This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hoefer, I. E. Articles by Ozaki, C. K. Search for Related Content PubMed PubMed Citation Articles by Hoefer, I. E. Articles by Ozaki, C. K. Related Collections Angiogenesis

(Circulation. 2002;105:1639.)
© 2002 American Heart Association, Inc.

Brief Rapid Communications

Direct Evidence for Tumor Necrosis Factor- Signaling in Arteriogenesis

Imo E. Hoefer, MD; Niels van Royen, MD; John E. Rectenwald, MD; Elizabeth J. Bray, MD; Zaher Abouhamze, BS; Lyle L. Moldawer, PhD; Michiel Voskuil, MD; Jan J. Piek, MD PhD; Ivo R. Buschmann, MD; C. Keith Ozaki, MD

From the Department of Cardiology, University of Freiburg, Germany; Department of Surgery (J.E.R., E.J.B., Z.A., L.L.M., C.K.O.), University of Florida College of Medicine and the Malcolm Randall VAMC, Gainesville, Fla; and Department of Cardiology (M.V., J.J.P.), University of Amsterdam, the Netherlands.

Correspondence to Imo Hoefer, MD, Department of Cardiology, University of Freiburg Hugstetterstr. 55 D-79106, Freiburg, Germany. E-mail hoefer{at}med1.ukl.uni-freiburg.de

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background— Arteriogenesis serves as an efficient mechanism for flow restoration after arterial occlusion. This process is associated with inflammatory mediators such as tumor necrosis factor- (TNF-), although their role in arteriogenesis remains unclear. We hypothesized that arteriogenesis is reduced in mice lacking functional TNF- or p55 receptor. To test this hypothesis, we developed a novel microsphere-based murine model of hindlimb perfusion measurement.

Methods and Results— Unilateral femoral arteries of nude (n=9), TNF-^-/- (n=9), TNF- receptor p55^-/- (n=8), and p75^-/- (n=8) mice as well as their appropriate genetic background controls were occluded. The nude mice underwent laser Doppler hindlimb flux measurements preoperatively, postoperatively, and after 7 days. Seven days after ligation, all animals underwent tissue perfusion determinations using fluorescent microspheres. Laser Doppler findings confirmed acute decrease in flux with falsely normal values after 1 week. Microsphere results from control mice showed perfusion restoration to values 50% of normal within 7 days. TNF-^-/- mice demonstrated a significant reduction (45.1%) in collateral artery perfusion compared with controls (TNF-^-/- 22.4±5.1% versus B6x129 49.7±9.3%; P<0.01). p55^-/- mice exhibited an almost identical 45.8% reduction in collateral artery formation (p55^-/- 28.3±4.3% versus C57BL/6J 61.8±9.1%; P<0.01), whereas p75^-/- mice were equivalent to controls (p75^-/- 54.5±5.5%; P=0.13).

Conclusions— Microsphere techniques in mice offer a tool for the molecular dissection of arteriogenesis mechanisms. These results suggest that TNF- positively modulates arteriogenesis probably via signaling through its p55 receptor.

Key Words: collateral circulation • inflammation • microspheres • blood flow • remodeling

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Arteriogenesis (remodeling of preexistent arteriolar collateral networks into large collateral conductance arteries) serves as the most efficient mechanism to restore flow after arterial occlusion.¹ Perivascular inflammation and monocyte/macrophage accumulation accompany this process.²

