Novel Tissue-Engineered Biodegradable Material for Reconstruction of Vascular Wall

Transcription

1 Novel Tissue-Engineered Biodegradable Material for Reconstruction of Vascular Wall Shigemitsu Iwai, MD, Yoshiki Sawa, MD, Satoshi Taketani, MD, Kei Torikai, MD, Koichiro Hirakawa, MS, and Hikaru Matsuda, MD Division of Cardiovascular Surgery, Department of Surgery, Osaka University Graduate School of Medicine, Osaka, and Senko Medical Instrument Manufacturing Co, Ltd, Tokyo, Japan Background. To solve several problems with artificial grafts, we sought to develop a novel bioengineered material that can promote tissue regeneration without ex vivo cell seeding and that has sufficient durability to be used for artery reconstruction. Here, we tested whether this biodegradable material could accelerate the in situ regeneration of autologous cardiovascular tissue, especially of the arterial wall, in various models of cardiovascular surgeries. Methods. The tissue-engineered patch was fabricated by compounding a collagen-microsponge with a biodegradable polymeric scaffold composed of polyglycolic acid knitted mesh, reinforced on the outside with woven polylactic acid. Tissue-engineered patches without precellularization were grafted into the porcine descending aorta (n 5), the porcine pulmonary arterial trunk (n 8), or the canine right ventricular outflow tract (as the large graft model; n 4). Histologic and biochemical assessments were performed 1, 2, and 6 months after the implantation. Results. There was no thrombus formation in any animal. Two months after grafting, all the grafts showed good in situ cellularization by hematoxylin/eosin and immunostaining. The quantification of the cell population by polymerase chain reaction showed a large number of endothelial and smooth muscle cells 2 months after implantation. In the large graft model, the architecture of the patch was similar to that of native tissue 6 months after implantation. Conclusions. A tissue-engineered patch made of our biodegradable polymer and collagen-microsponge provided good in situ regeneration at both the venous and arterial wall, suggesting that this patch can be used as a novel surgical material for the repair of the cardiovascular system. (Ann Thorac Surg 2005;80:1821 8) 2005 by The Society of Thoracic Surgeons Tissue engineering represents a promising approach to the in vitro creation of living, autologous tissue replacements with the potential to grow or be repaired or remodeled [1]. Particularly for the surgical treatment of congenital heart disease, there is a substantial need for such implantation materials. Recently, a biodegradable material for cardiovascular surgery was developed and is being used clinically [2 5]. However, its usefulness for reconstructing the arterial wall has not yet been demonstrated. Moreover, the ex vivo cell-seeding pretreatment ( precellularization ) that was reported previously, is complicated, invasive, and can lead to infection [2 4]. Therefore, the development of a biodegradable material that allows good in situ cellularization without precellularization with sufficient durability for arterial reconstruction is an important strategy. Recently, a novel tissue-engineered material made of a biodegradable polymer and collagen-microsponge was reported that had good cell-attachment properties because of its cellular affinity and porous three-dimensional structure [6]. We recently reported that this tissueengineered prosthesis showed good in situ cellularization and a sufficient durability for vascular reconstruction, and its engraftment led to the regeneration of autologous vessel tissue in a dog model [7]. In the present study, we evaluated whether this newly developed tissue-engineered material could, even without precellularization, accelerate in situ cellularization by inducing the proliferation and differentiation of endothelial and smooth muscle progenitor cells after implantation in a large animal model of arterial wall reconstruction that could be used to simulate various cardiovascular surgeries. Material and Methods Scaffold Design The tissue-engineered patch (TEP) was fabricated by compounding a collagen-microsponge with a biodegradable polymeric scaffold composed of a polyglycolic acid (PGA) knitted mesh that was reinforced on the outside with woven polylactic acid (PLA) (Fig 1A). The PGA/PLA mesh was provided by Senko Medical Instrument Mfg Co, Ltd (Tokyo, Japan). The techniques for preparing the TEP with the