an additive effect on EC proliferation and a synergistic effect on EC migration. Moreover, in vitro administration of SF/HGF to HVSMCs induces upregulated expression of both VEGF mRNA and protein. In vivo, angiographic, physiological, and histological findings indicate that SF/HGF-induced angiogenesis is superior to that achieved with VEGF. SF/HGF is thus a potent agent for strategies designed to promote therapeutic angiogenesis, possibly as a result of the combination of direct effects of SF/HGF on ECs and the indirect effects, including paracrine upregulation of VEGF on SMCs, as demonstrated in vitro.
rhSF/HGF and rhVEGF165
rhSF/HGF was produced in Chinese hamster ovary cells transfected with a plasmid encoding the full-length, natural human SF/HGF sequence. The rhSF/HGF was then converted into the bioactive, heterodimer form by a modification of the method described by Naka et al25 and purified by conventional chromatography, as previously described.26 The 165-amino-acid homodimeric species of rhVEGF165 was purified from transfected Chinese hamster ovary cells as previously described.27 The purity of the material was assessed by a silver-stained SDS-PAGE gel and the presence of a single NH2-terminal amino acid sequence.
HUVECs were prepared from umbilical cord vein as previously described28 and grown in medium 199 (M199) (GIBCO BRL) supplemented with 20% FBS, EC growth supplement (100 μg/mL), and heparin (50 U/mL). HVSMCs were cultured by explant outgrowth from unused portions of internal mammary arteries obtained at coronary artery bypass graft surgery as previously described.29 Cells were cultured in DMEM supplemented with 10% FBS.
Northern Blot Analysis in Cultured HVSMCs
Confluent HVSMCs (passages 3 to 5) were growth-arrested in filtered DMEM containing 0.5% FBS for 48 hours before each experiment. The cells were then treated with rhSF/HGF for various time intervals and different concentrations as indicated. Stimulation of HVSMCs with PMA (Sigma Chemical Co) was used as a positive control for VEGF gene expression.30 Total RNA from HVSMCs was isolated by phenol/chloroform extraction,31 and Northern blot analysis was performed as previously described.21 The DNA probe for human VEGF was a 675-bp EcoRI/Bgl II fragment (gift of N. Ferrara, Genentech Inc).
RT-PCR Analysis in Cultured HVSMCs
First-strand cDNA synthesis and PCR analysis were performed according to standard procedures.32 33 The primer chosen for human VEGF (Genbank accession number M32977) was, for sense, (5′-3′) GAACTTTCTGCTGTCTTGGG and, for antisense, (5′-3′) TCACCGCCTCGGCTTGTCAC. PCR resulted in three bands (438, 570, and 642 bp) corresponding to the three principal VEGF isoforms, 121, 165, and 189, respectively, expressed in HVSMCs.30 For human GAPDH (X01677), the following primers were used: sense, (5′-3′) TGAAGGTCGGAGTCAACGGATTTG and antisense, (5′-3′) CATGTGGGCCATGAGGTCCACCAC. PCR resulted in a 983-bp band. The linear phase of the exponential range amplification was determined for each primer set to allow semiquantitative PCR analysis.34 The number of cycles was then chosen in the linear phase of amplification: 29 cycles for VEGF and 20 cycles for GAPDH. Ten microliters of each PCR reaction mixture was electrophoresed in a 1.5% agarose gel, and bands were visualized by ethidium bromide staining. Bands were quantified densitometrically by scanning and analyzing with Eagle’s Eye II software. To normalize signals for VEGF165, the value was divided by the signal for GAPDH, a widely invariant and highly expressed gene. The data presented as relative values (VEGF165/GAPDH) were plotted against time. The results represent three independent amplifications from two separate studies.
Analysis of VEGF Protein Expression in Cultured HVSMCs
For analysis of VEGF protein expression in SMCs by Western blotting, cells were lysed by addition of 1 mL RIPA buffer (1% NP-40, 0.5% sodium deoxycholic acid, 0.1% SDS in PBS, pH 7.4; 1 μmol/L leupeptin; 5 μmol/L aprotinin; 1 mmol/L PMSF; and 1 μmol/L pepstatin, all Sigma Chemical Co) per 100-mm plate. Protein extracts (100 μg) were separated on a 10% SDS-PAGE and transferred to a 0.2-μm PVDF membrane (Bio Rad). The membranes were blocked in 10% nonfat dry milk/0.2% Tween-20 in PBS, pH 7.4, then immunoblotted with a rabbit polyclonal anti-human VEGF antibody (1:200; Santa Cruz) overnight at 4°C. Blots were washed with 0.2% Tween-20 in PBS and incubated with horseradish peroxidase–linked goat anti-rabbit antibody (1:7500; Santa Cruz) for 45 minutes. Immunoreactive bands were visualized with ECL reagent (Amersham).