Tumor necrosis factor- (TNF-) is an important proximal mediator of inflammation. Most prototypical inflammatory TNF- signaling occurs by TNF- binding to its p55 receptor.³ Previous studies have associated TNF- with arteriogenesis.² We hypothesized that arteriogenesis proceeds by way of TNF- signaling via the p55 receptor. Specifically, we hypothesized that arteriogenesis is reduced in mice lacking functional TNF- or p55 receptor. We developed and validated a novel model of perfusion measurement in the mouse hindlimb to test this hypothesis.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Animal Model
This study conforms to the Guide for the Care and Use of Laboratory Animals (NIH publication No. 85-23, revised 1996). Nine athymic nude, 8 TNF- receptor p55^-/-3,4 (N5 backcross onto C57BL/6J, maintained by brother/sister matings) and 9 TNF-^-/-5 (on B6x129 F1 background, maintained by sister/brother matings) (both provided by Dr Carl K. Edwards III, Amgen, Inc, Thousand Oaks, Calif), 8 TNF- receptor p75^-/-, 7 B6x129, and 9 C57BL/6J (Jackson Laboratory, Bar Harbor, Maine) mice underwent unilateral femoral artery ligation. The right femoral artery was ligated immediately distal to the inguinal ligament. Because collateral arteries develop from preexisting arteriolar connections spanning from the profunda femoris and circumflexa femoris to the genualis and saphena parva arteries,⁴ the femoral artery was not excised to leave these vessels intact. In rabbits, this technique does not lead to ischemia in the region of collateral artery growth or distal to the site of ligation.⁵ Wounds were closed and animals recovered.

Laser Doppler Technique
Nude mice, lacking pigmentation and fur that obscure signal transmission, were anesthetized, and flux was evaluated using a laser Doppler (LD) device. Fur removal of normal mice only leads to slight improvement because of changes in skin circulation attributable to mechanical or chemical irritation. Hindlimbs were divided into 3 zones: complete hindlimb, femoral artery tract, and region of collateral artery growth in the thigh. Flux was measured directly before ligation, acutely after ligation, and 7 days later.

Microsphere Technique
Seven days after ligation, mice were anesthetized for microsphere-based flow calculations. To ensure perfusion of both hindlimbs and optimal distribution of the fluorescent microspheres, the abdominal aorta was cannulated with a polyethylene catheter (inner diameter, 0.58 mm; outer diameter, 0.96 mm). After stabilization of systemic pressure, both legs were perfused at 4 different pressures (70, 80, 90, and 100 mm Hg) with saline and continuous infusion of adenosine (1.0 mg/kg per min, Sigma) to achieve maximal vasodilatation. At each pressure level, microspheres with a different fluorescent dye (red, blue-green, orange, or yellow-green; diameter, 15 µm; Molecular Probes) were thoroughly mixed and injected into the perfusion system. At the distal end of the perfusion system, a reference sample was withdrawn for 2 minutes at a rate of 0.2 mL/min for each microsphere dye. By placing the catheter in the proximal descending aorta, using the described technique, renal perfusion did not show any significant difference between right and left kidneys. This is consistent with our data from mice that did not undergo femoral ligation, showing no difference in tissue perfusion between both hindlimbs.

Quantification of microspheres in muscle was performed as previously described in rabbits.⁴ The muscles were dissected from the leg and divided into groups: anterior thigh, posterior thigh, and calf muscles. Each muscle sample was weighed, homogenized, and digested with SDS/proteinase solution for FACS analysis and microsphere counting.⁴

Histology
An additional 4 TNF-^-/-, 4 B6x129, 4 p55^-/-, and 4 C57BL/6J mice underwent unilateral femoral ligation. Seven days later, muscles from both hindlimbs were harvested and grouped by anterior thigh, posterior thigh, and calf. Samples then underwent formalin fixation and H&E staining.

Statistical Analysis
Differences between genotypes were assessed using unpaired Student’s t test.

	Results

Top Abstract Introduction Methods Results Discussion References

 
No animal suffered gangrene or gross impairment of hindlimb function.

LD Versus Microsphere Techniques
Using LD, immediately before femoral ligation there was no significant difference between the right and the left hindlimb in any of the regions (Figure 1). Acutely after ligation, flux decreased in the occluded hindlimb when measured over either the femoral artery tract distal to the ligation site or the complete hindlimb (P<0.05). In contrast, flow in the collateral artery region showed no significant difference compared with the nonoccluded hindlimb. Acutely after ligation, perfusion distal to the ligation fell to 20.2% of baseline. Seven days after femoral occlusion, the LD measurements suggested complete flux restoration in the whole hindlimb and in the distal femoral artery relative to values of nonoccluded hindlimbs. These results are in contrast to the findings using the microsphere technique, which showed only 58.9±7.3% (occluded/nonoccluded) hindlimb flow restoration. Flux in the collateral artery region at 7 days showed a significant increase compared with the nonoccluded hindlimbs (P<0.05).

