The FASEB Journal • Research Communication
Differential roles of PDGFR-␣ and PDGFR- in angiogenesis and vessel stability Junhang Zhang,*,†,1 Renhai Cao,‡,1 Yun Zhang,§ Tanghong Jia,储 Yihai Cao,‡,2 and Eric Wahlberg*,2 *Division of Vascular Surgery, Department of Molecular Medicine and Surgery, and ‡Laboratory of Angiogenesis Research, Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, Stockholm, Sweden; §Qilu Hospital and 储Clinical Medical College, Shandong University, Jinan, Shandong, China; and †Department of Developmental Biology, China Medical University; Liaoning Cancer Hospital and Institute, Shenyang, China Preclinical and clinical evaluations of individual proangiogenic/arteriogenic factors for the treatment of ischemic myocardium and skeletal muscle have produced unfulfilled promises. The establishment of functional and stable arterial vascular networks may require combinations of different angiogenic and arteriogenic factors. Using in vivo angiogenesis and ischemic hind-limb animal models, we have compared the angiogenic and therapeutic activities of fibroblast growth factor 2 (FGF-2) in combinations with PDGF-AA and PDGF-AB, two members of the platelet-derived growth factor (PDGF) family, with distinct receptor binding patterns. We show that both PDGF-AA/FGF-2 and PDGF-AB/FGF-2 in combinations synergistically induce angiogenesis in the mouse cornea. FGF-2 upregulates PDGFR-␣ and - expression levels in the newly formed blood vessels. Interestingly, PDGF-AB/ FGF-2, but not PDGF-AA/FGF-2, is able to stabilize the newly formed vasculature by recruiting pericytes, and an anti-PDGFR- neutralizing antibody significantly blocks PDGF-AB/FGF-2-induced vessel stability. These findings demonstrate that PDGFR- receptor is essential for vascular stability. Similarly, PDGF-AB/FGF-2 significantly induces stable collateral growth in the rat ischemic hind limb. The high number of collaterals induced by PDGF-AB/FGF-2 leads to dramatic improvement of the paw’s skin perfusion. Immunohistochemical analysis of the treated skeletal muscles confirms that a combination of PDGF-AB and FGF-2 significantly induces arteriogenesis in the ischemic tissue. A combination of PDGF-AB and FGF-2 would be optimal proangiogenic agents for the treatment of ischemic diseases.—Zhang, J., Cao, R., Zhang, Y., Jia, T., Cao, Y., Wahlberg, E. Differential roles of PDGFR-␣ and PDGFR- in angiogenesis and vessel stability. FASEB J. 23, 153–163 (2009)
ABSTRACT
Key Words: neovascularization 䡠 ischemia 䡠 growth factor 䡠 receptor Preclinical and clinical evaluations of individual proangiogenic/arteriogenic factors for the treatment of ischemic myocardium and skeletal muscle have pro0892-6638/09/0023-0153 © FASEB
duced unfulfilled promises. The idea of promoting angiogenesis and arteriogenesis for the treatment of ischemic diseases, including coronary arterial diseases (CADs) and various periphery arterial diseases (PADs) was proposed more than 30 years ago (1). Initially, it was thought that delivery of several proangiogenic factors, including fibroblast growth factor 2 (FGF-2) and vascular endothelial growth factor (VEGF) would be a straightforward approach for ischemic disease therapy. Encouraged by this idea, numerous preclinical studies have shown that single proangiogenic factors such as FGF-1, FGF-2, and VEGF could significantly stimulate neovascularization in ischemic myocardium and skeletal muscles (2–5). Inspired by these preclinical studies, several early small open trials for the treatment of patients with CAD and PAD have claimed to be successful (2). However, after more than a decade of clinical efforts, all double-blinded, randomized, and placebo-controlled trials have failed to conclusively demonstrate a clinical benefit (6). These negative clinical data have raised several important issues about our understanding of molecular mechanisms of angiogenesis and arteriogenesis and translation of this preclinical information into clinical practice. Despite their angiogenic activity in animal models, selection of proangiogenic molecules for clinical trials lacks biological and mechanistic rationales. Clinical designs for these trials have not always taken into the consideration the basic mechanisms of angiogenesis and arteriogenesis and blood vessel stability. While angiogenesis is usually referred to as the growth of new blood vessels (more specifically, sprouting of endothelial cells from preexisting vessels), arteriogenesis is probably 1
These authors contributed equally to this work. Correspondence: E.W., Division of Vascular Surgery, Department of Molecular Medicine and Surgery, N1:06, Karolinska Institutet, SE-17176, Stockholm, Sweden. E-mail:
[email protected]; Y.C., Laboratory of Angiogenesis Research, Microbiology, Tumor and Cell Biology, Nobels Va¨g16, Karolinska Institutet, SE-17177, Stockholm, Sweden. E-mail:
[email protected] doi: 10.1096/fj.08-113860 2
153
confined to maturation of blood vessels or the establishment of collateral conduits (7). For therapeutic angiogenesis, both processes of angiogenesis and arteriogenesis are crucial for establishment of functional vascular networks in the ischemic tissues. For arteriogenesis, it is still unknown whether the growth of new arteries de novo or remodeling or enlargement of preexisting collaterals is the key process. Nevertheless, both de novo growth and remodeling of collaterals target the compartment of vascular smooth muscle cells (VSMCs), which also play a pivotal role in vascular development (8). In fact, our recent study demonstrates that a combination of FGF-2 and platelet-derived growth factor (PDGF) -BB not only synergistically stimulates angiogenesis but also stabilizes the newly formed arterial vascular networks (8 –10). In the present work, we have systematically compared the angiogenic/arteriogenic and therapeutic efficacy of PDGF-AA and PDGF-AB, two other members of the PDGF family, in combinations with FGF-2 in the mouse corneal angiogenesis and in the ischemic hind-limb animal models. Since these members of the PDGF family bind to different receptors, our study also provides important information understanding the roles of individual PDGF receptors in mediating arteriogenic signals and vascular stability. Thus, these results might provide important guidelines and rationales for selections of angiogenic/arteriogenic factors for future clinical trials.