EC Proliferation Assay
For evaluation of proliferation, the colorimetric MTT assay was performed as previously described.35 Briefly, HUVECs were seeded in a 96-well plate (5000 cells/well) in 0.2 mL M199 (GIBCO BRL) containing 5% FBS. MTT was dissolved in PBS at 5 mg/mL to constitute a stock solution. After the incubation of HUVECs with rhSF/HGF, rhVEGF165, or a combination of the two for 48 hours, the MTT solution was added to each well (10 μL/100 μL medium), and plates were incubated at 37°C for 4 hours. The medium was decanted, and DMSO was added to all wells to dissolve the dark blue crystals. Plates were read on an SLT EAR 400 AT automatic plate reader (SLT Labinstruments) with a test wavelength of 560 nm and a reference wavelength of 650 nm. All experiments were performed in quadruplicate.
EC Migration Assay
EC migration assays were performed in a 48-well microchemotaxis chamber (Neuroprobe Inc).36 PVP-free polycarbonate filters with a pore size of 8 μm (Nuclepore Corp) were coated with 0.1% gelatin for at least 6 hours at room temperature and dried under sterile air. Test substances were diluted to appropriate concentrations in M199 supplemented with 1% FBS, and 25 μL of the final dilution was placed in the lower chamber of the modified Boyden apparatus. Subconfluent, early-passage2 3 4 5 6 HUVEC cultures were washed and trypsinized for the minimum time required to achieve cell detachment. After the filter was placed between the lower and upper chambers, 2.5×105 cells suspended in 50 μL M199 containing 1% FBS were seeded in the upper compartment. The apparatus was then incubated for 5 hours at 37°C in a humidified chamber with 5% CO2 to allow cell migration. After the incubation period, the filter was removed, and the upper side of the filter with the nonmigrated cells was scraped with a rubber policeman. The filters were fixed with methanol and stained with a Giemsa solution (Diff-Quick, Baxter). Migration was quantified by counting cells of three random high-power fields (×100) in each well, and all experiments were performed in quadruplicate.
The physiological response to administration of rhVEGF165 and rhSF/HGF was investigated in a previously described rabbit ischemic hindlimb model.37 All protocols were approved by the St Elizabeth’s Institutional Animal Care and Use Committee. Twenty-seven male New Zealand White rabbits (3.5 to 4.0 kg) (Pine Acre Rabbitry, Norton, Mass) were anesthetized with a mixture of amine (50 mg/kg) and acepromazine (0.8 mg/kg) after premedication with xylazine (2 mg/kg). They underwent surgical removal of one femoral artery by a previously described technique.37 In this model, the excision of the femoral artery results in retrograde propagation of thrombus and occlusion of the external iliac artery. Consequently, blood flow to the ischemic limb is dependent on collateral vessels originating from the internal iliac artery.37
Design of the In Vivo Experiment
The dose of rhVEGF165 used in the present experiment (500 μg) has previously been shown to induce an optimal angiogenic effect of rhVEGF165 with a single bolus administration protocol.3 38 To ensure a maximum effect of rhVEGF165 alone, we administered three separate injections of 500 μg: the first dose locally (intra-arterially into the ischemic limb) and then two additional doses intravenously; the latter has also been shown to achieve a reliable biological effect.39 Because the dose-response curves observed with rhVEGF165 and rhSF/HGF on migration and proliferation of cultured ECs were similar (see “Results”), the same dose of SF/HGF (500 μg per injection) was chosen for comparison.
Ten days after surgery (day 0) and after measurement of baseline body weight as well as baseline noninvasive and invasive hemodynamic parameters (see below), animals received a single intra-arterial bolus of rhSF/HGF (500 μg, n=9), rhVEGF165 (500 μg, n=9), or vehicle solution (3 mL saline with 0.1% rabbit serum albumin [Sigma], n=9) administered as a bolus over 1 minute through a 3F end-hole infusion catheter (Tracker-18, Target Therapeutics) positioned in the internal iliac artery of the ischemic limb. On days 2 and 4, the same dose of drug (or placebo) was administered intravenously.
On day 30, all the measurements were repeated, and the animals were killed. The liver, kidney, and hindlimb muscles were weighed, and specimens of each were obtained for histological analysis. Muscle and kidney samples were also retrieved for microsphere analysis of tissue flow.
Lower Limb Calf Blood Pressure Index
Calf blood pressure was measured at days 0 and 30 in both hindlimbs with a Doppler flowmeter (model 1050, Parks Medical Electronics) and a cuff connected to a pressure manometer.37 40 The calf blood pressure index was defined for each rabbit as the ratio of systolic pressure of the ischemic limb to systolic pressure of the normal limb (×100).