View larger version (32K): [in this window] [in a new window]   Figure 1. LD measurements in mouse hindlimbs before, acutely after, and 7 days after femoral artery ligation. Flux was evaluated in three different regions: whole hindlimb, femoral artery tract, and thigh, the region of collateral artery growth.

TNF-^-/- and TNF- Receptor^-/- Mice
Seven days after femoral ligation, B6x129 wild-type mice showed a 49.6±9.3% flow restoration in the occluded hindlimbs compared with the nonoccluded hindlimbs (Figure 2). TNF-^-/- mice, in contrast, showed only a 22.4±5.1% flow restoration. Thus, the arteriogenic response of the TNF-^-/- mice was only 45.1% (P<0.01) of control animals.

View larger version (28K): [in this window] [in a new window]   Figure 2. Microsphere perfusion measurements. Collateral artery flow was evaluated with fluorescent microspheres 7 days after femoral artery ligation. TNF-/- mice as well as p55-/- mice show a significant reduction, whereas p75-/- mice do not show any significant difference compared with their matching controls. *P<0.05. ns indicates nonsignificant.

C57BL/6J mice demonstrated a 61.8±9.1% flow restoration 7 days after femoral ligation. In the p55^-/- mice, this restoration was 28.3±4.3%. Thus, the arteriogenic response of p55^-/- mice was only 45.8% (P<0.01) of controls. p75^-/- mice (54.5±5.5%), however, did not show any significant difference compared with the C57BL/6J controls (P=0.13).

Histology
There was no observed difference in the baseline vascular architecture between genotypes (unligated hindlimb). There was no evidence for increased tissue destruction in the transgenic mice compared with their WT controls (ligated hindlimb).

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
We report a novel murine model of arteriogenesis using hindlimb perfusion measurements after femoral artery ligation. In addition, we show the limitations of LD to survey arteriogenesis compared with the presented model. Most prior murine studies relied on LD approaches.^6,7 Blood flow evaluation using LD offers the advantage of serial measurements. We used LD to evaluate flux and thus estimate arteriogenesis after femoral ligation. Acutely after femoral occlusion, flux decreased in the complete hindlimb and in the femoral artery distal to the ligation. Seven days after femoral ligation, LD measurements in these regions were restored to values not significantly different from normal values. However, based on the results using microspheres, these were probably false normal values. With microspheres, flow was only restored to 60% of normal after 1 week. Radioactive and fluorescent microspheres have been widely used to investigate blood flow and tissue perfusion in various models and constitute the gold standard of perfusion measurements.⁸

LD results primarily reflect superficial blood flow, ie, skin circulation. Blood supply of the skin, however, does not necessarily parallel the situation in deeper structures, such as muscle. Collateral arteries lie within these deeper tissues. When examining arteriogenesis in vivo, low penetration devices are thus of limited value. A final critical limitation of LD also applies to high-penetration systems. To estimate changes in collateral artery flow, the vasculature has to be challenged with high flow under conditions of maximal vasodilation. Such hemodynamic conditions cannot be reached without animal instrumentation. In our microsphere system, we give large doses of adenosine to ensure maximal vasodilation. There is no reason to suspect that any of the strains would have differential vasodilatory responses to such large doses of adenosine.

The remodeling of preexistent arteriolar collateral networks into large collateral conductance arteries that we observed after acute femoral ligation is distinctly different than angiogenesis, and our methods are designed to quantify the former specifically. After femoral ligation, we use 15-µm microspheres, which are not able to pass into the distal thigh and calf muscles through newly formed 7-µm capillaries in the thigh. Passage of the 15-µm microspheres from the aorta to the distal thigh and calf requires larger-conductance >30-µm collateral vessels. Capillary sprouting does not translate into improved flow, which depends on the growth of larger collateral vessels.¹