collagen-microsponge were described in Accepted for publication March 28, Address correspondence to Dr Sawa, 2-2 Yamada-oka, Suita, Osaka, , Japan; sawa@surg1.med.osaka-u.ac.jp. Dr Hirakawa discloses a financial relationship with Senko Medical Instrument Manufacturing Co, Ltd by The Society of Thoracic Surgeons /05/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 1822 IWAI ET AL Ann Thorac Surg TISSUE-ENGINEERED GRAFT 2005;80: Fig 1. Formation of a biodegradable scaffold reinforced with woven polylactic acid mesh (arrow) cross-linked with collagen-microsponge (A). Scanning electron microscopy image of the tissue-engineered patch shows the uniformly distributed and interconnected pore structure (pore size m) of the collagen-microsponge (B) (magnification 40 ). detail by Chen and colleagues [6]. Briefly, the PGA/PLA mesh was immersed in a bovine type I collagen 0.5% solution (Koken Co, Ltd, Tokyo, Japan). It was then frozen at 80 C and lyophilized under vacuum to allow the collagen-microsponges to form. The collagenmicrosponges were further cross-linked by treatment with glutaraldehyde vapor. The completed patch was 800- m thick. A scanning electron microscopy (SEM) image of the patch is shown in Figure 1B. In Vitro Study To verify the cellular affinity and proliferation of the collagen-microsponge, three sets of PGA mesh (10 10 mm), including collagen-microsponge (n 5), a simple collagen-coat, or no collagen (n 5), were cultured dynamically with seeded endothelial cells ( cells) from the canine vena saphena magna. The numbers of attached cells were determined by the 3-[4.5 dimethylthiazol 2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay (DO JINDO Laboratories, Japan) on experimental days 1 and 3. In Vivo Study To examine the TEP in various situations, three experimental groups were used for TEP implantation. One was a porcine model (body weight, 15 kg) in which the TEP was implanted into the descending aorta for histologic evaluation and to test the patch s durability in the systemic circulation. Anesthesia was induced with an intramuscular bolus infusion of 3 mg/kg ketamine and 3 mg/kg barbiturate, and maintained by intravenous continuous infusion of propofol (5 mg/kg/hour). The heart was approached by a left anterolateral thoracotomy, and the chest was entered through the fourth intercostal space. The descending aorta was partially clamped and a TEP (20 15 mm) without precellularization was implanted using 5-0 monofilament running sutures. The TEP was explanted 2 months after implantation (n 5) and evaluated by further experiments. Another model was implantation of the TEP into the porcine (body weight, 15 kg) pulmonary arterial trunk, which we used to evaluate early endothelialization and histologic changes. Using the above surgical approach, the pulmonary arterial trunk was partially clamped. A TEP (20 15 mm) without precellularization was implanted using 5-0 monofilament running sutures. The TEP was explanted either one (n 3) or two (n 5) months after implantation. The explanted TEP was evaluated by further experiments. The final model was implantation into the canine right ventricular outflow tract to evaluate a larger size graft and midterm follow-up. Mongrel dogs (body weight, 20 kg) (n 4) were used. Anesthesia was induced with an intravenous bolus infusion of 3 mg/kg ketamine and 3 mg/kg sodium barbiturate, and maintained by inhalation of sevofluorane. Normothermic cardiopulmonary bypass was performed by means of the femoral artery and right atrial venous cannulation. With the heart beating, the pulmonary trunk to the right ventricular outflow tract was incised longitudinally. A large TEP (50 25 mm) without precellularization was implanted using 5-0 monofilament running sutures. Morphologic changes in the large patch model with right ventricular outflow tract reconstruction were evaluated by right ventricular angiography 6 months after the implantation. After the termination, the TEP was explanted and evaluated by further experiments. All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Histologic and Immunohistologic Examination Explanted tissue specimens were studied as hematoxylin/eosin, elastica van Gieson, and von Kossa stained paraffin or immunostained frozen sections. The antibodies for immunohistochemistry included monoclonal antibodies against CD31 (clone JC/70A, DAKO, Carpinteria, CA), -smooth muscle actin (clone HHF35, DAKO), and

