- Email: [email protected]
5.19 In Vivo Bioreactors☆
5.19 In Vivo Bioreactors☆ O Bleiziffer, Inselspital Bern University Hospital, Bern, Switzerland U Kneser, BG Trauma Center Ludwigshafen, Ludwigshafen, Germany r 2017 Elsevier Ltd. All rights reserved.
5.19.1 5.19.2 5.19.2.1 5.19.3 5.19.3.1 5.19.3.2 5.19.3.3 5.19.3.3.1 5.19.3.4 5.19.3.5 5.19.3.5.1 References
Introduction Elements of the In Vivo Bioreactor Stem Cells or Progenitor Cells Applications and Tissues Skin Bone Small Intestine Transplant tolerance induction and restoration of metabolic and organ function by cotransplantation of thymus with solid organs Local Immunosuppression in Vascularized Composite Tissue Allotransplantation (VCA) Urinary Tract The arteriovenous loop model: An “in vivo bioreactor” to generate vascularized tissues and organs
Abbreviations apoE Apolipoprotein E AV Arteriovenous bFGF Basic fibroblast growth factor BMP Bone morphogenetic protein CD Cluster of differentiation CSF Colony-stimulating factor CT Computed tomography CTA Computed tomography angiography ELISA Enzyme-linked immunosorbent assay EPC Endothelial progenitor cells FLC Fetal liver cells
5.19.1
343 344 344 345 345 345 346 347 347 347 348 349
IE Islet equivalents IK Islet kidneys MGH Massachusetts General Hospital MHC Major histocompatibility complex MLR Mixed lymphocyte reaction MRA Magnetic resonance angiography SBS Short bowel syndrome SIS Small intestinal mucosa TESI Tissue-engineered small intestine TIK Thymo-islet-kidneys VEGF Vascular endothelial growth factor
Introduction
The purpose of an in vivo bioreactor is to maximize the role of the patient’s own body in the regeneration of tissues and organs. The elements traditionally contributing to tissue engineering applications include cells that may be modified to serve specific purposes or stem cells, signal molecules, that is, growth factors and cytokines that exert specific effects on these cells, bioactive scaffolds that may replace or support form or function of impaired tissues or organ systems, and the extracellular environment that maintains protection and nutrition (Fig. 1). Many different approaches to combine cells, scaffolds, and signal molecules have been employed in tissue engineering and regenerative medicine, emphasizing either the in vitro or the in vivo approach to a varying extent. A bioreactor has thus far been defined as a device or system that supports a biologically active environment and has been extensively used in the past to optimize in vitro culture conditions, particularly in the field of liver1 and bone2 tissue engineering. On the contrary, the concept of an “in vivo bioreactor” is relatively new. To our knowledge, the term was first coined in 2005 by two independent research groups3,4 in a successful attempt to induce formation of tissue-engineered bone in vivo. Even though the animal models and the study design were entirely different from each other, the advantages of an in vivo approach toward tissue engineering were similar and continue to be so: availability and chemotaxis of autologous pluripotent stem or progenitor cells, eliminating the need for ex vivo cell manipulation and transplantation; autologous supply of growth factors as well as other
☆
Change History: October 2016. O. Bleiziffer and U. Kneser updated the text and references section.
This is an update of O. Bleiziffer, R.E. Horch and U. Kneser, 5.512 – In Vivo Bioreactors. In Comprehensive Biomaterials, edited by Paul Ducheyne, Elsevier, Oxford, 2011, pp. 169–175.
