High-Resolution Quantitative Computed Tomography Demonstrating Selective Enhancement of Medium-Size Collaterals by Placental Growth Factor-1 in the Mouse Ischemic Hindlimb
Background— The process of arteriogenesis after occlusion of a major artery is poorly understood. We have used high-resolution microcomputed tomography (μ-CT) imaging to define the arteriogenic response in the mouse model of hindlimb ischemia and to examine the effect of placental growth factor-1 (PlGF-1) on this process.
Methods and Results— After common femoral artery ligation, μ-CT imaging demonstrated formation of collateral vessels originating near the ligation site in the upper limb and connecting to the ischemic calf muscle region. Three-dimensional μ-CT and quantitative image analysis revealed changes in the number of segments and the segmental volume of vessels, ranging from 8 to 160 μm in diameter. The medium-size vessels (48 to 160 μm) comprising 85% of the vascular volume were the major contributor (188%) to the change in vascular volume in response to ischemia. Intramuscular injections of Ad-PlGF-1 significantly increased Sca1+ cells in the circulation, α-actin–stained vessels, and perfusion of the ischemic hindlimb. These effects were predominantly associated with an increase in vascular volume contributed by the medium-size (96 to 144 μm) vessels as determined by μ-CT.
Conclusions— High-resolution μ-CT delineated the formation of medium-size collaterals representing a major vascular change that contributed to the restoration of vascular volume after ischemia. This effect is selectively potentiated by PlGF-1. Such selective enhancement of arteriogenesis by therapeutically administered PlGF-1 demonstrates a desirable biological activity for promoting the growth of functionally relevant vasculature.
Received September 7, 2005; revision received March 20, 2006; accepted March 23, 2006.
In vascular beds with atherosclerotic ischemic disease, collateral formation plays an important role in preserving blood flow and organ function. In the heart, collaterals protect against damage resulting from multiple episodes of ischemia.1 In patients or animals models with cardiovascular risk factors, including hypercholesterolemia, hypertension, diabetes, obesity, and smoking, collateral development is impaired.2–5 The biological bases for the reduced collateral formation are not well understood. Impaired recruitment of inflammatory cells or production of angiogenic factors may attenuate arteriogenesis. For example, reduced angiogenesis in older,5 diabetic,6 and hypercholesterolemic2 animals is associated with reduced expression of vascular endothelial growth factor (VEGF). Recent studies have also shown that circulating progenitor cells that enhance angiogenesis are reduced in patients with cardiovascular risk factors.7,8
Clinical Perspective p 2453
In the embryo, large arteries are formed de novo from the capillary network. However, collateral development in the adult is not well understood. Several possibilities have been suggested, including remodeling of preexisting collaterals9–11 and formation of capillaries followed by remodeling into arterioles.12 After femoral artery ligation in the mouse, lower limb blood flow is restored to 50% to 60% of preligation levels.13,14 On the basis of hemodynamic estimates, millions of capillaries generated through angiogenesis would be required to generate the lumen needed to replace the flow provided by the conduit artery.11 Experimentally, this has not been observed, except in situations of hemangioma.15 The currently used histochemical, laser Doppler, and x-ray angiography techniques are inadequate to visualize, quantify, and characterize vascular development in response to the loss of a major arterial conduit. Thus, the nature of the vasculature contributing to the recovery after therapy is poorly defined. Therefore, to more thoroughly understand the nature of the vasculature responsible for the restoration of blood flow in these settings, we have developed a quantitative, 3D, high-resolution microcomputed tomography (μ-CT) imaging approach. Using μ-CT in a murine hindlimb ischemia model, we now show that ligation of the common femoral artery results in changes in the number and volume of vessels ranging from 8 to 160 μm in diameter, with a 188% increase in new vascular volume attributable to medium-size (48 to 160 μm) vessels.