TNF- has been localized to the macrophages of growing collateral arteries.² Using reverse transcriptase–polymerase chain reaction, we find TNF- mRNA in the region of collateral artery growth 2 days after femoral artery occlusion in the mouse (C. Keith Ozaki, unpublished data, 2000). However, the role of TNF- in arteriogenesis is uncertain—does it serve as a pivotal modulator or innocent bystander? Our in vivo results suggest that the role of TNF- is probably that of a positive modulator for arteriogenesis. A recent in vivo study demonstrated that TNF- positively modulates occlusive arterial lesions that form in response to low shear stress.⁹ Anti–TNF- approaches attenuate intimal thickening after balloon angioplasty,¹⁰ but unfortunately such approaches may also abrogate host arteriogenic responses. As it becomes apparent that multiple vascular processes proceed by way of inflammatory mediators, ¹¹ additional research must explore the mechanisms for specificity (occlusive versus enlargement) in these pathways.

In summary, this study suggests superiority of fluorescent microsphere perfusion studies over present LD techniques to survey arteriogenesis in mice. This novel method provides direct in vivo evidence that TNF- positively modulates arteriogenesis, probably signaling via the p55 receptor.

	Acknowledgments

 
This work was supported by the National Institutes of Health, VA Merit Review, Lifeline Foundation, The William J. von Liebig Foundation, and Volkswagen Foundation.

Received December 17, 2001; revision received February 19, 2002; accepted February 19, 2002.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Hershey JC, Baskin EP, Glass JD, et al. Revascularization in the rabbit hindlimb: dissociation between capillary sprouting and arteriogenesis. Cardivasc Res. 2001; 49: 618–625.[Abstract/Free Full Text]
   
2. Arras M, Ito WD, Scholz D, et al. Monocyte activation in angiogenesis and collateral growth in the rabbit hindlimb. J Clin Invest. 1998; 101: 40–50.[Medline] [Order article via Infotrieve]
   
3. Rothe J, Lesslauer W, Lotscher H, et al. Mice lacking the tumor necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature. 1993; 364: 798–802.[CrossRef][Medline] [Order article via Infotrieve]
   
4. Hoefer I, van Royen N, Buschmann I, et al. Time course of arteriogenesis following femoral artery occlusion in the rabbit. Cardiovasc Res. 2001; 49: 609–617.[Abstract/Free Full Text]
   
5. Deindl E, Buschmann I, Hoefer I, et al. Role of ischemia and of hypoxia-inducible genes in arteriogenesis after femoral artery occlusion in the rabbit. Circ Res. 2001; 89: 779–786.[Abstract/Free Full Text]
   
6. Duan J, Murohara T, Ikeda H, et al. Hyperhomocysteniemia impairs angiogenesis in response to hindlimb ischemia. Arterioscler Thromb Vasc Biol. 2000; 20: 2579–2585.[Abstract/Free Full Text]
   
7. Couffinhal T, Silver M, Kearney M, et al. Impaired collateral vessel development associated with reduced expression of vascular endothelial growth factor in ApoE^-/- mice. Circulation. 1999; 99: 3188–3198.[Abstract/Free Full Text]
   
8. Unger E. Experimental evaluation of coronary collateral development. Cardiovasc Res. 2001; 49: 497–506.[Abstract/Free Full Text]
   
9. Rectenwald J, Moldawer L, Huber T, et al. Direct evidence for cytokine involvement in neointimal hyperplasia. Circulation. 2000; 102: 1697–1702.[Abstract/Free Full Text]
   
10. Krasinski K, Spyridopoulos I, Kearney M, et al. In vivo blockade of tumor necrosis factor accelerates functional endothelial recovery after balloon angioplasty. Circulation. 2001; 104: 1754–1756.[Abstract/Free Full Text]
    
11. Ross R. Atherosclerosis: an inflammatory disease. N Engl J Med. 1999; 340: 115–126.[Free Full Text]
    

This article has been cited by other articles:

				D Versteeg, I E Hoefer, A H Schoneveld, D P V de Kleijn, E Busser, C Strijder, M Emons, P R Stella, P A Doevendans, and G Pasterkamp Monocyte toll-like receptor 2 and 4 responses and expression following percutaneous coronary intervention: association with lesion stenosis and fractional flow reserve Heart, June 1, 2008; 94(6): 770 - 776. [Abstract] [Full Text] [PDF]