MATERIALS AND METHODS Reagents and animals Recombinant PDGF-AA, PDGF-AB, PDGF-BB, and FGF-2 were obtained from PeproTech Inc. (Rocky Hill, NJ, USA) and Pharmacia & Upjohn (Milan, Italy). Male 5– 6-wk-old C57BL/6 mice (MTC; Karolinska Institute, Stockholm, Sweden) were used for the corneal pocket assay and acclimated in groups of 6 or less per cage. For the rat hind-limb ischemia model, Sprague-Dawley male rats (B&K Universal, Sollentuna, Sweden) weighing 280 ⫾ 20 g were used. Before all procedures, mice were anesthetized by injection of a mixture of dormicum and hypnorm in a ratio of 1:1 and were killed with a lethal dose of CO2 at the end of the experiments. The rats were anesthetized with isoflurane (Abbott Scandinavia AB, Stockholm, Sweden) using a special inhalation device (G.H. Medic Instrument, Stockholm, Sweden). Induction was made by 4% isoflurane in a glass box, and anesthesia was maintained by 2.5% isoflurane throughout the surgical procedures. All animal studies were reviewed and approved by the animal care and use committee of the North Stockholm Animal Board. Mouse corneal micropocket assay The mouse corneal assay was performed as described previously (11). Micropellets (0.35⫻0.35 mm) of sucrose aluminum sulfate (Bukh Meditec, Copenhagen, Denmark) were coated with hydron polymer type NCC containing 160 ng of PDGF-AA, -AB, or -BB; or 80 ng of FGF-2; or a combination of 80 ng of FGF-2 with 160 ng of PDGF-AA or 154
Vol. 23
January 2009
-AB; or a combination of 40 ng FGF-2 with 160 ng of PDGF-BB. Micropellets were surgically implanted into each micropocket in the mouse eyes. Corneal neovascularization was examined and quantified on day 5 after pellet implantation. Vascularization areas were calculated by measuring vessel length and clock hours (the circumferential area of neovascularization if the eye is considered as a clock). For vascular stability analysis, corneal neovascularization was examined on days 5, 12, 24, and 70 after pellet implantation. The corneal blood vessels were quantified from the average of 6 or 7 eyes at days 12, 24, and 70 by accounting the numbers of vessels across the midline of the circumferential surface of each cornea. Rat ischemic hind-limb model The rat ischemic hind limb model was carried out using a two-stage procedure on the left hind limb, whereas the right hind limb serves as a control. During the first operation, a midline laparotomy provided access with an aid of a dissecting microscope (Carl Zeiss, Oberkochen, Germany) to all arterial branches originating from the aorta distal to the renal arteries, including the spermatic, left lumbar, ileolumbar, inferior mesenteric, andcaudal arteries, and all branches from the left iliac artery on the left side down to the inguinal ligament. All of these branches were carefully ligated using 6.0 resorbable sutures. At day 5, rats were again anesthetized, and the second operation was performed. Through a left inguinal incision, the femoral artery was ligated at a position close to the origin of the superficial epigastric artery, which was also subsequently ligated. An analgesic (Temgesic, Schering-Plough Europe, Brussels, Belgium) was injected subcutaneously using a dose of 0.01 mg/ml/100 g every 12 h for 2 days after the first operation and 7 days after the second operation, respectively. The operated animals were randomly divided into 6 groups (6 rats/group) for treatment with FGF-2, PDGF-AB, PDGFAA, FGF-2/PDGF-AB, FGF-2/PDGF-AA, or phosphate-buffered saline (PBS). During the second operation, these individual growth factors or the combinations (0.8 g FGF-2, 1.6 g PDGF-AB, 1.6 g PDGF-AA, or 0.8 g FGF-2/1.6 g PDGF-AB, 0.8 g FGF-2/1.6 g PDGF-AA) in sucrose sulfate/ hydron slow-release polymers were implanted into intramuscular pockets near the ligation site. After completion of the operation, growth factors (1.5 g FGF-2, 3.0 g PDGF-AB, 3.0 g PDGF-AA, or 1.5 g FGF-2/3.0 g PDGF-AB, 1.5 g FGF-2/3.0 g PDGF-AA) in 400 l PBS were injected into 3 sites close to the femoral ligation site and the treatment continued every other day for 12 days. Whole-mount immunostaining and confocal analysis Growth factor-implanted mouse eyes were enucleated at day 5 or 9 after implantation. The corneal tissue was dissected and flattened before fixation with 3% paraformaldehyde (PFA) overnight. The tissues were digested with proteinase K (20 g/ml), followed by staining overnight at 4°C with a mixture of a rat anti-mouse CD31 monoclonal antibody (Pharmingen, San Diego, CA, USA) and a rabbit anti-NG2 polyclonal antibody (Chemicon, Temecula, CA, USA). After rigorous rinsing, blood vessel endothelial cells and pericytes were detected with secondary antibodies coupled to fluorescent (Chemicon). After washing, slides were mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA) and analyzed under a confocal microscope (Zeiss Confocal LSM510 microscope; Carl Zeiss, Jena, Germany). The images were further analyzed with Adobe Photoshop CS2 (Adobe, San Diego, CA, USA). Pericyte recruitment index
The FASEB Journal
ZHANG ET AL.