Placental growth factor-1 (PlGF-1) is a member of the VEGF gene family with unique biological activities.16 Unlike VEGF, PlGF-1 is not an endothelial cell mitogen but rather promotes smooth muscle cell and macrophage chemotaxis.17 In particular, PlGF has been suggested to act as an arteriogenic agent.18,19 To define the nature of the neovascularization in response to ischemia and the arteriogenic properties of PlGF-1, the new μ-CT imaging approach was used. We found that adenoviral PlGF-1 gene delivery selectively promoted the vascular volume of medium-size vessels in the ischemic hindlimb.
Male BALB/c mice (12 weeks old) from Harlan Sprague-Dawley (Indianapolis, Ind) were used for all studies. Studies were approved by Eli Lilly’s institutional animal care and use committee.
PlGF Adenoviral Vector
Adenoviral human-PlGF-1 (Ad-PlGF-1) was generated by cloning of 2 restriction fragments of the PlGF-1 gene. The p-shuttle transfer vector was linearized with Pme1, and the resulting fragment was cotransformed with use of the pAdEasy-1 vector (Stratagene, La Jolla, Calif) into Escherichia coli BJ5183. On homologous recombination in the bacteria, plasmid was generated in which the expression cassette was inserted into the E1 region of the adenovirus driven by the human cytomegalovirus promoter. After characterization, the recombinant adenoviral vector was purified and linearized with restriction enzyme PacI. The linearized DNA was transfected into HEK293 cells to obtain viral plaques, which were then amplified and purified (Viraquest Inc, North Liberty, Iowa).
The activity of the PlGF-1 generated from the adenoviral construct was determined in cells expressing human VEGFR1 (flt-1) and VEGFR2 (KDR). The PlGF-induced signal was 3- to 4-fold higher in cells transfected with both flt-1 and KDR than with flt-1 alone; therefore, the former were used for routine assay. Cells were cotransfected with the luciferase reporter gene containing transcriptional activator Ets-like protein (ELK), which is activated on phosphorylation by mitogen-activated protein kinase (Stratagene, La Jolla, Calif). HEK293 cells were transfected with flt-1, KDR, and ELK-luciferase plasmids with use of the Mirus LT-1 transfection system. After 18 hours, cells were treated with Ad-PlGF-1 and incubated for another 24 hours. Luciferase activity, an indication of PlGF-1 response, was determined with a luminometer (La Roche Diagnostics, Indianapolis, Ind). PlGF-1 protein was determined by ELISA (R&D Systems, Minneapolis, Minn). Recombinant human PlGF-1 expressed in E coli was obtained from R&D Systems; mammalian cell–derived human PlGF-1 was prepared at Eli Lilly by cloning and expression in Chinese hamster ovary cells.
In Vivo Expression of PlGF
Control vector or Ad-PlGF (25 μL total) was administered into the calf muscle in 3 separate injections. After 3 days, mice were euthanized, and the calf muscles were frozen in LN2 and stored at −80°C. Total RNA was prepared by use of the Trizol method (Invitrogen, Carlsbad, Calif). The PlGF-1 mRNA was amplified with a human PlGF-1–specific probe and primers in a real-time polymerase chain reaction (ABI 7900 sequence detection system). Levels of human PlGF-1 mRNA were normalized to glyceraldehyde 3-phosphate dehydrogenase.
Progenitor Cell Mobilization
Ad-PlGF-1 was administered to the calf muscle of nonligated mice. After 3 days, the mice were anesthetized with isoflurane, and blood samples were collected by cardiac puncture. Human granulocyte colony stimulating factor (G-CSF; R&D Systems, Minneapolis, Minn) was administered subcutaneously daily (200 μg per mouse). On day 5, blood was collected by cardiac puncture. Red blood cells were lysed by mixing 300 μL of blood with 5 mL of the lysis buffer (Santa Cruz Biotechnology, Inc, Santa Cruz, Calif). After 10 minutes, cells were pelleted by centrifugation and resuspended in 400 μL of phosphate-buffered saline containing 0.1% bovine serum albumin. Aliquots (100 μL) of the cell suspension were then labeled with PE-antimouse Sca1 antibodies (BD Biosciences, San Jose, Calif). Similarly, staining with other markers (c-kit, CD45, CD11b, KDR, and flt-1) was performed with antibodies from BD Biosciences. Cells were washed, fixed with 0.5% formaldehyde, and analyzed by fluorescence-activated cell sorting. Peripheral blood cells were analyzed with use of a Hemavet 850 hematology system (Drew-Scientific Inc, Ramsy, Minn).