Angiography and Doppler Guidewire Measurements
In addition to measurements performed at rest, endothelium-dependent and endothelium-independent responses were evaluated with intra-arterial administration of acetylcholine chloride and nitroprusside (Sigma) over 2 minutes via a constant infusion pump (1 mL/min). Each drug was administered into the iliac artery of the ischemic limb at a dose of 1.5 μg · min−1 · kg−1.
Doppler-Derived Blood Flow in the Ischemic Hindlimb
The luminal diameter of the internal iliac artery was determined by angiography at the site of the Doppler sample volume, at rest and after drug infusion, with an automated edge-detection system (Quantum 2000I; QCS) as previously described.37 Doppler-derived flow was calculated as QD=(πd2/4)(0.5×APV), where QD is Doppler-derived time-averaged flow, d is vessel diameter, and APV is time average of the spectral peak velocity.41 The mean velocity was estimated as 0.5×APV by assuming a time-averaged parabolic velocity profile across the vessel. The Doppler-derived flow calculated in this fashion has previously been validated in vivo.41 In the ischemic hindlimb, the internal iliac artery supplied blood flow for the entire hindlimb.37
Angiographic Analysis of Collateral Vessels
Morphometric analysis of collateral vessel development in the ischemic limb was performed by use of the 4-second angiograms recorded after injection of contrast media into the internal iliac artery. A grid overlay composed of 2.5-mm-diameter circles arranged in rows spaced 5 mm apart was placed over the angiogram at the level of the medial thigh. The number of contrast-opacified arteries crossing over circles and the total number of circles encompassing the medial thigh area were counted in single-blind fashion. An angiographic score was calculated as the ratio of crossing opacified arteries divided by the total number of circles in the ischemic thigh. This angiographic score reflects vascular density in the medial thigh.40 42
Measurement of Muscle, Renal, and Hepatic Blood Flow
Blood flow to hindlimb muscles and kidneys was evaluated by use of colored polystyrene spheres 15 μm in diameter.40 42 Quantification of microspheres is based on analysis of a linear relationship between spectrophotometrically determined absorbance of dye and number of microspheres. Thus, after the invasive measurements described above were completed, two different sets of 3×106 Dye-Trak colored microspheres (Triton Technology) were injected through a 3F Teflon catheter into the left ventricle at days 0 and 30. At each time point, the first injection of the set was performed under baseline conditions and the second after a nitroprusside infusion (15 μg · min−1 · kg−1) into the abdominal aorta. Each time, a reference blood sample was withdrawn with a syringe pump (Sage 351, Orion Research) to collect microspheres at a constant rate of 1.2 mL/min from a peripheral artery (central ear artery). When the animals were killed, a total of 14 tissue samples (2 g each) from seven different muscles (tensor fasciae latae, vastus lateralis, vastus medialis, adductor, semimembranosus, gastrocnemius, and tibialis anterior) in each hindlimb (ischemic and nonischemic) were retrieved. Samples from right and left kidneys and from the liver were also collected and used (1) to control for right-left homogeneity of blood microsphere content and (2) to evaluate the effect of administered angiogenic factors on blood flow to nonischemic tissues. After tissue and blood sample digestion with potassium hydroxide, microsphere extraction by filtering, and complete dye removal with dimethyl formamide, each sample was analyzed by a conventional spectrophotometer (model 8452A, Hewlett Packard).40 42 From the optical density (OD) measurements, the muscle perfusion expressed in mL · min−1 · 100 g−1 was calculated from the following equation: blood flow of muscle sample (mL · min−1 · g−1)=(OD of tissue sample/OD of reference blood sample)×[withdrawal rate of reference blood sample (mL/min)/weight of tissue sample (g)].
Blood flow of muscle sample (mL · min−1 · g−1)=(OD of tissue sample/OD of reference blood sample)×[withdrawal rate of reference blood sample (mL/min)/weight of tissue sample (g)].
Muscle blood flow in each hindlimb was expressed as the mean of the 14 samples. For each animal, muscle blood flow in the ischemic hindlimb was also expressed as a percentage of muscle blood flow in the nonischemic hindlimb.
The extent of vascularity was further examined by measurement of the number of capillaries in light microscopic sections retrieved from ischemic and nonischemic hindlimbs, as previously described.40 Tissue specimens were obtained as transverse sections from the adductor muscles and the semimembranosus muscles at the time of euthanization (day 30). Frozen sections were stained for alkaline phosphatase with an indoxyl tetrazolium method to detect capillary ECs and then were counterstained with eosin.43 A total of 20 different fields under an ×20 objective from the two muscles were randomly selected, and the number of capillaries were counted to determine the capillary density (mean number of capillaries per square millimeter). To ensure that analysis of capillary density was not overestimated because of muscle atrophy, capillary density was also evaluated as a function of the number of muscle fibers in the histological section.