Comprehensive Biomaterials II, Volume 5
doi:10.1016/B978-0-12-803581-8.10225-5
343
344
In Vivo Bioreactors
Fig. 1 Principles and mechanisms involved in the in vivo bioreactor.
supportive stimuli; and potential for in situ creation of a vascular network supplying the tissue-engineered construct, including a main vessel axis, enabling a later transfer of the vascularized neo-tissue or organoid to the desired site. Our overview will highlight attempts toward generating “in vivo bioreactors” in different compartments of the body with a broad range of applications which include not only tissue regeneration but also the investigation of tumor and metastasis formation,5 tolerance induction in tissue and organ transplantation,6 and substitution of metabolic function.7,8 As insufficient oxygen supply has been one of the major obstacles in the generation of bioartificial tissues and organs, we will particularly emphasize vascularization strategies and present the application possibilities of an axially vascularized separation chamber that could be employed not only to create transplantable tissues or organoids with their own blood supply, based on a microsurgically created arteriovenous axis (reviewed in Polykandriotis et al.9) but may also be used as a platform for cancer research (reviewed in Hutmacher et al.5).
5.19.2 5.19.2.1
Elements of the In Vivo Bioreactor Stem Cells or Progenitor Cells
Stem cells or progenitor cells are considered among the most promising tools today for regenerative medicine applications, but despite extensive research, control over trafficking, survival, proliferation, and differentiation remains extremely challenging and the underlying mechanisms largely unknown (review in Discher et al.10). The interplay of growth factors, cell–cell contacts, and cell–matrix adhesions taking place in vivo, determines the response of these cells to a certain stimulus determining their action and effect, but is very difficult to reproduce under in vitro conditions. Drawbacks of this ex vivo approach lie in the fact that stem and progenitor cell populations may be very heterogeneous, making it difficult to obtain purity of the cell population, thereby limiting reproducibility. In addition, growth in two-dimensional (2D) culture systems may lead to changes in phenotype and differentiation and proliferation status, altering the cells' potential to exert their proper inherent function. Absence of physiological stimuli and factors such as the extracellular matrix, mechanical forces, and signaling molecules may all contribute to undesired changes in the cell's features and capacity for regeneration. The in vivo bioreactor makes use of the fact that potentially useful cell populations already exist in the body and attraction of these cells to a desired anatomic site may provide new therapeutic options. At the same time, complexity of manipulation and cost associated with in vitro bioreactors can be reduced. Stevens et al. were the first to coin the term “in vivo bioreactor” and demonstrated the potential of de novo bone growth solely based on the creation of an artificial space between bone and
In Vivo Bioreactors
345
periosteum, which was subsequently filled with a calcium alginate gel that cross-linked in situ and provided the scaffold for powerful generation of functional native bone.3 An increasing number of studies have been aimed at the mobilization of endothelial progenitor cells (EPC) to ischemic tissues with chemotactic agents, yielding disappointing results as far as retention of the transplanted cells is concerned. Most trial protocols have attempted to mobilize progenitor cells using injections of granulocyte colony-stimulating factor (G-CSF). G-CSF, however, induces irreversible cleavage of cell–surface adhesion molecules a4-integrin and CXCR4 by proteases, thereby reducing retention of progenitor cells in the ischemic target areas. Indeed, use of alternative mobilization agents appear to be promising in this regard.11 In vivo vascularization of a given space containing a matrix or a scaffold is gaining increasing attention to prebuild a vascularized structure and deliver either pluripotent or specialized cells to it as the blood vessel tree continues to grow.12 This approach avoids the typical inevitable problem of hypoxic periods encountered in the primarily in vitro approach where cells are combined with a matrix in a bioreactor followed by transplantation into an initially avascular recipient bed where the blood supply has yet to grow into the construct. As mentioned above, EPC have been under intense investigation in the past due to their angiogenic (either directly or via secondary cytokines such as VEGF13) and vasculogenic (via direct integration and subsequent de novo blood vessel formation).14 In an attempt to increase homing of transplanted human EPC to a growing prevascularized tissue-engineered organoid in a separation chamber in a nude rat model, Simcock et al. administered the chemokine CXC chemokine ligand 12 (CXCL12) and achieved enhanced homing of xenogenic EPC to an angiogenic site developing around an arteriovenous loop inserted in a tissue engineering chamber.12 Similar approaches will undoubtedly be developed and optimized to attract vascular progenitor cells and other precursor or stem cells to the in vivo tissue engineering site.