Hindlimb Ischemia Model
Mice were anesthetized with a mixture of 1.5% isoflurane in O2 (1.5 L/min). On day 1, both the iliac artery (and the vein proximal to the pudenoepigastric vein (and the femoral artery (and vein) were ligated with a 6-0 silk suture. After ligation, mice received 3 injections (total, 25 μL) of saline, control vector, or Ad-PlGF-1 in the adductor muscle of the left hindlimb at 3 sites near the ligation. The first site was on the ventral surface, the second on the dorsal surface (femoral biceps muscle) at the same height, and the third in the adductor muscle below. Preligation and postligation laser Doppler scans were taken to ensure consistency of ligation. The hindlimbs were then scanned weekly. At the end of 4 weeks, the mice were prepared for μ-CT and immunohistochemistry.
Laser Doppler Blood Flow
Flow images of the foot were acquired with a Moor LDI infrared laser Doppler imager (Moor Instruments Ltd, Wilmington, Del). Skin temperature was maintained at 36.5°C to 37.5°C. Scanning was performed before and immediately after surgery and on days 7, 14, 21, and 28. Mice were positioned ventrally with their rear legs outstretched to allow scanning of the entire plantar surface. The data were analyzed with a Moor LDI image processor V3.09 and reported as left/right foot flow after background subtraction.
The hindlimbs were fixed in 3% paraformaldehyde (EMS, Fort Washington, Pa) for 24 hours and preserved in 70% ethanol. A 4-mm segment of the calf muscle medial to the left femur was embedded in paraffin. Ten serial sections (5 μm thick) were cut and mounted. Sections were treated 3 times sequentially with xylene, 100% ethanol, and 95% ethanol/water. Antigen retrieval was performed by sequential incubation at 96°C and at room temperature for 15 minutes. The sections were washed with water, TBST (10 mmol/L Tris-HCl, pH 7.4; 130 mmol/L NaCl; and 0.1% Tween-20), and 3% H2O2. After protein blocking, the sections were washed and then incubated with either α-actin monoclonal antibody (Dako, Carpinteria, Calif) or isolectin B4 (Vector Labs, Burlingame, Calif). Histology images at ×200 magnification were captured with a digital camera. Images were quantified by Image*Pro v126.96.36.199 analysis software (Media Cybernetics, Silver Spring, Md). Data were expressed as the mean±SEM and compared by ANOVA.
Three-Dimensional High-Resolution μ-CT
Mice were injected with heparin (100 U) 5 minutes before euthanasia with isoflurane. The vasculature was flushed with saline containing papaverine (4 mg/L) and adenosine (1 g/L) through a needle inserted into the left ventricle. The inferior vena cava was cut to allow outflow of the perfusate. For fixation, 3.7% paraformaldehyde was perfused for 5 minutes at 100 mm Hg pressure. The descending aorta was then cannulated, and 0.7 mL of contrast agent (50% bismuth [Sigma-Aldrich, St. Louis, Mo], 5% gelatin) was injected. The vasculature was imaged by μ-CT (GE Healthcare, Waukesha, Wis) at a detector pixel size of 8 μm; x-ray tube voltage, 80 kV; x-ray tube current, 80 μA; detector bin mode, 1×1; and 500 view angles. The field of view was 2 cm in the axial direction and 1.04 cm in the slice. Reconstruction at 8 μm of 500 slices or at 16 μm of 250 slices was performed with a cone-beam algorithm. Bone and preexisting vessels were used as anatomic references. A 2D maximum-intensity projection image was generated with the use of Microview Software (GE Healthcare). A threshold was chosen on the basis of visual 3D vessel number, volume, and eliminated bone according to a modified Image Pro-Plus 5.0 algorithm, Vessel Tree 0.9 (Media Cybernetics, Silver Spring, Md). The data are expressed as vascular segment number and volume counted in 500 slices in the 3D μ-CT images.