3 Ann Thorac Surg IWAI ET AL 2005;80: TISSUE-ENGINEERED GRAFT Table 1. The Primers and a for Quantitative Reverse Transcription-Polymerase Chain Reaction GAPDH Forward primer 5=-CTGCACCACCAACTGCTTAGC-3= 5=-GCCATGCCAGTGAGCTTCC-3= 5=-CCTGGCCAAGGTCATCCATGACCACTTC-3= vwf Forward primer 5=-ATGGAGTACACGGCTTTGCTG-3= 5=-CAGGCACATGCTGTGACACAT-3= 5=-ATGAATGTCCACCTCCTCTTCAGACCGG-3= VEGF Forward primer 5=-GACGTCTACCAGCGCAGCTACT-3= 5=-TTTGATCCGCATAATCTGCATG-3= 5=-TTCCAGGAGTACCCCGATGAGATCGA-3= SM22 Forward primer 5=-GCTCCATTTGCTTGAAGACCAT-3= 5=-GTAATGCAGTGTGGCCCTGA-3= 5=-CTCAAAATCACGCCGTTCTTCAGCCA-3= ACTA2 Forward primer 5=-TATGTGGCTATTCAGGCGGTG-3= 5=-AGGATCTTCATGAGGTAGTCGGTG-3= 5=-ATGGTGTCACCCACAACGTGCCCATT-3= Vimentin Forward primer 5=-AGGTGGCAATCTCAATGTCGA-3= 5=-AAATGAGTACCGGAGACAGGTGC-3= 5=-CTCTTCCATTTCCCGCATCTGGCGTT-3= 1823 ACTA2 vascular smooth muscle -actin 2; GAPDH glyceraldehyde-3-phosphate dehydrogenase; SM22 smooth muscle 22 ; VEGF vascular endothelial growth factor; vwf von Willebrand factor. vimentin (clone V9, DAKO), and a polyclonal antibody against von Willebrand factor (Factor VIII related antigen, rabbit, N1505, DAKO). Quantitative Reverse Transcription-Polymerase Chain Reaction (RT-PCR) To quantify the cellular population in the explanted TEPs 2 months after implantation, the expression of von Willebrand factor (vwf), vascular endothelial growth factor (VEGF), vascular smooth muscle -actin 2 (ACTA2), smooth muscle 22 (SM22 ), and vimentin were determined with quantitative real-time RT-PCR, as described previously [8]. The primers and a fluorogenic probe are shown in Table 1. The quantitative real-time RT-PCRs were carried out in quadruplicate using the TaqMan RT-PCR kit (ABI) in the Applied Biosystems 7700 sequence detector system (Applied Biosystems, Foster City, CA), according to the manufacturer s instructions. We first determined the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) messenger RNA (mrna) levels in each sample by RT-PCR using differentially diluted, heat-treated, single-stranded complementary DNA (cdna) mixtures under different PCR cycle conditions, and then amplified the target cdnas using defined amounts of heat-denatured, single-stranded cdna mixtures containing an equal amount of GAPDH cdna. Standard curves covering the range from 10 2 to 10 7 copies were constructed using the in vitro synthesized transcripts. For an assay to be considered successful, the correlation coefficient of the standard curve had to be greater than The levels of these mrnas were compared with those of the native pulmonary artery and descending aorta from the same animal and expressed as the percentage of the native content. Biochemical Examination A 4-hydroxyproline assay was used to measure the collagen content in the explanted TEPs 2 months after implantation. The elastin content was quantified by determining the dry weight of the insoluble material after delipidation in acetone-diethyl ether and solubilization in 0.1 N NaOH. The levels of these extracellular components of the explanted TEPs were compared with those of the native vascular tissue from the same animal and expressed as the percentage of the native content. Statistical Analysis All results were expressed as the mean SD and analyzed using a statistical analysis software package (Stat-View version 5.0; Abacus Concept, Inc, Berkeley, CA). One-way analysis of variance, Bonferroni s test, and paired t test were used to analyze the data. A pvalue of less than 0.05 was considered statistically significant. In quantitative RT-PCR and biochemical examination, statistical analyses were performed to compare explanted TEP tissue with native vascular tissue. Results In Vitro Study On experimental day 1, the number of seeded cells attached to the PGA mesh with the collagenmicrosponge was significantly higher than to the PGA mesh alone, by the MTT assay (PGA mesh with collagenmicrosponge, ; PGA mesh with collagen-coat, ; PGA mesh alone, ) (p 0.05). On day 3, the proliferation of cells seeded onto the PGA mesh with collagen-microsponge was significantly higher than the proliferation on the PGA mesh with a