For each hindlimb, muscles were divided into three groups: external thigh, internal thigh, and calf. At the time the animals were killed, the weight of each group (including the samples used for microsphere and histological analyses) was measured in each hindlimb. In each animal, an index of muscle atrophy was calculated for each group of muscles and for the entire hindlimb as follows: index of muscle atrophy=[1−(muscle weight in ischemic hindlimb/muscle weight in nonischemic hindlimb)]×100.
Histological Examination of Liver and Kidneys
To evaluate potential systemic effects of administered angiogenic factors, tissue samples from liver and kidney were systematically harvested, weighed, and cut into 5-mm slices to allow inspection for morphological abnormalities.
Results were expressed as mean±SEM. Statistical significance was evaluated by unpaired Student’s t test for comparisons between two means, ANOVA followed by Scheffé’s procedure for more than two means, and two-way ANOVA to test for interaction. A value of P<.05 was interpreted to denote statistical significance.
SF/HGF Induces VEGF mRNA and Protein Expression in Cultured HVSMCs
Administration of 25 ng/mL rhSF/HGF to cultured HVSMCs resulted in time-dependent induction of VEGF mRNA expression, as demonstrated by Northern blot analysis (Fig 1A⇓). Whereas the level of VEGF mRNA was low to nondetectable among quiescent SMCs maintained in low-serum medium, induction of VEGF mRNA was already pronounced 3 hours after stimulation, peaked at 6 hours, and persisted for at least 24 hours. Induction of VEGF mRNA expression was dose dependent, with a maximal effect between 25 and 50 ng/mL rhSF/HGF (Fig 1B⇓). To exclude nonspecific effects, BSA was added to SMCs and yielded no change in VEGF mRNA baseline expression (Fig 1C⇓). PMA was used as a positive control. RT-PCR analysis (Fig 2A⇓) disclosed equivalent induction of all three major VEGF isoforms, and semiquantitative evaluation of the VEGF165 PCR product confirmed that peak induction occurred 6 hours after stimulation with rhSF/HGF (2.9±0.13-fold increase compared with unstimulated controls). After 18 and 24 hours, VEGF165 mRNA decreased to 2.4±0.16 and 2.3±0.11-fold, respectively (Fig 2B⇓).
Western blotting confirmed the effect of rhSF/HGF on VEGF expression at the protein level (Fig 3⇓).
In Vitro Effects of rhSF/HGF and rhVEGF165, Alone and Together, on EC Migration and Proliferation
The impact of rhSF/HGF and rhVEGF165 on proliferation and migration of cultured HUVECs is shown in Fig 4⇓. No statistical difference was detected in the response of ECs to the two cytokines over a concentration range of 0.1 to 1000 ng/mL. For rhSF/HGF, maximal effect (expressed as percent of baseline) on HUVEC proliferation was 139.1±13.4% and for rhVEGF165, 158.2±7.6%, P=NS (Fig 4A⇓). Similarly, the maximal effect of rhSF/HGF and rhVEGF165 on HUVEC migration was 233.9±6.4% and 278.1±6.7%, respectively, P=NS (Fig 4B⇓).
The combined effect of rhSF/HGF and rhVEGF165 administered together was at least additive for both EC proliferation (P<.01, Fig 4A⇑) and migration (P<.0001, Fig 4B⇑). In addition, when submaximal concentrations (1, 10, 50, and 100 ng/mL) were used, a synergistic effect of the two cytokines was observed on EC migration (P<.0001). At 10 ng/mL, for example, rhSF/HGF+rhVEGF165250.3% of baseline versus an expected additive value of +178.9% (rhSF/HGF alone76.3% and rhVEGF165 alone102.6%; P<.0001).
Representative examples of the EC migration assay performed at a growth factor concentration of 10 ng/mL are shown in Fig 5⇓. Similar results were obtained with human microvascular ECs (data not shown).
Effect of rhSF/HGF and rhVEGF165 on Collateral Vessel Development in Ischemic Hindlimb
On the basis of the above results showing similar direct effects of both cytokines on ECs in vitro, we choose the same doses of rhSF/HGF and rhVEGF165 to evaluate their respective effects on angiogenesis in vivo, using the ischemic hindlimb model. We speculated that in addition to the direct effect of SF/HGF on EC proliferation and the synergistic effect, when administered with VEGF, on EC migration, the potential for SF/HGF to induce VEGF synthesis in SMCs could yield a superior effect, relative to rhVEGF165 alone, on angiogenesis in vivo.
The quantitative analysis of collateral blood vessel development in the medial thigh of rabbits with hindlimb ischemia is summarized in Fig 6A⇓. Before treatment (day 0), angiographic scores did not differ significantly among the experimental groups. By day 30, the angiographic score in both rhVEGF165- and rhSF/HGF-treated groups exceeded (P<.001) that of the control group (control=0.55±0.03; rhVEGF165=0.71±0.04; rhSF/HGF=0.89±0.04). Moreover, the angiographic score for the rhSF/HGF group was significantly higher than that of the rhVEGF165 group (P=.02).