Quantitative Image Analysis
Image-Pro Plus 5.0.1 (Media Cybernetics, Silver Spring, Md) was used for image analysis and quantification of the number, length, and volume of vessels. The algorithm was built within the AutoPro programming language (Visual Basic SAX engine) inside Image-Pro Plus. The 3D imaging data were analyzed on each of the 2D planes that made up the 3D volume. On each plane, the “spots” of vasculature are counted by width and area. The total number of spots at each width across all planes corresponds to the vessel length at that width. The total area at each width corresponds to the vascular volume for that vessel size. The minimum width was used in the 3D analysis.
Statistical significance was determined by ANOVA from SAS (version 8.2) and JMP (version 5.1) programs (SAS Institute, Cary, NC). With small sample sizes and unequal variances between groups, the averaged ranked data of each variable were used in the ANOVA. Logarithmic transformation was applied to vessel number and vessel volume whose data distributions were skewed. For time-dependent or segment-dependent effects on relative blood flow or vessel number and vessel volume from animals treated independently, a repeated-measures ANOVA was applied. A repeated-measures test by considering a split-plot experiment with correlated whole plots was applied when the animals were not treated independently, and the multiple segments were measured. Significance was declared when P<0.05. Results are expressed as mean±SEM.
The authors have full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.
Arteriogenesis in the Ischemic Hindlimb
To more precisely define the nature of arteriogenesis in the murine hindlimb ischemia model, we used a μ-CT–based method. To this end, a 3D μ-CT image analysis system was developed to quantify the changes in the number as well as volume of the vessel segments of different sizes. Figure 1 is a high-resolution μ-CT image of normal leg vasculature (A) and 4 weeks after femoral artery ligation (B). The major vessels of the normal leg principally contributing to the blood supply are designated. The medium and small vessels appear straight, with little or no evidence of tortuosity. In contrast, 4 weeks after ligation, several distinct vessels of &80 to 120 μm, not previously visible, become highly prominent (Figure 1B and 1C, arrows). These vessels originated from the regions near or above the ligation site and connected all the way to the lower limb. The 2D μ-CT images clearly show primary collateral development in response to ischemia. To define the extent of the vasculature, high-resolution scanning with 8-μm reconstruction enabled visualization of vessels in the 8- to 300-μm range. A single slice of the z-position shows the range of vessel size detectable by μ-CT imaging, as depicted by the white spots in Figure 1C and the connecting sites with the large artery in the lower limb (Figure 1D).
Image analysis of the z-slices enabled determination of the number of vessels ranging from 8 to 250 μm in size at various points along the leg. The vessel diameter determination by μ-CT was validated by using wires of specified diameters shown in Figure 2. The actual size, the number of μ-CT pixels, and the estimated size after image analysis are shown in the table within Figure 2. These data confirm that the size estimation by μ-CT and image analysis are highly accurate, providing vessel size values within 10% of the actual size.
Figure 3 shows the representative longitudinal z-position slices of the upper and lower limbs from control and ligated animals at their matched sites. Bone and preexisting vessels were used as anatomic references to closely match the sections from the 2 groups. These representative sections selected from different sites provided visual evidence indicating that femoral artery ligation resulted in an increased number of vessels. We then used 500 serial z-position slices representing 4000 μm of tissue length to determine the total number and the volume of vessel segments within the indicated size range. Although changes in the number as well as volume of vessels ranging from 8 to 160 μm in size were observed (Figure 4), a large change in vessel volume ranging from 48 to 160 μm in diameter occurred in response to ischemia. For determination of the contribution by different size vessels to total vessel number and volume, we arbitrarily categorized the vessels into 2 groups of small (8 to 40 μm) and medium (48 to 160 μm) size. Vessel segments >160 μm were few, which included the preexisting large vessels, and therefore were not included in the quantification. The Table shows that even though 62% of vessel segments (20 574, versus 32 981 in the ligated leg) were small, they only contributed to &15% (165 396 versus 1 067 526 mm3) of the vascular volume. The major proportion (86%) of the vascular volume (902 131 versus 1 067 526 mm3) was attributed to the medium-size vessels. These data suggest that measurement of vessel volume is a more relevant parameter than is capillary number obtained by histology. Ligation resulted in an 86% and a 128% increase in the number of small and medium-size vessel segments, respectively. More important, the major change in vessel volume (188%) was contributed by medium-size vessels. These data show that the increase in vessel volume due to the formation of medium-size collaterals was a major contributing factor to recovery after femoral artery ligation. Because μ-CT is a contrast perfusion–based method, the ability to discriminate veins and arteries remains a limitation. Because angiogenesis and arteriogenesis are the major responses after ischemia, despite the limitation, the method has proven highly effective in discerning the vascular anatomy and quantifying and discriminating vascular changes previously not achieved.