4 1824 IWAI ET AL Ann Thorac Surg TISSUE-ENGINEERED GRAFT 2005;80: ) (Fig 2). These results proved that the collagenmicrosponge had a higher cellular affinity than did the collagen-coated PGA mesh or the polymeric scaffold alone. Fig 2. The attachment and proliferation of seeded cells cultured on the polyglycolic acid (PGA) mesh with collagen-microsponge were significantly higher than when the culture was done on PGA mesh with a simple collagen-coat or on PGA mesh alone. * p 0.05 versus PGA mesh only; ** p 0.05 versus PGA mesh with collagen-coat and PGA mesh only; PGA mesh only; PGA mesh with collagencoat; ΠPGA mesh with collagen-microsponge. (MTT 3-[4.5 dimethylthiazol 2-yl]-2,5-diphenyltetrazolium bromide.) collagen-coat or the PGA mesh alone (PGA mesh with collagen-microsponge, ; PGA mesh with collagen-coat, ; PGA mesh alone, ) (p In Vivo Study All the animals survived the operative procedure for the in vivo experiments. Macroscopic inspection of the explanted TEP showed no intimal thickening in TEP and no aneurism formation, even in the aortic implantation model (Figs 3A and 3B). No patch showed thrombus formation on the internal surface of the TEP, even though we did not provide anticoagulant therapy (Figs 3A and 4A). Right ventricular angiography of the large patch model performed 6 months after implantation showed no evidence of stenosis or aneurismal change (Fig 5). Histologic Findings In the descending aortic model 2 months after implantation, the histologic examination showed almost complete absorption of the PGA polymer and good recellularization, as assessed by hematoxylin eosin staining. The CD31 and -smooth muscle actin staining showed uniform cellularization of endothelial cells and SMCs (Figs 3C and 3D). In the pulmonary arterial model, hematoxylin-eosin staining showed excellent in situ cellularization of the TEP 1 month after implantation (Fig 4C [1 month]), and Fig 3. Macroscopic and histologic findings of the descending aorta model, 2 months after implantation. Macroscopically, there was no aneurysm formation in any animal (A) and (B). The CD31 (C) and -smooth muscle actin staining (D) showed uniform cellularization of the endothelial and smooth muscle cells (magnification 100 ).

5 Ann Thorac Surg IWAI ET AL 2005;80: TISSUE-ENGINEERED GRAFT 1825 Fig 4. Macroscopic (A and B) and histologic (C E) findings of the pulmonary trunk model 1 month and 2 months after implantation. Macroscopically, there was no thrombus formation or neointimal thickness in any sample (A and B). Hematoxylin-eosin staining (C) (magnification 100 ) showed in situ cellularization by 1 month (1 month [C]). The CD31 staining (D) (100 ) showed a monolayer of endothelial cells by 1 month (1 month [D]). Alpha-smooth muscle actin staining (E) (100 ) revealed the parallel alignment of smooth muscle cells at 2 months (2 months [E]). CD31 staining showed a monolayer of endothelial cells covering the surface of the TEP (Fig 4D [1 month]). Alpha-smooth muscle actin staining revealed a parallel alignment of smooth muscle cells (SMCs) 2 months after implantation (Fig 4E [2 months]). In the large patch model 6 months after implantation, no calcification was detected by von Kossa staining, and the architecture of the implanted TEP was similar to that Fig 5. Right ventricular angiography of the large patch model. Arrow indicates tissue engineered patch. There was no stenosis or aneurysmal change from the right ventricular outflow tract to the pulmonary artery. of the native pulmonary artery wall (Fig 6A). Scanning electron microscope observations showed that the luminal surface of the TEP was uniformly covered by a confluent endothelium (Fig 6B). Elastica van Gieson staining revealed the expansion of native tissue into the TEP (Fig 6C). Factor VIII staining showed a number of microvessels in the regenerated tissue (Fig 6D). Quantitative RT-PCR The expression of vwf mrna in the TEPs was comparable with that in the native vascular tissue in both the porcine descending aortic and pulmonary arterial implantation models, even 2 months after implantation (descending aortic model, % of native tissue [p 0.26]; pulmonary arterial model, % of native tissue [p 0.39]). The VEGF mrna expression in both groups was higher than that of native vascular tissue (descending aortic model, % of native tissue [p 0.13]; pulmonary arterial model, % of native tissue [p 0.12]). This may have been the result of angiogenesis in the TEP. The ACTA2 and SM22 expression were also at a level similar to that of the native tissue 2 months after implantation (descending aortic model, ACTA2, 84 15% of native tissue [p 0.16]; pulmonary arterial model, SM22, 72 24% of native tissue [p 0.17]). Vimentin expression revealed less excessive fibroblast migration in the TEP tissue (descending aortic model, 93 59% of native tissue [p 0.78]; pulmonary arterial model, 56 36% of native tissue [p 0.31]).