was calculated by associating the NG2 pericyte-positive signals with CD31 positive signals from randomized 6 –10 fields of each sample. In situ hybridization In situ hybridization was carried out according to a standard method using radiolabeled oligonucleotide probes and high stringency conditions. Two probes complementary to PDGFR-␣ (nucleotides 423– 470 and 3083–3130) and two probes complementary to PDGFR- (nucleotides 946 –996 and 2610 –2657) were used. All probes were used separately and did not match any known sequence in GenBank except those of the intended genes. A control 50-mer random probe, not complementary to any sequence deposited in GenBank, was also used. Following 3⬘ end-labeling with [33P]dATP (NEN Dupont, Boston, MA, USA) by terminal deoxynucleotidyl transferase (Amersham, Arlington Heights, IL, USA), probes were purified (QIAquick Nucleotide Removal Kit Protocol, Qiagen, Chatsworth, CA, USA). Corneal histological slides at day 5 after implantation were incubated overnight (42°C) with 0.1 ml hybridization cocktail, containing 50% formamide, 4⫻ SSC (0.15 M NaCl, 15 mM sodium citrate, pH 7.0), 1⫻ Denhardt’s solution, 1% Sarcosyl, 0.02 M Na3PO4 (pH 7.0), 10% dextran sulfate, 0.06 M dithiothreitol (DTT), 0.1 mg/ml sheared salmon sperm DNA, and hot probe. Slides were then rinsed 4 times (45 min) in 1⫻ SSC at 60°C and allowed to adjust to room temperature during a fifth rinse in 1⫻ SSC. Further rinsing was carried out in distilled water and increasing concentrations of ethanol. Air-dried slides were then dipped in emulsion (Kodak NTB2, diluted 1:1 with water; Eastman Kodak, Rochester, NY, USA). After 5 wk of exposure, slides were developed, counterstained with cresyl violet, and mounted (Entellan; Merck, Rahway, NJ, USA). The control probe was hybridized and processed together with the other probes and gave rise to no specific pattern of hybridization signals in the mouse tissue. Specific labeling was confirmed by similar expression patterns revealed by two probes (complementary to different parts of the mRNA) for each PDGFR-␣ and PDGFR-. Detection of positive autoradiographic signals was based on serial observations of adjacent sections from each tissue specimen and accumulation of silver grains in the emulsion above specific cells and tissues identified by the staining procedures. Only cells over which silver grain accumulation was clearly above surrounding background levels and could be confirmed by both dark-field and high-magnification bright-field were regarded as positive. Blockage of PDGFR- receptor Using the same mouse corneal assay, a micropellet of sucrose aluminum sulfate coated with hydron polymer containing 160 ng of PDGF-AB and 80 ng of FGF-2 was surgically implanted into a mouse corneal micropocket. After operation, PBS or 800 g/mouse of a neutralizing PDGFR- antibody (kindly provided by Dr. Zhenping Zhu, Imclone, New York, NY, USA) was injected i.p. The treatment continued every other day for 12 days. The corneal neovascularization was examined on day 19 after pellet implantation. Angiography At day 42 after the second operation, rats were anesthetized, and a midline incision was performed to expose and place a ligature around the abdominal aorta. It was then cannulated with a polyethylene tube (PE-50; Becton Dickinson, Sparks, MD, USA) distal to the ligature and the tip of catheter placed just above the bifurcation. The rats were positioned supine on MECHANISMS OF ANGIOGENESIS BY PDGFS AND FGFS
an image plate, directly on the collimator of a mobile X-ray system (Siemens, Erlangen, Germany). Papaverin (Recip AB, Stockholm, Sweden) of 1.0 ml volume (4 mg/ml) was injected followed by 0.5– 0.8 ml of iodixanol (Visipaque, Amersham Health AB, Stockholm, Sweden). An image was obtained after injection of the contrast solution. The developed films were examined to identify collateral vessels that bridged the ligated femoral artery. It was quantified in the left thigh by direct counting of the number of arteries crossing a line, drawn vertically across the midthigh. Immunohistochemistry After sacrificing the animals at day 42 after the second operation, muscle tissues from the ligated and treated areas of the ischemic and control hind limbs were dissected. Tissue samples were fixed with 3% paraformaldehyde overnight at 4°C, dehydrated, embedded in paraffin, and cut as 5-m cross-sections. The muscle tissue was then stained with a monoclonal antibody against smooth-muscle alpha-actin (Neomarkers, Fremont, CA, USA). After deparaffinization, hydrogen peroxidase (3% diluted in distilled water) was added for 5 min, followed by 3 washes in PBS for a total of 15 min. Blocking solution (5% goat serum in PBS) was applied for 1 h at room temperature (RT), and the slides were incubated overnight at 4°C with mouse monoclonal antibody against rat smooth muscle alpha-actin at 1:200, diluted in 1% bovine serum albumin (BSA) in PBS. Incubation was followed by three 5-min washes in PBS. A secondary biotinylated goat anti-mouse antibody was added at 1:200 in 1% BSA in PBS for 1 h at RT, followed by 3 washes in PBS for a total of 15 min. Labeled avidin-conjugated peroxidase complex (Vectastain ABC kit; Vector Laboratories) was used for color development according to the manufacturer’s recommendations for 30 min at RT. After rinsing in PBS, DAB (3,3⬘-diaminobenzidine) peroxidase substrate kit (Vector Laboratories) was added for 7 min to localize the immune complexes. The sections were counterstained with Mayer’s hematoxylin (Vector Laboratories) for 3 min and mounted with Mountex (Histolab products AB; Stockholm, Sweden). Omission of the first antibody was used as a negative control. Analysis of tissue samples was conducted using a Nikon epifluorescence microscope (Eclipse E800; Nikon, Yokohama, Japan). Collateral analysis Collateral numbers and sizes were measured from alpha-actin positively stained vessels, excluding the ones with a venous shape, at ⫻10. Collaterals with a diameter greater than 10 m were counted and measured. A total number of collaterals were counted manually from 10 different fields selected from two different sections from each sample in a masked fashion. The results were expressed as the number of collateral arteries per field of view. Laser-Doppler perfusion imaging The laser-Doppler perfusion imager (MoorLDI-VR, Moor Instruments Ltd, Axminster, UK) was used to assess the limb perfusion. At day 42 after the second operation, perfusion in the ischemic and control hind paws of each animal was measured. The laser-Doppler source was mounted on a desktop stand, and a laser beam scanned the tissue using a moving mirror. The laser beam reflected from moving red blood cells in nutritional capillaries, as well as in arterioles and venules, was recorded and processed to provide a flux value. The information was color coded to provide a map of tissue perfusion. Poor or no perfusion was displayed as dark 155
blue, and the highest perfusion level was displayed as red to white. Mean flux values were calculated using the Moor LDI V3.09 image processing software. Improvements in blood perfusion were calculated as the average of the percentages of perfusion in the ischemic paw and the control paw. The hemoglobin oxygen saturation was measured with a pulse oximeter (Nonin 8500V; Nonin Medical, Inc., Plymouth, MN, USA) using a flexible sensor on a forepaw. Supplementary oxygen was delivered on demand through a face mask, maintaining the saturation at or above 90%. The rats were placed on a 37°C heating pad to reduce heat loss during measurements. Statistics Data are presented as means ⫾ sd. Statistical evaluation of the results was made by two-tailed Student’s t test and MannWhitney U test using Statview software (Adept Scientific Inc., Acton, MA, USA) and Microsoft Excel (Microsoft, Redmond, WA, USA). Differences with a value of P ⬍ 0.05 were considered significant.