Expression and Biological Activity of Ad-PlGF-1
Having established the μ-CT features of developing collaterals in the mouse hindlimb, we then used this model to define the biological effects of PlGF-1. Ad-PlGF-1 was first characterized for expression of biologically active PlGF-1 in vitro and in vivo. Transduction of the Ad-PlGF-1 construct into HEK293 cells produced a dose-dependent increase in PlGF-1 expression (Figure 5A). Adenoviral vector–generated PlGF-1 was biologically active, as demonstrated by activation of flt-1 in a reporter assay (Figure 5B). On the basis of the determination of PlGF-1 protein by ELISA and of its biological activity by flt-1 receptor activation, the specific activity of Ad-PlGF-1 was comparable to that of recombinant PlGF-1 produced in E coli (R&D Systems) or mammalian cells generated at Eli Lilly and Co.
Determination of in vivo PlGF-1 expression from the adenoviral construct after administration into the calf muscle showed a dose-dependent expression of PlGF-1 mRNA in muscle (Figure 5C) and the appearance of low-level PlGF-1 in plasma. A 108 plaque-forming-unit (pfu) dose of Ad-PlGF-1 produced 608±288 pg/mL (n=6) of PlGF-1 in plasma after 3 days of injection, which became undetectable by day 7. Similar expression of PlGF-1 mRNA was observed when Ad-PlGF was administered to the adductor muscle. Because of the convenience of quantification of local expression, we routinely used calf muscle to test the in vivo expression of PlGF-1.
To demonstrate the in vivo biological activity of PlGF-1, we studied the effect of Ad-PlGF-1 on the mobilization of progenitor cells. We found that PlGF-1 selectively mobilized Sca1+ cells into the circulation. On day 3 after Ad-PlGF-1 administration, the number of circulating Sca1+ cells was 2.5-fold higher than in controls (Figure 5D). The cells exhibiting other markers, including c-kit, CD45, CD11b, KDR, and flt-1, were not altered. Complete blood counts showed that unlike G-CSF, PlGF-1 treatment did not affect circulating leukocytes, lymphocytes, or monocytes (Figure 6). These results demonstrated that Ad-PlGF-1 gene delivery resulted in a dose-dependent expression of biologically active PlGF-1 in vitro and in vivo.
Improvement in Blood Flow and Arteriogenesis by Ad-PlGF-1
To study the effect of PlGF-1 on arteriogenesis, the gene construct was injected into the adductor muscle after femoral artery ligation. Laser Doppler measurements showed a time-dependent recovery of tissue perfusion, reaching 52±5% of preischemia blood flow after 4 weeks. Control vector did not produce any significant change in the laser Doppler perfusion measurement (53±4%). Administration of 108 pfu of Ad-PlGF-1 resulted in an overall significant improvement in blood flow (67±7%; Figure 7A). Determination of capillary density by isolectin staining and arterial vessel density by α-actin staining showed that total capillary density in vector- and Ad-PlGF-1–treated limbs was similar (Figure 7B and 7C), whereas Ad-PlGF-1 produced a 60% (P<0.05) increase in α-actin–stained vessels (Figure 7B and 7D).