5.19.3 5.19.3.1
Applications and Tissues Skin
The skin is the largest organ in the body and primarily functions as a barrier between the organism and the environment. Major skin loss as a result of trauma and burns has been a major concern to physicians and lead to the development of a multitude of strategies to promote skin wound healing and restore surface integrity. Among them, skin tissue engineering has gained particular attention in recent years and has led to a wide variety of commercially available skin substitutes of different compositions. Significant drawbacks, however, include high costs or the need for extensive ex vivo manipulation. The Eriksson group made the first attempt to use the skin's accessibility and inherent regenerative capacities for healing by creating a standardized and isolated in vivo liquid wound environment using a transparent flexible chamber for sealing. This system allowed the delivery of analgesics, antibiotics, growth factors via direct or cell-based gene transfer, growth media, and cells into the chamber, creating an in vivo incubator-like environment. The group demonstrated healing of clean wounds as fast as or faster than any other method, with less scarring.15 The beneficial effects were also demonstrated in a clinical study.16 Easy access to the skin has induced interest beyond wound healing and skin regeneration. Keratinocytes secrete a wide variety of proteins that have a role in skin physiology, including proteins and proteinases for matrix remodeling, growth factors, and cytokines.7 In addition, it has been demonstrated that skin cells, namely keratinocytes, secrete proteins that can reach the bloodstream and gain systemic impact. This effect was initially demonstrated by Taichman and coworkers who grafted human keratinocytes onto athymic mice and rats. Human apolipoprotein E (apoE) was thereafter detected in the systemic circulation of the animals as long as the graft remained on the animals.17 Human apoE was reduced to undetectable levels within 24 h after graft removal. The potential of using genetically altered skin cells such as keratinocytes and fibroblasts, given that a skin-derived protein attained systemic distribution, has been exploited in numerous studies since then.7 Transgene products were not only detected in the serum, but also exhibited physiological effects. A study by Meng et al.,18 in which an interleukin-10 expressing vector was injected into the dorsal skin of rats, demonstrated that the transgene was detected in the bloodstream and led to an inhibition of contact hypersensitivity at distant areas of the skin. A serious drawback of methods of transient gene transfer to exploit the skin's accessibility as a delivery route for systemically effective agents has been the fact that efficiency is low and the effect only transient, likely due to promoter silencing of the transgene and immunological responses.7 Therefore, transgenic mouse models have been used for experimental evaluation of the skin as a bioreactor, as gene expression was shown to be higher and more ubiquitous, and simultaneously eliminating a host immune response. Targeting of the therapeutic gene to specific layers of the dermis was achieved by stratum-specific promoters such as keratin 5 and keratin 14. Gene delivery exploiting the skin’s potential as an in vivo bioreactor may provide therapeutic options for metabolic diseases such as growth hormone for dwarfism, coagulation factors for coagulopathies, and insulin for diabetes mellitus.19
5.19.3.2
Bone
Treatment of large bone defects still requires autologous transfer of vital bone. Due to the discrepancy between the need on the one hand and limited supply as well as significant donor site morbidity on the other hand, generation of bone by the principles of tissue engineering has become a subject of extensive research efforts.20,21 Most experimental setups employ traditional principles
346
In Vivo Bioreactors
of tissue engineering where harvested cells are cultured in vitro, combined with porous scaffolds, and stimulated with exogenous morphogens and mitogens in vitro and/or in vivo. In contrast to this, the Langer group designed a rabbit model where an artificial space between the tibia and the adjacent periosteum was created, which was referred to as an “in vivo bioreactor.”3 Owing to the exposition of this mesenchymal layer which is rich in pluripotent cells, the creation of significant volumes of bone was possible in a predictable manner, without the necessity for delivery of growth factors or cell transplantation. Biological functionality of the newly created bone was demonstrated by successful transplantation to contralateral tibia defects, resulting in complete integration without significant donor site morbidity. This study also showed that inhibition of angiogenesis in this setting and the creation of a more hypoxic environment promoted cartilage formation. A similar model has been described by Miller and coworkers who used a plastic chamber for periosteum-guided bone formation from the rib in a sheep model.22–24 The lack of a defined vascular axis, however, limits the size and potential to transplant ectopically grown neo-bone to the desired location. Therefore, the ideal bone graft would not only include osteoconductive scaffolds, osteoinductive proteins, and osteogenic cells, but also a vascular network based on a main vascular access constituted by an artery and a vein. Prefabrication25 is defined as “the process of neovascularization by the implantation of a vascular