RESULTS Comparison of angiogenic synergism among combinations of PDGF-AA/FGF-2, PDGF-AB/FGF-2, and PDGF-BB/FGF-2 To compare synergistic effects of combinations of FGF-2 with PDGF-AB, PDGF-AA, or PDGF-BB, we chose the mouse corneal angiogenesis model that has several advantages over other angiogenesis assays to study angiogenic functions of various combinatorial molecules. Because of the avascular nature of the corneal tissue, all blood vessels induced by a given factor are newly formed blood vessels. This model is particularly important for the present work in the aspect of studying vascular stability. It allowed us to monitor the fate of newly formed blood vessels in the corneal tissue without any mechanical disruption of tissues or blood vessels. Thus, this assay system provided us a unique opportunity to systematically compare angiogenic synergisms and vascular stability of various molecules alone or in combinations in mice. At day 5 after pellet implantation, all four individual growth factors, including PDGF-AA, PDGF-AB, PDGFBB, or FGF-2, were able to induce corneal neovascularization (Fig. 1B–E). It appeared that FGF-2 displayed a similar potent angiogenic activity as PDGF-AB or PDGF-BB when these angiogenic effects were quantified as vessel areas (Fig. 1I ). In contrast, PDGF-AA exhibited only weak angiogenic activity as compared with PDGF-AB, PDGF-BB, or FGF-2 (Fig. 1B–E, I ). Interestingly, combinations of FGF-2 with PDGF-AB or PDGF-BB induced robust angiogenic responses in the cornea, which was quantified as a synergistic effect (more than additive to two individual factors, P ⬍ 0.001, Fig. 1G–J ). Similarly, a combination of FGF-2 with PDGF-AA also induced a significantly synergistic effect of corneal neovascularization (Fig. 1F, I). It seemed that PDGF-AB or PDGF-BB displayed more 156
Vol. 23
January 2009
Figure 1. Angiogenic synergisms between FGF-2 and PDGF-AA, PDGF-AB, or PDGF-BB. Slow-release polymer containing PDGF-AA (B), PDGF-AB (C), PDGF-BB (D) FGF-2 (E), PDGFAA/FGF-2 (F), PDGF-AB/FGF-2 (G), PDGF-BB/FGF-2 (H), or polymer alone (A) was implanted into each mouse corneal micropocket. At day 5 after implantation, corneal neovascularization was examined and photographed for each cornea. The corneal neovascularization was quantitatively measured as clock hours and vessel lengths, and vascularization areas were calculated (I, J ). Arrows point to the implanted micropellets.
The FASEB Journal
ZHANG ET AL.
potent angiogenic synergism with FGF-2 than the PDGF-AA with FGF-2 together (Fig. 1I, J ). However, comparison of angiogenic synergy between PDGF-AB/ FGF-2 and PDGF-BB/FGF-2 did not reveal a statistical significance, although there is a trend of difference between the two groups (P⫽0.06). These findings demonstrate that all three PDGFs, including PDGF-AA, PDGF-AB, and PDGF-BB, produce angiogenic synergies with FGF-2 in the mouse corneal angiogenesis model, albeit the latter two combinations are significantly more potent than the former combination. Combinations of PDGF-AB or PDGF-BB, but not PDGF-AA, with FGF-2 induced long-lasting vascular networks We next studied the vascular stability effect of single or combinatorial angiogenic factors in this mouse corneal model. Although at day 5 after pellet implantation, all three combinations of PDGF-AA/FGF-2, PDGF-AB/FGF-2, and PDGF-BB/FGF-2 produced significantly synergistic effects on corneal neovascularization, the vascular network induced by these combinatorial factors seemed to be premature and primitive at this early time point (Fig. 2A, E, I ). Particularly, the combinations of PDGF-AB and PDGF-BB with FGF-2 induced a rather primitive and disorganized vasculature with the fusion of multiple microvessels into a vascular plexus in the leading front (Fig. 2A, E, I; black arrows). At day 12 after pellet implantation, the primitive vascular networks underwent dramatic remodeling with the appearance of most well-structured vessels remaining in the corneal tissue (Fig. 2B, F, J). Surprisingly, the majority of vascular networks induced by the combinations of PDGF-AB/FGF-2 and PDGF-BB/FGF-2 remained in the cornea and continued to grow into a mature vascular network that covered almost the entire cornea (Fig. 2F, J). In contrast, the majority of PDGF-AA/FGF-2-induced vascular networks vanished, and only a few blood vessels remained in the cornea by day 12 (Fig. 2B). Long-term follow-up of these vascular networks showed that the PDGF-AB/FGF-2- or PDGF-BB/FGF2-induced vasculature permanently remained in the corneal tissue at day 70 (Fig. 2H, L). By day 70, the PDGF-AA/FGF-2-induced vessels nearly completely regressed (Fig. 2D). Time course and quantification analyses showed that PDGF-BB plus FGF-2 exhibited more potent effect on vascular stability than the combination of PDGF-AB and FGF-2 did (Fig. 2M–O). The other interesting finding was that at day 12 after implantation, most of the implanted pellets had disappeared from the corneal micropockets, probably due to high levels of vascularization. Even without the angiogenic supply, PDGF-AB/FGF-2- and PDGF-BB/FGF-2induced vessels continued to grow and were permanently stabilized in the cornea. These data indicate that a combination of PDGF-AB or PDGF-BB, but not PDGFMECHANISMS OF ANGIOGENESIS BY PDGFS AND FGFS
Figure 2. Stability of corneal blood vessels. A slow-release polymer containing PDGF-AA/FGF-2 (A–D), PDGF-AB/ FGF-2 (E–H) or PDGF-BB/FGF-2 (I–L) was implanted into each mouse corneal micropocket. Corneal neovascularization was examined and photographed at the indicated time points. Yellow arrows point to the implanted micropellets; black arrows point to the leading edges of corneal neovascularization on day 5 after pellet implantation. Corneal vessels were quantified at different time points (M–O). Asterisks indicate positions of pellets in those corneas that lost implanted pellets.