Fig 7⇓ shows representative internal iliac angiograms recorded on day 30 from control and rhVEGF165- and rhSF/HGF-treated animals. In the rhVEGF165- and rhSF/HGF-treated groups, collateral artery development was more marked.
The impact of rhSF/HGF and rhVEGF165 on vascular density was also investigated at the capillary level, with light microscopy used to quantify capillaries/mm2 in tissue sections harvested at necropsy from the medial thigh muscles of the ischemic and nonischemic limbs. Morphometric analysis revealed that capillary density was significantly higher (P<.001) in the rhVEGF165- and rhSF/HGF-treated groups than in the control group (control=153±10/mm2; rhVEGF165=231±14/mm2; and rhSF/HGF=288±18/mm2). Capillary density in the rhSF/HGF group was also significantly higher than in the rhVEGF165 group (P=.02) (Fig 6B⇑). Analysis of the ratio of capillaries to muscle fiber yielded similar results (control=0.41±0.03; rhVEGF165=0.65±0.04; and rhSF/HGF=0.84±0.05).
Analysis of capillary density and ratio of capillaries to muscle fiber in sections from medial thigh muscles of the nonischemic limb showed no differences among groups (control=201±8; rhVEGF165=204±8; and rhSF/HGF=210±9; P=NS) (Fig 6B⇑).
Representative examples of histological sections stained for alkaline phosphatase to identify capillary density in the three experimental groups are shown in Fig 8⇓.
Effect of rhSF/HGF and rhVEGF165 on Pressure Perfusion Ratio
Calf blood pressure ratio was similar in all groups at day 0. By day 30, the blood pressure ratio had improved in all groups; in the rhVEGF165- and rhSF/HGF-treated groups, however, the blood pressure ratio was higher (P<.01) than in the control group (control=0.51±0.05; rhVEGF165=0.71±0.04; and rhSF/HGF=0.88±0.04). The blood pressure ratio observed in the rabbits treated with rhSF/HGF was significantly higher than that observed in rabbits treated with rhVEGF165 (P<.05).
Effect of rhSF/HGF and rhVEGF165 on Iliac Arterial Blood Flow
Blood flow parameters were measured from the Doppler-tipped guidewire positioned in the internal iliac artery of the ischemic limb at days 0 and 30 (Fig 9A⇓). At day 0, there were no differences among groups in resting or maximum flow. At day 30, flow at rest was higher in rhVEGF- (P<.05) and rhSF/HGF-treated animals (P≤.005) than in controls (control=17.9±0.9 mL/min; rhVEGF165=21.2±0.7 mL/min; and rhSF/HGF=23.9±1.2 mL/min). Maximum flow after nitroprusside infusion was significantly higher (P<.004) in both rhVEGF165- and rhSF/HGF-treated groups, compared with the maximum flow in the control group (control=36.1±2.3 mL/min; rhVEGF165=48.0±2.7 mL/min; and rhSF/HGF=59.4±3.7 mL/min). Maximum flow in the rhSF/HGF-treated animals was higher than in the rhVEGF165-treated animals (P=.02). The significant increase in maximum flow in the rhVEGF165- and rhSF/HGF-treated rabbits resulted in a significantly higher (P<.005) flow reserve than in the control group (control=2.01±0.06; rhVEGF165=2.26±0.06; and rhSF/HGF=2.47±0.05). The flow reserve in the rhSF/HGF-treated animals was also significantly higher than that of the rhVEGF165-treated animals (P=.016).
Effect of rhSF/HGF and rhVEGF165 on Muscle Blood Flow
Muscle blood flow, assessed with colored microspheres, was determined for days 0 and 30 (Fig 9B⇑). In each animal, the muscle blood flow in the ischemic limb was expressed as a percentage of the muscle flow measured in the nonischemic limb (see above). The homogeneity of microsphere distribution was verified by measurement of blood flow to the right and left kidneys (r=.99; slope=0.97±0.04, intercept=0.14±0.10).
At day 0, there were no differences among groups for muscle flow index measured at rest (control=65.6±3.9%; rhVEGF165=62.9±4.2%; and rhSF/HGF=64.6±3.7%; P=NS) or after administration of nitroprusside (control=45.4±3.0%; rhVEGF165=44.4±3.7%; and rhSF/HGF=39.8±2.9%; P=NS). At day 30, muscle flow index at rest was higher in rhVEGF165- and rhSF/HGF-treated animals (P<.002) than in controls (control=71.1±2.3%; rhVEGF165=87.6±3.7%; and rhSF/HGF=99.1±3.4%). The muscle flow index measured after nitroprusside was significantly higher (P<.0005) in both rhVEGF165- and rhSF/HGF-treated groups than in the control group (control=49.1±3.9%; rhVEGF165=66.9±3.7%; and rhSF/HGF=80.3±3.3%). The muscle flow index after nitroprusside was also higher for the rhSF/HGF-treated than in the rhVEGF165-treated animals (P=.01).