pedicle into tissues to be transferred later to a distant site either on the pedicle or as a free graft by microvascular anastomosis.”26 Many attempts focused on using the inferior epigastric pedicle as a vascular axis to supply the in vivo bone bioreactor created around a wide variety of scaffolds consisting of hydroxyapatite27 and coralline,4 among others, in a separation chamber. A certain degree of bone formation could be demonstrated after addition of BMP-2 and establishment of a vascular network. This in vivo bone bioreactor was also shown to be useful as a model of tumor–bone interaction.28 A clinical application of in situ-generated vascularized engineered bone was reported by Terheyden’s group to repair an extended mandibular defect. They created a custom-made construct based on a titanium mesh cage filled with bone mineral blocks and infiltrated with recombinant human bone morphogenetic protein-7 (BMP-7) and the patient’s bone marrow. Thus prepared, the transplant was implanted into the latissimus dorsi muscle and 7 weeks later, transplanted as a free bone–muscle flap to repair the mandibular defect. Follow-up by skeletal scintigraphy and CT showed bone remodeling and mineralization inside the mandibular transplant both before and after in vivo transplantation, thus providing radiological evidence of new bone formation.29 In recent years, 3-D bioprinting has emerged as a new and promising tool towards generation of a variety of tissues, including bone. As reviewed by Atala, it enables precise positioning of cells, biochemical materials under spatial control and layer-by-layer positioning. Principal approaches to 3D bioprinting are referred to as biomimicry, autonomous self-assembly and mini-tissue building blocks.30 The same author and his group recently introduced an integrated tissue–organ printer (IOP) which is expected to enable generation and engineering of stable, human-scale tissue constructs of a desired shape. As proof of principle, they generated a human-sized mandible fragment based on a computer-aided design model (CAD) using data from a CT scan of a human mandible defect. Human amnionic-fluid derived stem cells were employed in combination with a hydrogel consisting of a combination of PCL and tricalcium phosphate (TCP), and Pluronic F127.31 The CAD model was translated into a program that controls motion of printer nozzles which are responsible for 3D delivery and dispension of cells into preformed microchannels. This fashion of delivery circumvents the diffusion limit of 100–200 mm for cell survival in engineered tissues. After induction of osteogenic differentiation, 3 D mandible-shaped bone formation was successfully demonstrated.31 The principle of in vitro 3D printing may efficiently be combined with the in vivo bioreactor principle in the future.
5.19.3.3
Small Intestine
The management of patients with short bowel syndrome (SBS) has been unsatisfactory with conventional treatment strategies so far. Various studies from the Vacanti laboratory demonstrated improved recovery after massive small bowel resection (SBR) using tissue-engineered intestine.32 The tissue-engineered small intestine (TESI) was constructed applying “in vivo bioreactor” principles. First, segments of intestine harvested from neonatal rats were minced and full-thickness plugs of intestine created. These so-called organoid units were then seeded onto a tubular biodegradable scaffold polymer and then wrapped into the omentum of adult rats which served as the in vivo bioreactor. These organoid units, arranged on the biodegradable scaffold, became vascularized by the omentum to form a cyst, which histologically resembled the small bowel.33 Several biochemical features of digestive enterocyte function and polarity have been demonstrated in these cysts in later studies.34 As a next step, the mature, vascularized TESI cyst was placed in continuity with the gastrointestinal tract at the time of an 85% enterectomy in rats. TESI animals returned to preoperative weight more rapidly when compared with the animals undergoing SBR alone. On histologic analysis, the TESI cysts appeared to resemble normal adjacent intestine. Furthermore, the neointestine had preserved expression of green fluorescent protein that had been introduced into the organoid units prior to implantation. These marker studies strongly suggested that the neointestine was derived primarily from the neonatal-derived intestine, as opposed to ingrowth of adjacent recipient intestine. Additional studies showed lymphangiogenesis as detected by expression of the lymphatic endothelial marker vascular endothelial growth factor receptor (VEGFR)-3 using immunohistochemistry, and the lymphangiogenic growth factor VEGF-C was quantified by ELISA. Moreover, tubular structures, which resembled lymphatics architecturally, developed, and were distinct from CD34-positive blood vessels, and lacked luminal erythrocytes. This was the first demonstration of lymphatic vessels in an engineered tissue.35 Tissue-engineered intestine also had the capacity to develop a mucosal immune system with an immunocyte population similar to that of native small intestine depending on both exposure to luminal stimuli and the duration of this exposure.36 These results demonstrate the possibility of ectopic formation of a
In Vivo Bioreactors
347
biologically functional bioartificial organoid with access to blood and lymphatic circulation, which could even be transferred to an orthotopic site with subsequent proof of maintained biological function.