6 1826 IWAI ET AL Ann Thorac Surg TISSUE-ENGINEERED GRAFT 2005;80: Fig 6. Histologic and scanning electron microscopy findings of large tissue engineered patch ( TEP) model 6 months after implantation. Hematoxylin-eosin staining (magnification 100 ) showed that the architecture of the implanted TEP was similar to that of the native pulmonary artery wall (A). The scanning electron microscopy observations (500 ) showed that the luminal surface of the TEP was uniformly covered by a confluent endothelium (B). Elastica van Gieson staining (100 ) revealed the expansion of native tissue into the TEP (C). Factor VIII staining (100 ) showed a number of microvessels in the regenerated tissue (D). Biochemical Examination The 4-hydroxyproline assay demonstrated that 2 months after implantation, the patch had almost the same collagen content as the native vascular tissue (descending aortic model, % of native tissue [p 0.26]; pulmonary arterial model, 92 6% of native tissue [p 0.83]). The elastin content increased to half that of native tissue 2 months after implantation (descending aortic model, 63 15% of native tissue [p 0.64]; pulmonary arterial model, 47 25% of native tissue [p 0.24]). Comment In this study, we demonstrated the feasibility of using a bioengineered material made of collagen-microsponge and a biodegradable polymer scaffold to promote autologous vascular tissue regeneration without precellularization in various situations, especially arterial reconstruction. Among the experimental animals, whether they underwent arterial or venous reconstruction, there was no case of neointimal hyperplasia, thrombus or aneurysm formation, or rupture, even in the animals whose implant was left in place for 6 months. These findings suggested the TEP patch had sufficient durability after implantation to sustain vascular reconstruction. In our previous study, the PGA patch showed good mechanical strength. However, reinforcement of the patch seemed to be necessary in some cases: for example, in pulmonary hypertension and systemic circulation. Except for this caveat, PGA appeared to be a useful biodegradable polymer with a longer degradation time than 6 months, and with the potential to reinforce the wall of the graft [9]. Therefore, we applied a woven PLA mesh to reinforce the outside of the patch to avoid rupture and aneurysm formation in the present study. Our histologic findings showed the architecture of the implanted TEP was similar to the native vascular wall. No calcification was detected 6 months after implantation. The cellular population, which was determined by RT- PCR, demonstrated excellent in situ cellularization of the TEP. The increase in the collagen content would be reflected by quantifying both the bovine collagen of the initial graft and the newly generated porcine collagen, which could not be discriminated in this experiment. However, the increase in the elastin content to half that of native tissue would only be due to newly generated porcine elastin. In our study, we documented that the percentage of cellular and extracellular components in the graft had increased to levels similar to those in native tissue at six months [7]. The present results suggested that the collagenmicrosponge facilitated the in vivo attachment and spreading of autologous cells throughout the patch, and did so without precellularization. Previously, the redifferentiation of dedifferentiated cells was shown in vitro using a biodegradable polymer with collagenmicrosponge [10]. We also showed that the biodegradable scaffold without the collagen-microsponge resulted

7 Ann Thorac Surg IWAI ET AL 2005;80: TISSUE-ENGINEERED GRAFT in immature regeneration and thrombus formation. The cellular affinity for collagen and the interconnected pore structure of microsponge accelerate the in situ regeneration of autologous tissue without precellularization or the addition of growth factors. Regarding the precellularization, conventional tissue engineering techniques have required autologous cell harvesting and ex vivo culture before implantation. To avoid these pretreatments, a bone marrow cell-seeding technique was recently reported by Matsumura and colleagues [5]. The idea of using bone marrow cells has attracted much interest. The results with bone marrow cells inspired us to test whether we could promote the regeneration of vascular tissue from vascular progenitor cells. Our concept of inducing vascular progenitor cell growth is based on the idea of supplying an in situ bioreactor and incubator for circulating progenitor cells. Therefore, our findings on the efficacy of this material without precellularization may not be reproducible if the graft is used in an organ without direct contact with circulating blood. In our previous study, we proved that there was no remarkable difference in the histologic and biochemical findings after implantation of this material into the dog pulmonary artery between groups with and without precellularization, from 2 to 6 months. A significant