Effect of rhSF/HGF and rhVEGF165 on Muscle Atrophy
Muscle atrophy, evaluated by the muscle atrophy index (as described above), was significantly reduced in animals that received rhVEGF165 or rhSF/HGF (P<.01) compared with controls (control=28.67±2.47%; rhVEGF165=14.25±2.52%; and rhSF/HGF=4.27±2.65%). The effect observed in the rhSF/HGF group was significantly greater than that observed in the rhVEGF165 group (P=.013) (Table 1⇓).
Systemic Effects of rhSF/HGF and rhVEGF165 )
Although a trend toward greater weight gain was observed in rhVEGF165- and rhSF/HGF-treated animals, there were no significant differences among groups. Similarly, neither the weight nor the regional blood flow in liver or kidney differed among the groups. Blinded examination of microscopic sections from remote organ sites disclosed no pathological findings (Table 2⇓).
SF/HGF is a pleiotropic growth factor; its receptor, c-Met, is widely expressed on different cell populations, including both ECs and vascular SMCs.24 Cells transfected with full-length met cDNA respond to SF/HGF with a wide spectrum of biological effects.44 We therefore considered the possibility that SF/HGF could also upregulate VEGF expression in SMCs. In our first series of in vitro experiments, we could in fact demonstrate that SF/HGF upregulates VEGF expression in HVSMCs. Induction of VEGF mRNA by SF/HGF has previously been suggested by results obtained with implanted Matrigel.45 In that previous report, neither the identity of the cells responsible for the increase in VEGF mRNA nor the possibility of an increase in VEGF at the protein level was investigated. The present results establish that SMCs increase both VEGF mRNA and protein production after stimulation by SF/HGF. Furthermore, expression of all three principal VEGF isoforms was augmented.
The second series of in vitro experiments indicates that the effect of SF/HGF on EC migration and proliferation is similar to that induced by VEGF. When SF/HGF and VEGF are administered together, however, the mitogenic and chemotactic response of cultured ECs exceeds that achieved with either cytokine alone. Taken together, these findings suggest that in a milieu in which both ECs and SMCs are present, as is the case in vivo, SF/HGF may exert a potent combination of direct and indirect effects, including direct effects on ECs and indirect effects mediated via an increase in the production of VEGF. The consequence of such combined effects would be expected to be a potent means for stimulating angiogenesis in vivo.
The results of the in vivo experiments, in fact, confirm this notion. Experiments performed in the rabbit ischemic hindlimb model established that SF/HGF may be used therapeutically to augment collateral vessel development and blood flow to ischemic tissue. Previous investigators have evaluated the effect of SF/HGF in alternative in vivo assays, including Matrigel plug and cornea pocket assays.16 18 The present observations extend these previous results by showing that the neovasculature that develops in response to SF/HGF is functional, as indicated by the concomitant increase in arterial blood flow in the ischemic hindlimb, improvement in skeletal muscle perfusion, and a significant reduction in muscle atrophy.
The demonstration of a potent therapeutic effect of exogenously administered SF/HGF on angiogenesis in this ischemic hindlimb model is consistent with previous descriptions of transient expression in skeletal muscle during embryogenesis and after ischemic injury,14 in contrast to absent expression in normal adult muscle.13 14 The possibility that SF/HGF may play a key role in modulating angiogenesis that develops in response to ischemia is further supported by the recent finding that expression of SF/HGF is enhanced after myocardial ischemia and reperfusion.20
The present in vivo studies also indicate that SF/HGF-induced neovascularity exceeds that achieved with VEGF. Despite a similar dose-response relationship in vitro and the same schedule of administration (500 μg×3) in vivo, angiographic as well as histological evidence of angiogenesis in animals receiving SF/HGF was more marked than that in those treated with VEGF. These in vivo results extend our in vitro findings by suggesting that the ability of SF/HGF to induce angiogenesis by direct effects on EC proliferation and migration may be potentiated in vivo by the ability to induce angiogenesis indirectly by upregulating one or more cytokines, as shown here for VEGF.