5.19.3.3.1
Transplant tolerance induction and restoration of metabolic and organ function by cotransplantation of thymus with solid organs
Organ transplantation to restore impaired or lost function is hampered by the short- and long-term complications of immunosuppression and the current worldwide shortage of donor organs.37 The induction of donor-specific tolerance in allogeneic and xenogeneic transplantation offers a promising solution to these problems. The thymus has been identified as critical for systemic central tolerance to self-antigens in which potentially autoreactive T cells are deleted or anergized by exposure to the appropriate self-antigens presented either by bone marrow-derived cells or thymic stromal cells.33,34 Based on these findings and their MGH miniature swine model, David Sachs and coworkers have developed composite allografts consisting of vascularized thymic tissue and the kidney as a solid organ to induce tolerance across major histocompatibility complex (MHC)-mismatched barriers.38 A new transplantable organ (thymokidney) was generated by autologous thymic grafts under the renal capsule. The thymic grafts exhibited normal macroscopic and microscopic structure, and normal thymocyte composition. To evaluate the thymic function of the graft, a composite thymokidney was transplanted into a recipient who had previously been thymectomized, had few circulating CD4-single positive cells, and had lost MLR reactivity. The number of CD4 þ /CD45RA þ cells in this animal increased steadily indicating functionality of the thymus within the composite thymokidney. In addition, MLR assays demonstrated that the recipient recovered immunocompetence.39 Additional work from the same laboratory demonstrated that transplantation of allogeneic thymic tissue as part of a composite vascularized graft was far more successful in terms of both engraftment and long-term survival than transplantation of thymic tissue or cells alone.40 This concept of the vascularized composite neo-organ was subsequently extended to transplantation of allogeneic islets, comparing survival of islet cell suspensions to that of vascularized composite islet-kidneys (IK), prepared by injection of autologous islets underneath the renal capsule 2–3 months prior to allogeneic transplantation of the composite organ. Composite IK were transplanted successfully across minor and full MHC mismatch barriers, using treatment regimens previously demonstrated to induce long-term tolerance of kidney allografts across these barriers. IK allografts containing Z5000 islet equivalents (IE) per kilogram recipient body weight were found capable of reversing surgically induced diabetes, while injection of comparable numbers of purified islets via the portal vein or under the renal capsule did not. An additional study was directed toward preparation of autologous “thymo-islet-kidneys” (TIK), for potential use as xenografts, in which the thymic component was intended to induce tolerance and the islets to reverse diabetic hyperglycemia. To evaluate the function of the transplanted islets, three animals bearing TIK and IK underwent total pancreatectomy 3 months following islet transplantation. All three animals maintained normoglycemia thereafter. In two of these animals, the IKs were removed 2 months after the pancreatectomy, and in both cases normoglycemia was maintained thereafter by the TIK. The combination of replacement of a lost metabolic function and vascular competence of the grafts, as well as the potential for immune tolerance induction, makes this model a particularly successful example of an unusual application of an in vivo bioreactor. Use of both types of composite organ transplants (IK and TIK) may eventually be applicable to the treatment of type I diabetic patients suffering from end-stage diabetic nephropathy.40
5.19.3.4
Local Immunosuppression in Vascularized Composite Tissue Allotransplantation (VCA)