High-resolution, 3D μ-CT image analysis was further used to examine the effect of Ad-PlGF-1 on the vascular response to ischemia. Ad-PlGF-1 doses were selected on the basis of the initial studies demonstrating a minimum, nonspecific effect on Sca1+ cell mobilization. Control vector >5×108 pfu induced mobilization of bone marrow cells and a nonspecific inflammatory response in the tissue. Therefore, doses <108 pfu were used in these studies. The vessel number and segmental volume were determined in a block of 700×900×500-μm tissue from control and Ad-PlGF-1–treated animals. (Note that a larger tissue block [700×1000×500 μm] was used in Figure 4, which resulted in a larger number of total vessels and volume compared with those in Figure 8.) Determination of vessel number and segmental volume after 4 weeks of treatment showed that PlGF-1 did not significantly alter the number or the volume of small vessels (Figure 8). However, Ad-PlGF-1 treatment produced a significant (P<0.05) increase in the volume of vessel segments in the 96- to 136-μm-diameter range. These effects of Ad-PlGF-1 were dose dependent. For example, relative to vector control, the increase in vascular volume in vessel size ranging from 104 to 128 μm was 58% (P≤0.05) at 108, 29% (P<0.05) at 5×107, and 0.2% (not significant) at 107 pfu Ad-PlGF dose. These data show that PlGF-1 gene delivery produced a significant and selective increase in vascular volume contributed by medium-size collaterals.
Collateral development plays an important role in the restoration of distal blood flow and tissue salvage in a number of physiological and clinical settings.12 Deficiency in collateral development has been well documented in patients and animals with cardiovascular risk factors.2–6 At the present time, the nature and mechanism of collateral development are not well characterized. In large measure, this stems from our inability to visualize and quantify changes in the entire vascular tree in an organ. Indeed, commonly used histological and laser Doppler techniques are limited to semiquantitative vessel counts within a limited area and measurement of superficial blood flow, respectively, whereas x-ray angiography is limited by its spatial resolution.
In this study, we used a high-resolution, 3D μ-CT imaging technique to investigate the nature of vascular development and the effect of PlGF-1 gene delivery in a mouse model of hindlimb ischemia. Our principle findings are that arteries in the 48- to 160-μm range accounted for the predominant change and the principle component responsible for the restoration of arterial perfusion. Importantly, PlGF-1 selectively enhanced growth of arterial vessels that overlapped with this size class.
Delineation of a specific size of arterial vessel that appears to play a key role in the restoration of arterial perfusion is an unexpected and important finding. Because the number of vessels required to restore functional perfusion in the capillary category would be very high, the development of few larger collaterals would be a physiologically beneficial mechanism. This is also supported by the x-ray angiographic findings in large animals.20 Using μ-CT, we have clearly established that the increase in vessel volume results predominantly from the generation of new medium-size vessels (48 to 160 μm) that are likely to be associated with the change in blood flow and tissue perfusion. Because the majority of the change in vessel volume in response to ischemia occurred in medium-size vessels, our studies are the first to demonstrate that measurement of vessel volume rather than the commonly used histological determination of capillary density is a more relevant parameter.
In a recent study, Duvall et al21 described a μ-CT analysis of vascular recovery after hindlimb ischemia in mice. Our studies have further extended the methodology and present new findings with regard to collateral development after ischemia, quantification of arteriogenesis, and the effect of therapy. Using an 8-μm voxel resolution, appropriate selection of tissue sections, and threshold settings, we were able to use a bismuth contrast agent without requiring bone decalcification. Use of high-resolution imaging and an advanced image analysis program has provided highly accurate quantification of vessel number and segmental volume. In an earlier preliminary study, we reported changes in vessel number in the mouse thigh region in response to PR-39.22 Here we show that determination of vessel volume at each thickness provides a more accurate picture of the vascular changes that enabled the detection of differences between control and PlGF-1–treated groups and the delineation of the exact vessel size modified by therapy.