				W. Schaper Prevention of Tissue Death by Killer Cells?: The Role of the Immune System in Arteriogenesis Arterioscler. Thromb. Vasc. Biol., November 1, 2007; 27(11): 2273 - 2274. [Full Text] [PDF]

				B. Mees, S. Wagner, E. Ninci, S. Tribulova, S. Martin, R. van Haperen, S. Kostin, M. Heil, R. de Crom, and W. Schaper Endothelial Nitric Oxide Synthase Activity Is Essential for Vasodilation During Blood Flow Recovery but not for Arteriogenesis Arterioscler. Thromb. Vasc. Biol., September 1, 2007; 27(9): 1926 - 1933. [Abstract] [Full Text] [PDF]

				D. Chalothorn, J. A. Clayton, H. Zhang, D. Pomp, and J. E. Faber Collateral density, remodeling, and VEGF-A expression differ widely between mouse strains Physiol Genomics, July 18, 2007; 30(2): 179 - 191. [Abstract] [Full Text] [PDF]

				N. R. Ferreri Estrogen-TNF interactions and vascular inflammation Am J Physiol Heart Circ Physiol, June 1, 2007; 292(6): H2566 - H2569. [Full Text] [PDF]

				D. A. Goukassian, G. Qin, C. Dolan, T. Murayama, M. Silver, C. Curry, E. Eaton, C. Luedemann, H. Ma, T. Asahara, et al. Tumor Necrosis Factor-{alpha} Receptor p75 Is Required in Ischemia-Induced Neovascularization Circulation, February 13, 2007; 115(6): 752 - 762. [Abstract] [Full Text] [PDF]

				D. Luo, Y. Luo, Y. He, H. Zhang, R. Zhang, X. Li, W. L. Dobrucki, A. J. Sinusas, W. C. Sessa, and W. Min Differential Functions of Tumor Necrosis Factor Receptor 1 and 2 Signaling in Ischemia-Mediated Arteriogenesis and Angiogenesis Am. J. Pathol., November 1, 2006; 169(5): 1886 - 1898. [Abstract] [Full Text] [PDF]

				C. E. Bergmann, I. E. Hoefer, B. Meder, H. Roth, N. van Royen, S. M. Breit, M. M. Jost, S. Aharinejad, S. Hartmann, and I. R. Buschmann Arteriogenesis depends on circulating monocytes and macrophage accumulation and is severely depressed in op/op mice J. Leukoc. Biol., July 1, 2006; 80(1): 59 - 65. [Abstract] [Full Text] [PDF]

				S. Grundmann, I. Hoefer, S. Ulusans, N. van Royen, S. H. Schirmer, C. K. Ozaki, C. Bode, J. J. Piek, and I. Buschmann Anti-tumor necrosis factor-{alpha} therapies attenuate adaptive arteriogenesis in the rabbit Am J Physiol Heart Circ Physiol, October 1, 2005; 289(4): H1497 - H1505. [Abstract] [Full Text] [PDF]

				I. E. Hoefer, S. Grundmann, S. Schirmer, N. van Royen, B. Meder, C. Bode, J. J. Piek, and I. R. Buschmann Aspirin, But Not Clopidogrel, Reduces Collateral Conductance in a Rabbit Model of Femoral Artery Occlusion J. Am. Coll. Cardiol., September 20, 2005; 46(6): 994 - 1001. [Abstract] [Full Text] [PDF]

				D. Chalothorn, S. M. Moore, H. Zhang, S. W. Sunnarborg, D. C. Lee, and J. E. Faber Heparin-Binding Epidermal Growth Factor-Like Growth Factor, Collateral Vessel Development, and Angiogenesis in Skeletal Muscle Ischemia Arterioscler. Thromb. Vasc. Biol., September 1, 2005; 25(9): 1884 - 1890. [Abstract] [Full Text] [PDF]

				N. van Royen, S. H. Schirmer, B. Atasever, C. Y.H. Behrens, D. Ubbink, E. E. Buschmann, M. Voskuil, P. Bot, I. Hoefer, R. O. Schlingemann, et al. START Trial: A Pilot Study on STimulation of ARTeriogenesis Using Subcutaneous Application of Granulocyte-Macrophage Colony-Stimulating Factor as a New Treatment for Peripheral Vascular Disease Circulation, August 16, 2005; 112(7): 1040 - 1046. [Abstract] [Full Text] [PDF]