AA, with FGF-2 is able to induce long-lasting blood vessels in the cornea. Significant recruitment of pericytes by PDGF-AB/FGF-2 Because pericytes are the primary vascular cell types involved in vascular stability of newly formed blood vessels, we studied pericyte coating on the newly formed corneal vessels induced by various factors. Significant recruitments of pericytes to the angiogenic vessels induced by various growth factors occurred at the early time point of day 5 after implantation (Fig. 3A–E ). Intriguingly, the number of pericytes was generally decreased in all angiogenic factor-stimulated corneas at day 9 after implantation, suggesting that significant remodeling occurred. Interestingly, pericytes were differentially distributed in a subset of corneal vessels (probably the arterioles) but were lacking in other microvessels (venules) (Fig. 3F–J ). In general, PDGF-AB/FGF-2-induced angiogenic vessels recruited significantly higher numbers of pericytes 157
did not significantly induce recruitment of pericytes (Fig. 3A–C). FGF-2 induces PDGFR-␣ and PDGFR- expression in the newly formed vasculature To study the underlying mechanisms by which FGF-2 and PDGF-AA or PDGF-AB synergistically stimulate angiogenesis and vessel stability, we detected expression levels of PDGFRs in the newly formed corneal vessels. In situ hybridization with two specific probes for mouse PDGFR-␣ and PDGFR- showed that FGF-2 significantly up-regulated both PDGFR-␣ and PDGFR- mRNA expression in the newly formed blood vessels. It appears that high expression levels of both receptors types are localized in all vascular cells, including endothelial cells and mural cells (pericytes and vascular smooth muscle cells) because all vascular cells exhibited strong positive signals (Fig. 4C, D). In contrast, VEGF-induced new blood vessels completely lacked detectable signals for PDGFR-␣ and PDGFR- (Fig. 4A,
Figure 3. Detection of pericytes on corneal vessels. A–J) Corneas implanted with PDGF-AA (A, F), PDGF-AB (B, G), FGF-2 (C, H), PDGF-AA/FGF-2 (D, I), or PDGF-AB/FGF-2 (E, J) were removed on days 5 and 9 after implantation and stained with an anti-CD31 antibody (red) and a NG2 antibody (green). Arrows point to NG2-positive vessel areas. Bottom panels represent approximately ⫻4 view of inserts in middle panels. K) Quantification of pericyte recruitment index of newly formed blood vessels induced by various growth factors and data represent the mean average of 6 –10 random fields of each sample.
than PDGF-AA/FGF-2-induced vessels at different time points (Fig. 3D, E, I–K ). Quantification of pericyte recruitment index showed that PDGF-AB/FGF-2-induced blood vessels were able to maintain the same number of pericytes at a later time point (day 9), whereas PDGF-AA/FGF-2-induced blood vessels were unable to maintain the initial number of pericytes (Fig. 3K). Thus, there were significant differences of pericyte numbers in PDGF-AB/FGF-2- and PDGFAA/FGF-2induced vessels at both early and later time points. In contrast to combinatorial factors, single angiogenic factors, including PDGF-AB, PDGF-AA, and FGF-2 158
Vol. 23
January 2009
Figure 4. PDGFR expression and neutralization of vessel stability effect by an anti-PDGFR- antibody. A–D) At day 5 after VEGF (A, B) and FGF-2 (C, D) implantation, corneal blood vessels were in situ hybridized with a mouse PDGFR-␣ probe (A, C) or a mouse PDGFR- probe (B, D). E, F) Mouse corneas treated with anti-PDGFR- neutralizing antibody (F) and control buffer (E). Stable corneal blood vessels were examined on day 19 after pellet implantation. Arrows point to in situ positive signals (A–D); corneal vessels (E, F).
The FASEB Journal
ZHANG ET AL.
B). These findings provide a molecular basis of angiogenic synergism promoted by FGF-2 together with PDGF-AA or PDGF-BB. PDGFR- mediates vessel stability The fact that PDGF-AB and PDGF-BB, but not PDGF-AA in combination with FGF-2, were able to promote longlasting blood vessels, and both PDGF-AB and PDGF-BB could activate the PDGFR- receptor suggests PDGFR- was crucial for vessel stability. To study this possibility, we treated PDGF-AB/FGF-2-implanted mice with an antiPDGFR- neutralizing antibody. Interestingly, treatment with this antibody significantly abrogated the vascular stability effect induced by PDGF-AB and FGF-2 (Fig. 4E, F ). These findings demonstrate that PDGFR- is essential for PDGF-AB/FGF-2-induced vascular stability. Stimulation of collateral growth The synergistic effect of angiogenesis and establishment of long-lasting blood vessels by a combination of PDGF-AB/FGF-2 promoted us to study the therapeutic efficacy for the treatment of ischemic hind limbs. We established a severe rat ischemic hind limb model, in which long-lasting severe ischemia is produced in the left leg. Ischemia was created by a two-step surgical procedure. During the first operation, all arterial branches originated in the aorta distal to the renal arteries, the spermatic, left lumbar, ileolumbar, inferior mesenteric, caudal arteries, and all branches from the left iliac artery on the left side, were ligated. After 5 days, the left femoral artery was ligated. This ischemic model is relevant for leg ischemia found in patients. To
monitor establishment of stable collateral vessels, we performed angiographic analysis at day 42 (relatively long term) after the second operation. Delivery of the single angiogenic factors PDGF-AA, FGF-2, or PDGFAB, to the muscle demonstrated only weak effects on collateral growth (Fig. 5B–D), albeit there was a trend that FGF-2 and PDGF-AB appeared to be more potent than PDGF-AA (Fig. 5G). Although the combination of PDGF-AA with FGF-2 produced a significant angiogenic synergy in the mouse corneal angiogenesis model, dual delivery of these two factors to the muscle tissue did not produce a statistically greater effect than the FGF-2 alone-induced collateral growth at this time point (Fig. 5D, E, G). This finding supports our data obtained from the cornea angiogenesis model in which PDGF-AA exhibited an angiogenic synergy with FGF-2, but it was not able to stabilize and develop the newly formed vasculature. In contrast to all other single and dual angiogenic factor deliveries, treatment of the muscle tissue with PDGF-AB/FGF-2 significantly increased the number of collaterals observed 6 wk after femoral artery ligation (Fig. 5F, G). In addition to high numbers, the diameters of collaterals seemed to be enlarged in the PDGF-AB/FGF-2-treated group (Fig. 5F, white arrows). To further validate the effect of stimulation of collateral growth by various growth factors, we performed immunohistochemical analysis to quantify vascular smooth muscle cell-positive vessels. Although smooth muscle cell ␣-actin-positive vessels were present in muscle tissues treated with various growth factors and control samples, PDGF-AB/FGF-2 together significantly increased the number of arter-