In the previous chapter, tolerance induction with thymus contransplantation was demonstrated to have potential to improve transplant survival. Up to now, systemic immunosuppression, however, constitutes the standard treatment after allogenic solid organ transplantation, even though patients suffer from serious side effects. To circumvent these drawbacks, it would appear desirable to deliver immunosuppression locally. Recently, this strategy has been tested in a rat VCA hindlimb transplantation model. The immunosuppressive drug tacrolimus was loaded into a self-assembled hydrogel. During the post-transplantation inflammation process proteolytic enzymes are overexpressed, leading to degradation of the hydrogel and thereby inducing release of the drug. In their study, Gajanayake et al. reported significantly prolonged graft survival in a Brown Norway–to–Lewis rat hindlimb transplantation model after one-time local injection of the tacrolimus-laden hydrogel.41 This therapeutic principle requires the biological in vivo bioreactor-like environment within the allograft, where a stimulusresponsive sustained release of the immunosuppressive drug can take place.42 Since VCA, eg, of the upper extremity, is not a lifesaving procedure, new therapeutic strategies to improve long term survival of the graft whilst providing more safety and reduction of side effects are particularly crucial compared to solid organ transplantation. The principle of VCA has also been combined with the evaluation of surgical angiogenesis in an experimental rabbit model of knee joint transplantation.43
5.19.3.5
Urinary Tract
Congenital and acquired ureteric defects cause significant morbidity and have been notoriously difficult to regenerate by tissue engineering approaches to create a neoureter in the past. Such attempts to generate a neoureter in animal models failed because of major inflammatory response. Avoidance of such inflammation requires a well-differentiated urothelium. Baumert et al. investigated whether omental maturation of a small intestinal submucosa (SIS) matrix seeded with previously cultured urothelial and
348
In Vivo Bioreactors
smooth muscle cells that had been derived from bladder biopsies in a pig model could achieve terminal differentiation of the urothelium to allow construction of a stricture-free neoureter. After 2 weeks of cell growth, the in vitro SIS-seeded construct was shaped around a silicone drain and wrapped by the omentum to obtain neoureters. These neoureters were left in the omentum without any contact with urine, and then harvested 3 weeks later for histologic and immunohistochemical studies. Before implantation, the in vitro constructs were composed of a mono- or bilayer of undifferentiated urothelium overlying a monolayer of smooth muscle cells. After 3 weeks of omental maturation, these constructs were vascularized and comprised a terminally differentiated multilayered urothelium with umbrella cells over connective tissue and smooth muscle cells, with no evidence of fibrosis or inflammation. This model of in vivo maturation using the omentum as an in vivo bioreactor enabled for the first time the generation of a mature neoureter composed of a well-differentiated multilayered urothelium.44 When looking at other parts of the genitourinary system, bladder regeneration is another challenging goal. One of the main problems is to limit the development of ischemic fibrosis during tissue maturation. Similar to their ureter tissue engineering, Baumert et al. described a model using the omentum as an in vivo bioreactor for a previously seeded scaffold. A highly vascularized tissue-engineered construct was generated that contracted in response to acetylcholine stimulation. The wall thickness was 4 mm, on average. Histologic and immunostaining analysis of the construct confirmed the presence of a multilayer urothelium on the luminal aspect and deeper fascicles of organized tissue composed of differentiated smooth muscle cells and mature fibroblasts without evidence of inflammation or necrosis. Large- and small-diameter vessels were clearly identified histologically in the tissue obtained. The omentum as an in vivo incubator permitted in vivo maturation of seeded scaffolds with the development of a dense vascularization that is anticipated to prevent fibrosis and loss of contractility. This in vivo maturation into the omentum could be the first step before in situ implantation of the construct.45
5.19.3.5.1
The arteriovenous loop model: An “in vivo bioreactor” to generate vascularized tissues and organs