Direct comparison of SF/HGF and VEGF carried out in these in vivo studies suggests that EC specificity is not necessarily an advantageous feature for cytokines designed to promote angiogenesis. In fact, from the standpoint of bioactivity, the in vivo results suggest that the pleiotropic effects of nonspecific EC mitogens such as SF/HGF may constitute an important means of optimizing neovascularization of ischemic tissues. The potential liability of pleiotropic effects associated with a non–EC-specific mitogen is the risk of stimulating superfluous cell populations. In this regard, it is noteworthy that no pathological consequences were observed with the dose and protocol of administration of SF/HGF used for these experiments. In particular, body weight was not altered and no evidence of angiogenesis or increased blood flow was observed in nonischemic tissues. In addition, whereas previous reports have suggested that SF/HGF administration is associated with a growth-promoting effect on liver and kidney,15 46 47 48 no such effect was observed in our study. It is important to point out, however, that in these previous reports, growth-promoting effects of SF/HGF were demonstrated in the case of injured organs15 46 47 and/or continuous administration of the angiogenic cytokine.47 48
The absence of accompanying pathological effects in the present series of experiments as well as in other reported work46 is consistent with the suggestion that the in vivo response to SF/HGF is facilitated by a priming stimulus. Indeed, similar site specificity has been demonstrated for basic fibroblast growth factor.7 The above-cited work of Ono et al20 suggested that local upregulation of the c-Met receptor acts to regulate the extent of SF/HGF bioactivity. A similar paradigm has been described for VEGF, in which site specificity appears to be due in part to paracrine upregulation of the KDR receptor by factors derived from hypoxic myocytes.49 It is also worth noting that previous studies have established that stimulation of the c-Met receptor in vascular SMCs does not evoke a proliferative response.24 49 Moreover, the risk of potentially unwanted effects in noninjured organs may be avoided by a protocol of restricted administration. In this regard, it is fortuitous that the time interval required for neovascularization of ischemic tissues in animals3 39 50 and patients9 10 is typically <30 days.
Selected Abbreviations and Acronyms
|HUVEC||=||human endothelial cell|
|HVSMC||=||human vascular smooth muscle cell|
|RT-PCR||=||reverse transcription–polymerase chain reaction|
|SF/HGF||=||scatter factor/hepatocyte growth factor|
|SMC||=||smooth muscle cell|
|VEGF||=||vascular endothelial growth factor|
This study was supported in part by grants HL-40518, HL-53354, and HL-57516 and an Academic Award in Vascular Medicine (HL-02824) from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md. Dr Van Belle is the recipient of a Fellowship from the French government “Bourse Lavoisier.”
Drs Van Belle and Witzenbichler contributed equally to this work.
- Received August 26, 1997.
- Revision received October 30, 1997.
- Accepted November 13, 1997.
- Copyright © 1998 by American Heart Association
Folkman J, Shing Y. Angiogenesis. J Biol Chem. 1992;267:10931–10934.
Takeshita S, Zheng LP, Brogi E, Kearney M, Pu LQ, Bunting S, Ferrara N, Symes JF, Isner JM. Therapeutic angiogenesis: a single intra-arterial bolus of vascular endothelial growth factor augments revascularization in a rabbit ischemic hindlimb model. J Clin Invest. 1994;93:662–670.
Banai S, Jaklitsch MT, Shou M, Lazarous DF, Scheinowitz M, Biro S, Epstein SE, Unger EF. Angiogenic-induced enhancement of collateral blood flow to ischemic myocardium by vascular endothelial growth factor in dogs. Circulation. 1994;89:2183–2189.
Harada K, Grossman W, Friedman M, Edelman ER, Prasad PV, Keighlcy CS, Manning WJ, Selke FW, Simons M. Basic fibroblast growth factor improves myocardial function in chronically ischemic porcine hearts. J Clin Invest. 1994;94:623–630.
Baumgartner I, Pieczek A, Manor O, Blair R, Isner JM. Evidence of therapeutic angiogenesis in patients with critical limb ischemia after intramuscular phVEGF165 gene transfer. Circulation. 1997;96(suppl I):I-32. Abstract.
Sonnenberg E, Meyer D, Weidner KM, Birchmeier C. Scatter factor/hepatocyte growth factor and its receptor, the c-met tyrosine kinase, can mediate a signal exchange between mesenchyme and epithelia during mouse development. J Cell Biol. 1993;123:223–235.
Jennische E, Ekberg S, Matejka GL. Expression of hepatocyte growth factor in growing and regenerating rat skeletal muscle. Am J Physiol. 1993;265:C122–C128.
Igawa T, Matsumoto K, Kanda S, Saito Y, Nakamura T. Hepatocyte growth factor may act as a renotropic factor for regeneration in rats with acute renal injury. Am J Physiol. 1993;265:F61–F69.
Grant SD, Kleinman HK, Goldberg ID, Bhargava MM, Nickoloff BJ, Kinsella JL, Polverini P, Rosen EM. Scatter factor induces blood vessel formation in vivo. Proc Natl Acad Sci U S A. 1993;90:1937–1941.