Figure 5. Angiographic analysis of the rat ischemic hind limbs after delivery of various angiogenic factors. Left hindlimb ischemia was created on each animal, as described in Materials and Methods. Immediately after the second operation (ligation of the femoral artery), a slow-release polymer containing buffer alone (A), PDGF-AA (B), PDGF-AB (C), FGF-2 (D), PDGF-AA/FGF-2 (E), or PDGF-AB/FGF-2 (F), same as described in the mouse micropocket neovascularization assay, was implanted into the ligation site. Soluble growth factors were also injected into the ligation site for 7 times every other day for 12 days total. At day 42 after ligation of femoral artery, collateral growth was analyzed using angiographic analysis and quantified by counting the number of collaterals in each leg (A–G). Yellow arrows in panels A–F point to the newly formed collaterals; star marks the ligation site of the femoral artery; white arrows point to the collaterals whose diameters seemed to be larger. MECHANISMS OF ANGIOGENESIS BY PDGFS AND FGFS
159
ies in the tissue (Fig. 6A–G). Consistent with angiographic analysis, it appeared that PDGF-AB/FGF-2induced collaterals were slightly wider (Fig. 6F). Thus, our results provide compelling evidence that dual delivery of PDGF-AB/FGF-2 significantly stimulates both the growth and establishment of longlasting collaterals in the muscle tissue, implying the therapeutic potentials of delivery of a combination of these angiogenic/ateriogenic factors for the treatment of ischemic diseases. Improvement of blood flow To study improvement of blood flow after delivery of various single and combinatorial angiogenic/arteriogenic factors in the ischemic hind limbs, a laserDoppler analysis of paw skin perfusion was carried out at day 42 after the second ligation procedure. Consistent with angiographic findings, dual delivery of PDGFAB/FGF-2 significantly improved paw perfusion in the ischemic hind limb (Fig. 7F, G). In contrast, delivery of PDGF-AA, PDGF-AB, or FGF-2 did not significantly improve paw blood flow in the ischemic hind limbs, nor did the combination of PDGF-AA with FGF-2 (Fig. 7A–E, G). These results show that dual delivery of PDGF-AB with FGF-2 induced functional collaterals that increased blood flow and enhanced distal skin perfusion. DISCUSSION The concept of delivery of proangiogenic factors to the tissue for therapeutic improvement of blood perfusion
was proposed more than 30 years ago (1). This concept was born almost simultaneously as the idea of therapeutic antiangiogenesis for the treatment of angiogenesisdependent diseases such as cancer and ocular diseases. It was assumed that therapeutic angiogenesis might be a straightforward approach for the treatment of ischemic cardiovascular disorders. Ironically, therapeutic antiangiogenesis, which was thought to be a more difficult approach at the time, has become one of the most successful fields for research and pharmaceutical development with a few new classes of drugs in the market for malignant and nonmalignant angiogenesisdependent diseases (12–16). Regardless of numerous successful studies of delivery of various proangiogenic factors in ischemic animal models, all large, randomized, double-blinded, and placebo-controlled clinical trials for treatment of cardiovascular diseases have not demonstrated significant beneficial effects (2). These failures have raised the classical concern of the ability to translate observation from animal models to human trials. Furthermore, the most important unresolved issue is probably our poor understanding of the complex basic cardiovascular biology and the fundamental aspects of angiogenesis and arteriogenesis processes. Better knowledge is essential for positive results in clinical trials. Development of therapeutic arteriogenesis and angiogenesis for the treatment of ischemic cardiovascular disorders has to fulfill the following criteria: 1) the ability to establish arterial blood vessels that perfuse highly oxygenated blood to the ischemic tissue; 2) newly formed vascular networks should be functional with normal vessel properties; 3) blood
Figure 6. Immunohistochemical analysis of collaterals in the ischemic muscle tissue. At day 42 after femoral ligation, ischemic muscle tissue samples were collected around the ligated and treated areas of the left hind limb of each animal. A–F ) Representative micrographs of samples treated with buffer (A), PDGF-AA (B), PDGF-AB (C), FGF-2 (D), PDGF-AA/FGF-2 (E), or PDGF-AB/FGF-2 (F) and stained with an anti-␣-smooth muscle cell actin antibody. G) Average numbers of collaterals in each sample were counted from 10 different fields; means ⫾ sd. Arrows point to stained positive structures (A–F). 160
Vol. 23
January 2009
The FASEB Journal
ZHANG ET AL.
Figure 7. Blood flow analysis. At day 42 after femoral artery ligation, paw blood flow of left hind limbs treated with buffer (A), PDGF-AA (B), PDGF-AB (C), FGF-2 (D), PDGF-AA/FGF-2 (E), or PDGF-AB/FGF-2 (F) was analyzed using a laser-Doppler technique. Right side of hind limb of each animal served as a control (A–F). Blood flow was quantified in each group by relating the left side perfusion to the right, then comparing between the groups (G). Arrows point to the left side ischemic hind limb.