				T. Tojo, M. Ushio-Fukai, M. Yamaoka-Tojo, S. Ikeda, N. Patrushev, and R. W. Alexander Role of gp91phox (Nox2)-Containing NAD(P)H Oxidase in Angiogenesis in Response to Hindlimb Ischemia Circulation, May 10, 2005; 111(18): 2347 - 2355. [Abstract] [Full Text] [PDF]

				M. Voskuil, I. E Hoefer, N. van Royen, J. Hua, S. de Graaf, C. Bode, I. R Buschmann, and J. J Piek Abnormal monocyte recruitment and collateral artery formation in monocyte chemoattractant protein-1 deficient mice Vascular Medicine, November 1, 2004; 9(4): 287 - 292. [Abstract] [PDF]

				G. S. Werner, E. Jandt, A. Krack, G. Schwarz, O. Mutschke, F. Kuethe, M. Ferrari, and H. R. Figulla Growth Factors in the Collateral Circulation of Chronic Total Coronary Occlusions: Relation to Duration of Occlusion and Collateral Function Circulation, October 5, 2004; 110(14): 1940 - 1945. [Abstract] [Full Text] [PDF]

				M. Heil and W. Schaper Influence of Mechanical, Cellular, and Molecular Factors on Collateral Artery Growth (Arteriogenesis) Circ. Res., September 3, 2004; 95(5): 449 - 458. [Abstract] [Full Text] [PDF]

				S. E. Epstein, E. Stabile, T. Kinnaird, C. W. Lee, L. Clavijo, and M. S. Burnett Janus Phenomenon: The Interrelated Tradeoffs Inherent in Therapies Designed to Enhance Collateral Formation and Those Designed to Inhibit Atherogenesis Circulation, June 15, 2004; 109(23): 2826 - 2831. [Full Text] [PDF]

				I. E. Hoefer, N. van Royen, J. E. Rectenwald, E. Deindl, J. Hua, M. Jost, S. Grundmann, M. Voskuil, C. K. Ozaki, J. J. Piek, et al. Arteriogenesis Proceeds via ICAM-1/Mac-1- Mediated Mechanisms Circ. Res., May 14, 2004; 94(9): 1179 - 1185. [Abstract] [Full Text] [PDF]

				M.H. Tayebjee, G.Y.H. Lip, and R.J. MacFadyen Collateralization and the response to obstruction of epicardial coronary arteries QJM, May 1, 2004; 97(5): 259 - 272. [Abstract] [Full Text] [PDF]

				N. van Royen, M. Voskuil, I. Hoefer, M. Jost, S. de Graaf, F. Hedwig, J.-P. Andert, T.A.M. Wormhoudt, J. Hua, S. Hartmann, et al. CD44 Regulates Arteriogenesis in Mice and Is Differentially Expressed in Patients With Poor and Good Collateralization Circulation, April 6, 2004; 109(13): 1647 - 1652. [Abstract] [Full Text] [PDF]

				W. Schaper and D. Scholz Factors Regulating Arteriogenesis Arterioscler. Thromb. Vasc. Biol., July 1, 2003; 23(7): 1143 - 1151. [Abstract] [Full Text] [PDF]

				N. van Royen, I. Hoefer, M. Bottinger, J. Hua, S. Grundmann, M. Voskuil, C. Bode, W. Schaper, I. Buschmann, and J.J. Piek Local Monocyte Chemoattractant Protein-1 Therapy Increases Collateral Artery Formation in Apolipoprotein E-Deficient Mice but Induces Systemic Monocytic CD11b Expression, Neointimal Formation, and Plaque Progression Circ. Res., February 7, 2003; 92(2): 218 - 225. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hoefer, I. E. Articles by Ozaki, C. K. Search for Related Content PubMed PubMed Citation Articles by Hoefer, I. E. Articles by Ozaki, C. K. Related Collections Angiogenesis