As indicated earlier, remarkably, the population of vessels carrying the bulk of arterial blood was precisely the population enhanced by PlGF-1. This observation implies that specific biological signals that distinguish this population of vessels exist and that selective therapy may conceivably be developed to augment their growth. We used adenoviral PlGF-1 gene delivery to achieve sustained expression of PlGF-1. Ad-PlGF-1 was first rigorously characterized with regard to its expression and biological activity, which established a dose-dependent expression of PlGF-1 in vitro and in vivo. Adenovirally derived PlGF-1 was biologically active, as demonstrated by flt-1 activation and mobilization of Sca1+ cells.
We observed that PlGF-1 selectively increased Sca1+ cells in the circulation without affecting other cell types. This contrasts with G-CSF, which, while mediating mobilization of the Sca1+ population, also mobilizes leukocytes and moncytes.23 Because nonspecific mobilization of inflammatory cells may be undesirable for therapy,24 selective mobilization of Sca1+ cells may be superior for inducing progenitor cell mobilization. Previous studies have demonstrated expression of flt-1 in monocytes25 and its role in arteriogenesis.19 The purpose of monitoring bone marrow cells in our studies was to use it as a pharmacodynamic biomarker for in vivo PlGF-1 activity. Whether these cells also play a role in arteriogenesis is unknown.
While inducing growth of medium-size arteries, PlGF-1 did not affect small vessels, including arterioles and capillaries. In this regard, our μ-CT data are consistent with the lack of a PlGF-1 effect on capillary density reported previously.18,19 The effect of PlGF-1 in promoting medium-size, stable, and perfusing vessels may be clinically important for the treatment of ischemic coronary and peripheral vascular diseases. Although the clinical efficacy of PlGF-1 remains to be tested, the quantitative 3D μ-CT imaging data strongly support that PlGF-1 represents an important therapeutic molecule to promote large conducting vessels in patients with ischemic coronary and peripheral arterial diseases.
We thank Dr J. Anthony Ware for his support and Dr Ebo D. deMuinck for valuable suggestions on mouse models of ischemia.
Dr Michael Simons, who has served as a scientific consultant, received a nominal honorarium from Eli Lilly and Company.
Couffinhal T, Silver M, Kearney M, Sullivan A, Wotzenbichler B, Magner M, Annex B, Peters K, Isner JM. Impaired collateral vessel development associated with reduced expression of vascular endothelial growth factor in ApoE-/- mice. Circulation. 1999; 99: 3188–3198.
Rivard A, Faber JE, Silver M, Chen D, Murohara T, Kearney M, Magner M, Asahara T, Isner JM. Age-dependent impairment of angiogenesis. Circulation. 1999; 99: 111–120.
Rivard A, Berthou-Soulie L, Principe N, Kearney M, Curry C, Branellec D, Semenza GL, Isner JM. Age-dependent defect in vascular endothelial growth factor expression is associated with reduced hypoxia-inducible factor 1 activity. J Biol Chem. 2000; 275: 29643–29647.
Tamarat R, Silver J-S, Huijberts M, Benessiano J, Ebrahimian TG, Duriez M, Wautier M-P, Wautier JL, Levy BI. Blockade of advanced glycation end-product formation restores ischemia-induced angiogenesis in diabetic mice. Proc Natl Acad Sci U S A. 2003; 100: 8555–8560.
Fadini PG, Miorin M, Facco M, Bonamico S, Baesso I, Grego F, Menegolo M, Kreutzenber SV, Tiengo A, Agostini C, Avogara A. Circulating endothelial progenitor cells are reduced in peripheral vascular complications of type 2 diabetes mellitus. J Am Coll Cardiol. 2005; 45: 1449–1457.
Schaper W, Scholz D. Factors regulating arteriogenesis. Arterioscler Thromb Vasc Biol. 2003; 23: 1143–1151.
Simons M. Angiogenesis: where do we stand now? Circulation. 2005; 111: 1556–1560.
Waters RE, Terjung RL, Peters KG, Annex BH. Preclinical models of human peripheral arterial occlusive disease: implications for investigation of therapeutic agents. J Appl Physiol. 2004; 10: 773–780.