flow in these networks should enable physiological regulation without leakage; and 4) the newly established vascular network should remain stable long term. In principle, arteries consist of two characteristic cell types: endothelial cells and pericytes/vascular smooth muscle cells. Endothelial cells are the primary cell type that delineates lumens, while pericytes/ smooth muscle cells stabilize and control blood flow within the vessels. The importance of these two functions and cell types in the arteries suggest that two different growth factors targeting each cell type should be considered for therapeutic development. To date, there is no known single angiogenic factor that could accomplish this complex task. Indeed, our present and previous observations demonstrate that delivery of single angiogenic factors, including VEGF-A, FGF-2, PDGF-BB, PDGF-AA, or PDGF-AB in the mouse angiogenesis and in the rat ischemic hind limb models is unable to establish functional arterial networks (8). In contrast, dual delivery of FGF-2 in combinations with PDGF-AB or PDGF-BB induces long-lasting vessel networks in both mouse corneas and skeletal muscles. Accordingly, our findings provide crucial information about the choice of proangiogenic agents for therapy. Although the angiogenic responses induced by the combination of PDGF-AB/FGF-2 are overwhelming in the corneal tissue, delivery of the same combination of these two growth factors did not result in a similar level of collateral growth in the ischemic muscle tissue. One possible reason could be that the corneal tissue is avascular and thus the angiogenic responses are easier to detect than in the skeletal muscle tissue. The other possibility is that a nonoptimal dose has been used for delivery in the muscle tissue. It would be extremely important for our future studies to use various dosages for delivery in the ischemic muscle tissues. Additionally, an optimal delivery system, including the choice of slow-release polymers, bioavailability of angiogenic MECHANISMS OF ANGIOGENESIS BY PDGFS AND FGFS
stimuli, the optimal ratios of growth factors and the optimal time of delivery warrant thorough studies. We plan to address these important issues in our future studies. Intriguingly, in PDGF-AA/FGF-2-, PDGF-AB/FGF-2-, or PDGF-BB/FGF-2-implanted corneas, blood vessels continuously grow even after removal of the implanted growth factors at day 5, suggesting that spatiotemporal kinetics of release of these factors are not crucial to subsequent neovascularization and stability induced by these two particular factors (Figs. 1 and 2). The underlying mechanism by which corneal vessels continue to grow without growth factor ligands is not understood. It is speculated that high levels of PDGFRs induced by FGF-2 could subsequently form receptor aggregates of dimers and oligomers that could transduce active signals without ligands. As illustrated in Fig. 8, FGF receptors are expressed in endothelial cells, and FGF-2 is able to activate these receptors in angiogenic vessels. The activation of the FGF signaling pathway leads to up-regulation of both PDGFR-␣ and PDGFR-, which are subsequently activated by PDGF ligands. Thus, in the presence of PDGF ligands and FGF-2, both FGF- and PDGF-signaling pathways become simultaneously activated, and synergistic angiogenic responses are induced. Withdrawal of PDGF ligands might continuously lead to activation of PDGF receptors, which are aggregated due to the extraordinary high levels of expression (Fig. 4C, D). However, only PDGF-AB/ FGF-2 and PDGF-BB/FGF-2 are able to stabilize the newly formed vasculature by recruiting vascular supportive cells, pericytes, onto the nascent vasculature by activating the PDGFR- expressed in pericytes. Indeed, the combination of PDGF-AB/FGF-2 significantly improves the recruitment of pericytes onto the corneal vessels. It is unclear why FGF-2 is also required for recruitment of pericytes since PDGF-AB alone is unable to significantly recruit these supportive cells. However, it is possible that FGF-2 might also able to significantly up-regulate PDGFR- in 161
Figure 8. Schematic diagram of possible mechanisms by which PDGFs in combination with FGF-2 promote angiogenic synergy and vascular stability. When a newly formed angiogenic vessel is exposed to FGF-2, PDGFRs are up-regulated in endothelial cells and become responsive to PDGF ligand stimulation. Thus, the FGF and PDGF signaling systems simultaneously become activated, and angiogenic synergisms are observed. Since the PDGFRs are up-regulated in high levels by FGF-2 and they potentially form receptor aggregates that continuously remain activated, even after removal of the PDGF ligands, persistent angiogenic responses persist. Both PDGF-AB and PDGF-BB could recruit pericytes via activation of the PDGFR-beta receptor and stabilize the newly formed blood vessels.
pericytes. In support of this hypothesis, in situ hybridization reveals that almost every vascular cell, including endothelial cells and pericytes, express high levels of PDGFRs. This interesting aspect warrants further mechanistic studies. In the case of a combination of VEGF and PDGF-BB, spatiotemporal kinetics seem to play a role in vessel formation and vascular patterning (17). Perhaps, different combinations of angiogenic and arteriogenic factors involve different mechanisms for promoting vessel growth and maturation. One of the most striking findings of our present study is that PDGF-AB, but not PDGF-AA, produces vascular stability in a combination with FGF-2. Since PDGF-AA only binds to PDGFR-␣ and PDGFR-AB binds to PDGFR-␣, the stabilization of blood vessels by PDGF-AB indicates that PDGFR- receptor is crucial for vascular stability. Indeed, the anti-PDGFR- neutralizing antibody nearly completely abolished the vascular stability effect of PDGF-AB/FGF-2, indicating that the PDGFR- receptor is required for vessel stability. This finding is further supported by our previous results obtained from PDGF-BB studies (8). Consistent with our data, previous genetic and vascular studies have indicated the roles of PDGFR- in stabilization of newly formed blood vessels (18, 19). Surprisingly, PDGF-AB and PDGF-BB alone are unable to stabilize newly formed vasculatures, suggesting that targeting vascular pericytes or smooth muscle cells only, is not sufficient for vascular stability. Although FGF-2 acts on a broad spectrum of cell types in vitro, this growth factor mainly targets endothelial cells in vivo (20). Thus a combination of PDGF-AB with FGF-2 would act on pericytes/ vascular smooth muscle cells and endothelial cells, respectively. Probably, only combinations of angiogenic factors, such as FGF-2, and arteriogenic factors, such as 162
Vol. 23
January 2009
PDGF-AB or PDGF-BB are able to provide survival signals for both cell types. Although PDGF-AA is unable to serve as a vascular stabilizer, the combination of PDGF-AA/FGF-2 produces a significant angiogenic synergism, indicating that PDGFR-␣ receptor mediates active angiogenic signals. Consistent with this finding, PDGFR-␣, the unique receptor for PDGF-AA, has been reported to be expressed on vascular endothelial cells (21). Taken together, previous clinical studies using proangiogenic factors for promoting angiogenesis for the treatment of ischemic cardiovascular and peripheral vascular disorders lacked robust mechanistic rationales. The choice of proangiogenic agents for trials appears to have been dependent on the availability of angiogenic factors and intellectual property rights (2). All previous trials were based on delivery of single proangiogenic molecules such as FGF-2, VEGF-A, VEGF-C, or PDGF-BB (2). Delivery of proangiogenic factor such as VEGF-A to the ischemic site not only induces disorganized vasculatures, but also increases vascular leakage, which may negatively influence healing and regeneration of ischemic cardiac muscle tissue. In the leg, the site for growth factor delivery is another issue often neglected in past. Accordingly, it is not surprising that delivery of these angiogenic factors to the ischemic muscle tissue does not produce beneficial effects. Our present study provides compelling evidence that dual delivery of FGF-2 in a combination with PDGF-AB but not with PDGF-AA promotes long-lasting vessels in the ischemic muscle tissue. Thus, our observations provide a guide for choosing angiogenic/arteriogenic factors for the treatment of ischemic diseases. We thank Dr. Ebba Bråkenhielm for performing the in situ hybridization studies. This study was supported by grants
The FASEB Journal
ZHANG ET AL.