Bussolino F, DiRenzo MF, Ziche M, Bocchietto E, Olivero M, Naldini L, Gaudino G, Tamagnone L, Coffer A, Comoglio PM. Hepatocyte growth factor is a potent angiogenic factor which stimulates endothelial cell mobility and growth. J Cell Biol. 1992;119:629–641.
Sato Y, Okamura K, Morimoto A, Hamanaka R, Hamanaguchi K, Shimado T, Ono M, Kohno K, Sakata T, Kuwano M. Indispensable role of tissue-type plasminogen activator in growth factor-dependent tube formation of human microvascular endothelial cells in vitro. Exp Cell Res. 1993;204:223–229.
Ono K, Matsumori A, Shioi T, Furukawa U, Sasayama S. Enhanced expression of hepatocyte growth factor/c-Met by myocardial ischemia and reperfusion in a rat model. Circulation. 1997;95:2552–2558.
Brogi E, Wu T, Namiki A, Isner JM. Indirect angiogenic cytokines upregulate VEGF and bFGF gene expression in vascular smooth muscle cells, while hypoxia upregulates VEGF expression only. Circulation. 1994;90:649–652.
Naka D, Ishii T, Yoshiyama Y, Miyazzawa K, Hara H, Hishida T, Kidamura N. Activation of hepatocyte growth factor by proteolytic conversion of a single chain form to a heterodimer. J Biol Chem. 1992;267:20114–20119.
Zioncheck TF, Richardson L, Liu J, Chang L, King KL, Bennett GL, Fugedi P, Chamow SM, Schwall RH, Stack RJ. Sulfated oligosaccharides promote hepatocyte growth factor association and govern its mitogenic activity. J Biol Chem. 1995;270:16871–16878.
Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins. J Clin Invest. 1973;52:2745–2756.
Tischer E, Mitchell R, Hartmann T, Silva M, Gospodarowicz D, Fiddes J, Abraham J. The human gene for vascular endothelial growth factor: multiple protein forms are encoded through alternative exon splicing. J Biol Chem. 1991;266:11947–11954.
Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual. Plainview, NY: Cold Spring Harbor Laboratory Press; 1989.
Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KG, Ehrlich HA. Primer-directed enzymatic amplification of DNA with thermostable DNA polymerase. Science. 1988;239:487–491.
Wang AM, Doyle MV, Mark DF. Quantification of mRNA by the polymerase chain reaction. Proc Natl Acad Sci U S A. 1989;86:9717–9721.
Bauters C, Asahara T, Zheng LP, Takeshita S, Bunting S, Ferrara N, Symes JF, Isner JM. Physiologic assessment of augmented vascularity induced by VEGF in ischemic rabbit hindlimb. Am J Physiol. 1994;267:H1263–H1271.
Takeshita S, Pu L-Q, Zheng L, Ferrara N, Stein LA, Sniderman AD, Isner JM, Symes JF. Vascular endothelial growth factor induces dose-dependent revascularization in a rabbit model of persistent limb ischemia. Circulation. 1994;90(suppl II):II-228-II-234.
Walter DH, Hink U, Asahara T, Van Belle E, Horowitz J, Tsurumi Y, Vandlen R, Heinsohn H, Keyt B, Ferrara N, Symes JF, Isner JM. The in vivo bioactivity of vascular endothelial growth factor/vascular permeability factor is independent of N-linked glycosylation. Lab Invest. 1996;74:546–556.
Doucette JW, Corl PD, Payne HM, Flynn AE, Goto M, Nassi M, Segal J. Validation of a Doppler wire for intravascular measurements of coronary artery flow velocity. Circulation. 1992;85:1899–1911.
Kowallik P, Schulz R, Guth BD, Schade A, Paffhausen W, Gross R, Heusch G. Measurement of regional myocardial blood flow with multiple colored microspheres. Circulation. 1991;83:974–982.
Weidner KM, Sachs M, Birchmeier W. The Met receptor tyrosine kinase transduces motility, proliferation, and morphogenic signals of scatter factor/hepatocyte growth factor in epithelial cells. J Cell Biol. 1993;121:145–154.
Silvagno F, Follenzi A, Arese M, Prat M, Giraudo E, Gaudino G, Camussi G, Comoglio PM, Bussolino F. In vivo activation of met tyrosine kinase by heterodimeric hepatocyte growth factor molecule promotes angiogenesis. Arterioscler Thromb Vasc Biol. 1995;15:1857–1865.
Roos F, Ryan AM, Chamow SM, Bennett GL, Schwall RH. Induction of liver growth in normal mice by infusion of hepatocyte growth factor/scatter factor. Am J Physiol. 1995;268:G380–G386.
Tabata H, Silver M, Isner JM. Arterial gene transfer of acidic fibroblast growth factor for therapeutic angiogenesis in vivo: critical role of secretion signal in use of naked DNA. Cardiovasc Res. 1997;35:470–479.