Neufeld G, Cohn T, Gengrinovitch S, Poltorak Z. Vascular endothelial growth factor (VEGF) and its receptors. FASEB J. 1999; 13: 9–22.
Selvaraj SK, Giri RK, Perelman N, Johnson C, Malik P, Kalra VK. Mechanism of monocyte activation and expression of proinflammatory cytochemokines by placenta growth factor. Blood. 2003; 102: 1515–1524.
Luttun A, Tjwa M, Moons L, Wu Y, Angelillo-Scherrer A, Lia F, Nagy A, Hooper A, Priller J, DeKeller B, Compernolle V, Daci E, Bohlen P, Dewerchin M, Herbert J-M, Fava R, Matthys P, Carmeliet G, Collen D, Dvora HF, Hicklin DJ, Carmeliet P. Revascularization of ischemic tissues by PlGF treatment and inhibition of tumor angiogenesis, arthritis and atherosclerosis by anti-Flt. Nat Med. 2002; 8: 381–840.
Pipp F, Heil M, Issbrucker K, Ziegelhoeffer T, Martin S, van den Heuvel J, Weich H, Fernandez B, Golomb G, Carmeliet P, Schaper W, Clauss M. VEGFR-1–selective VEGF homologue PlGF arteriogenic: evidence for a monocyte-mediated mechanism. Circ Res. 2003; 92: 378–385.
Duvall CL, Taylor WR, Weiss D, Guldberg RE. Quantitative microcomputed tomography analysis of collateral vessel development after ischemic injury. Am J Physiol Heart Circ Physiol. 2004; 287: H302–H310.
Tirziu D, Moodie K, Zhuang ZW, Singer K, Helisch A, Dunn JF, Li W, Singh JaiPal, Simons M. Delayed arteriogenesis in hypercholesterolemic mice. Circulation. 2005; 112: 2501–2509.
Kang HJ, Kim HS, Zhang SY, Park KW, Cho HJ, Koo BK, Kim YJ, Soo LD, Sohn DW, Han KS, Oh BH, Lee MM, Park YB. Effects of intracoronary infusion of peripheral blood stem cells mobilized with granulocyte-colony stimulating factor on left ventricular systolic function and restenosis after coronary stenting in myocardial infarction: the MAGIC cell randomized clinical trial. Lancet. 2004; 363: 751–756.
Swano A, Iwai S, Sakurai Y, Ito M, Shitara K, Nakahata T, Shibuya M. Flt-1, vascular endothelial growth factor receptor 1, is a novel cell surface marker for the lineage of monocyte-macrophage in humans. Blood. 2001; 97: 785–791.
For patients with severe occlusive arterial disease, a major medical need is the restoration of blood flow. Percutaneous transluminal angioplasty and drug-eluting stents provide highly successful device-based interventions in some of these patients. Therapeutic induction of collaterals is an attractive alternative approach that may be effective for a large number of patients. However, to date, pharmacological stimulation of therapeutic angiogenesis has not been achieved in humans. In part, this failure may be due to the use of agents that predominantly induce angiogenesis (capillary growth) rather than arteriogenesis (arterial/collateral growth). The data presented in the present study have 3 potential clinical implications: (1) Use of μ-CT for quantification of arterial vessels in the mouse model provides a direct translation of preclinical data to clinical studies. (2) We have shown that in response to ischemia, a major change occurred in vascular volume through the expansion of medium-size collaterals. This response largely contributed to the rapid restoration of flow and function. Therefore, measurement of vascular volume is a more relevant parameter than is superficial blood flow by laser Doppler methods or histological capillary density not relevant to clinical efficacy. (3) The present study has revealed a unique effect of placental growth factor-1 on medium-size arteries relevant to therapy in humans. Furthermore, a selective increase of progenitor cells in the circulation by placental growth factor suggests its potential use for pharmacological enhancement of progenitor cells that are deficient in patients with cardiovascular risk factors, including diabetes and dyslipidemia.