from China State Scholarship (Beijing, China), the National Basic Research Program of China (2006CB503803), the National Natural Science Foundation of China (30428005), the Swedish Heart and Lung Foundation (Stockholm, Sweden), the EU integrated project of VascuPlug (contract STRP 013811) and Angiotargeting (contract 504743), the So¨derberg Foundation (Stockholm, Sweden), and the Swedish Medical Research Council (Stockholm, Sweden).
11. 12.
13.
REFERENCES 14. 1. 2.
3.
4. 5. 6. 7. 8.
9. 10.
Kumar, S., West, D., Shahabuddin, S., Arnold, F., Haboubi, N., Reid, H., and Carr, T. (1983) Angiogenesis factor from human myocardial infarcts. Lancet 2, 364 –368 Toyota, E., Warltier, D. C., Brock, T., Ritman, E., Kolz, C., O’Malley, P., Rocic, P., Focardi, M., and Chilian, W. M. (2005) Vascular endothelial growth factor is required for coronary collateral growth in the rat. Circulation 112, 2108 –2113 Vale, P. R., Losordo, D. W., Milliken, C. E., Maysky, M., Esakof, D. D., Symes, J. F., and Isner, J. M. (2000) Left ventricular electromechanical mapping to assess efficacy of phVEGF(165) gene transfer for therapeutic angiogenesis in chronic myocardial ischemia. Circulation 102, 965–974 Simons, M., and Ware, J. A. (2003) Therapeutic angiogenesis in cardiovascular disease. Nat. Rev. Drug. Discov. 2, 863– 871 Carmeliet, P. (2000) Fibroblast growth factor-1 stimulates branching and survival of myocardial arteries: a goal for therapeutic angiogenesis? Circ. Res. 87, 176 –178 Simons, M. (2005) Angiogenesis: where do we stand now? Circulation 111, 1556 –1566 Carmeliet, P. (2000) Mechanisms of angiogenesis and arteriogenesis. Nat. Med. 6, 389 –395 Cao, R., Brakenhielm, E., Pawliuk, R., Wariaro, D., Post, M. J., Wahlberg, E., Leboulch, P., and Cao, Y. (2003) Angiogenic synergism, vascular stability and improvement of hind-limb ischemia by a combination of PDGF-BB and FGF-2. Nat. Med. 9, 604 – 613 Arbiser, J. L. (2007) Why targeted therapy hasn’t worked in advanced cancer. J. Clin. Invest. 117, 2762–2765 Lu, H., Xu, X., Zhang, M., Cao, R., Brakenhielm, E., Li, C., Lin, H., Yao, G., Sun, H., Qi, L., Tang, M., Dai, H., Zhang, Y., Su, R.,
MECHANISMS OF ANGIOGENESIS BY PDGFS AND FGFS
15.
16. 17. 18. 19. 20.
21.
Bi, Y., Zhang, Y., and Cao, Y. (2007) Combinatorial protein therapy of angiogenic and arteriogenic factors remarkably improves collaterogenesis and cardiac function in pigs. Proc. Natl. Acad. Sci. U. S. A. 104, 12140 –12145 Cao, Y., and Cao, R. (1999) Angiogenesis inhibited by drinking tea. Nature 398, 381 Yang, J. C., Haworth, L., Sherry, R. M., Hwu, P., Schwartzentruber, D. J., Topalian, S. L., Steinberg, S. M., Chen, H. X., and Rosenberg, S. A. (2003) A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N. Engl. J. Med. 349, 427– 434 Ferrara, N., and Kerbel, R. S. (2005) Angiogenesis as a therapeutic target. Nature 438, 967–974 Rosenfeld, P. J., Brown, D. M., Heier, J. S., Boyer, D. S., Kaiser, P. K., Chung, C. Y., and Kim, R. Y. (2006) Ranibizumab for neovascular age-related macular degeneration. N. Engl. J. Med. 355, 1419 –1431 Brown, D. M., Kaiser, P. K., Michels, M., Soubrane, G., Heier, J. S., Kim, R. Y., Sy, J. P., and Schneider, S. (2006) Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N. Engl. J. Med. 355, 1432–1444 Stone, E. M. (2006) A very effective treatment for neovascular macular degeneration. N. Engl. J. Med. 355, 1493–1495 Chen, R. R., Silva, E. A., Yuen, W. W., and Mooney, D. J. (2007) Spatio-temporal VEGF and PDGF delivery patterns blood vessel formation and maturation. Pharm. Res. 24, 258 –264 Lindahl, P., Johansson, B. R., Leveen, P., and Betsholtz, C. (1997) Pericyte loss and microaneurysm formation in PDGF-Bdeficient mice. Science 277, 242–245 Crosby, J. R., Seifert, R. A., Soriano, P., and Bowen-Pope, D. F. (1998) Chimaeric analysis reveals role of Pdgf receptors in all muscle lineages. Nat. Genet. 18, 385–388 Abraham, J. A., Mergia, A., Whang, J. L., Tumolo, A., Friedman, J., Hjerrild, K. A., Gospodarowicz, D., and Fiddes, J. C. (1986) Nucleotide sequence of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor. Science 233, 545–548 Kitami, Y., Inui, H., Uno, S., and Inagami, T. (1995) Molecular structure and transcriptional regulation of the gene for the platelet-derived growth factor alpha receptor in cultured vascular smooth muscle cells. J. Clin. Invest. 96, 558 –567 Received for publication May 16, 2008. Accepted for publication August 21, 2008.
163