difference may exist earlier than 2 months after implantation. However, we found endothelialization of the TEP even by 1 month and observed histologic changes associated with regeneration. In the present study, we obtained good in situ regeneration with many favorable histologic characteristics. Although the precise mechanisms behind the in situ regeneration have not been elucidated, we can speculate the following: (1) circulating vascular progenitor cells attach to the TEP and differentiate, as previously demonstrated [11 16]; (2) cellular and extracellular components from the native tissue expand into the TEP; (3) angiogenesis occurs in the newly formed tissue; and (4) fibroblasts that migrated from the adventitia side of the vessel proliferated in the TEP. Nonetheless, even though our findings seem to reveal site-specific regeneration of vascular tissue, further studies are needed for proof. Such experiments will be important in creating ideal conditions for autologous vascular cells to migrate and to proliferate in situ in artificial grafts. Thus our TEP, made of biodegradable polymer and collagen-microsponge, is a candidate for a novel surgical material; it should be useful for reconstructing autologous tissue in cardiovascular surgery, without the use of foreign material, to avoid later complications. The limitations of this study are the relatively small number of animals used and the short durations of the observations. Moreover, for the clinical use of this graft in the repair of congenital heart disease to be viable, its durability and potential for growth should be evaluated over a long-term follow-up period. The use of a xenogenous substance such as bovine collagen may also be a limitation for clinical application, although prosthetic vascular grafts coated with collagen are already widely used in clinical situations. As the next step, we should consider the use of atelocollagen, which carries lower immunologic and infectious risks. Nevertheless, we believe that our strategy of using this material without precellularization may be a useful option for a novel tissue engineering technique, especially for neonatal and pediatric cardiovascular surgery, which cannot wait for cell culture. Further investigation will be directed towards evaluating its potential and safety for clinical applications. In the present study, a bioengineered patch made of biodegradable polymer and collagen-microsponge used without precellularization, led to excellent histologic findings and was durable. This bioengineered patch is a promising material for cardiovascular surgery that could promote the in situ cellularization that leads to the regeneration of autologous vessel tissue, in both the venous and arterial wall. References Langer R, Vacanti JP. Tissue engineering. Science 1993;260: Shinoka T, Shum-Tim D, Ma PX, et al. Creation of viable pulmonary artery autografts through tissue engineering. J Thorac Cardiovasc Surg 1998;115: Hoerstrup SP, Kadner A, Melnitchouk S, et al. Tissue engineering of functional trileaflet heart valves from human marrow stromal cells. Circulation 2002;106:I Matsumura G, Hibino N, Ikada Y, Kurosawa H, Shinoka T. Successful application of tissue engineered vascular autografts: clinical experience. Biomaterials 2003;24: Matsumura G, Tomita SM, Shinoka T, Ikada Y, Kurosawa H. First evidence that bone marrow cells contribute to the construction of tissue-engineered vascular autografts in vivo. Circulation 2003;108: Chen G, Ushida T, Tateishi T. A biodegradable hybrid sponge nested with collagen-microsponges. J Biomed Mater Res 2000;51: Iwai S, Sawa Y, Ichikawa H, et al. Biodegradable polymer with collagen microsponge serves as a new bioengineered cardiovascular prosthesis. J Thorac Cardiovasc Surg 2004;128: Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Res 1996;6: Ozawa T, Mickle DA, Weisel RD, et al. Histologic changes of nonbiodegradable and biodegradable biomaterials used to repair right ventricular heart defects in rats. J Thorac Cardiovasc Surg 2002;124: Chen G, Sato T, Ushida T, Hirochika R, Tateishi T. Redifferentiation of dedifferentiated bovine chondrocytes when cultured in vitro in a PLGA-collagen hybrid mesh. FEBS Lett 542;2003: Asahara T, Masuda H, Takahashi T, et al. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological neovascularization. Circ Res 1999;85: Shi Q, Rafii S, Wu MH, et al. Evidence for circulating bone marrow-derived endothelial cells. Blood 1998;92: Yamashita J, Itoh H, Hirashima M, et al. Flk1-positive cells derived from embryonic stem cells serve as vascular progenitors. Nature 2000;408: Han CL, Campbell GR, Campbell JH. Circulating bone marrow cells can contribute to neointimal formation. J Vasc Res 2001;38: Shimizu K, Sugiyama S, Aikawa M, et al. Host bone marrow cells are a source of donor intimal smooth-muscle-like cells in murine aortic transplant arteriopathy. Nat Med 2001;7: Sata M, Saiura A, Kunisato A, et al. Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis. Nat Med 2002;8:403 9.