The development of a functional and adequate blood vessel network is a prerequisite for successful application of engineered tissues and organs. When in vitro-engineered constructs containing cells are transferred in vivo, they lack a defined axial blood supply consisting of a main arterial inflow and a venous outflow, respectively, and therefore rely on blood vessel ingrowth from their surroundings, that is, interstitial fluid diffusion and blood perfusion.20,21 This limitation in blood supply constitutes the main drawback when transferring in vitro tissue engineering approaches to the in vivo environment. Cell support via diffusion occurs only up to 200 mm into the matrix,46 resulting in insufficient nutrition and subsequent loss of cells located in the center of such in vitroengineered constructs. For these reasons, induction of vascularization is essential in virtually all successful tissue engineering applications, and a defined vascular axis is desirable whenever the bioengineered tissue or organ is yet to be transplanted to a different site within the host. In order to enable standardized transfer techniques of neo-tissues and organs to a different site, reconstructive surgeons have therefore aimed to generate axially vascularized tissues and organoids that are supplied by a defined arteriovenous axis and could hence be transferred to the desired site using microsurgical techniques of vascular anastomosis. As opposed to other techniques, these tissues are immediately connected to the systemic blood circulation upon implantation as free flaps are. The arteriovenous loop model based on the femoral vessels, including a venous graft from the contralateral side in a rat separation chamber, fulfills our criteria to constitute an in vivo bioreactor. The studies presented in this chapter are based on the initial description by Erol and Spira47 who successfully generated a prefabricated skin flap by creating an arteriovenous vessel loop using either arterial or venous grafts in a rat model. They observed the outgrowth of a dense 3D neovessel tree originating from the entire AV loop. This model was later further developed by isolation of the AV loop into isolation chambers and induction of vascularization in polymer matrices.48–50 Based on these works, a custom-made processed bovine cancellous bone matrix was successfully vascularized by means of the AV loop.20,21 A detailed analysis of the vascularization was established using invasive techniques such as corrosion casting as well as noninvasive methods such as micro-CT scanning.51 Using histology-based morphometric analysis, vascularization was meticulously quantified and it was demonstrated that more than 90% of the scaffold vascularized 8 weeks following implantation. Prevascularization of these matrices significantly improved survival of osteoblasts engrafted into the matrix, emphasizing the importance of an intrinsic blood supply for cell vitality within the construct.51 The experimental model has also been successfully modified to allow for combined intrinsic with extrinsic vascularization52 which was shown to significantly accelerate axial vascularization of a matrix within the isolation chamber. The AV loop model to vascularize tissue-engineered constructs is not limited to hard matrices and the attempt to create bone, but may be extended by using other matrices and cell populations. A fibrin gel approved for clinical use can be employed to generate a fibrin matrix, which was successfully vascularized and may be used as a platform to create soft tissues or organoids. The vascularization within this fibrin matrix was shown to respond to proangiogenic stimuli by accelerating the vascular outgrowth when the recombinant growth factors VEGF and bFGF were suspended in the matrix.53,54 Transplantation of myoblasts into a previously vascularized fibrin matrix containing the AV loop enhanced cell survival and maintenance of the myogenic phenotype compared to fibrin matrices without AV loops, even though no formation of skeletal muscle-like tissue could be observed, possibly due to the lack of potent myogenic stimuli.55 The vascularized matrix embedded in the AV loop may also be employed to create organoids or replace organ functions. As a new approach to hepatic tissue engineering, fetal liver cells (FLC) were suspended in the fibrin matrix, resulting in neo-tissue formation containing viable fetal donor cells in the highly vascularized fibrin matrix after 14 days.1 Previous work demonstrated that continuous hepatotrophic stimulation is necessary in tissue-engineered liver support systems, and it has been demonstrated that cocultivation with pancreatic islets may have stimulatory effects on hepatocytes and that they are mediated by soluble factors, and are dependent on insulin and glucagon.56,57
In Vivo Bioreactors
349
Fig. 2 Micro-CT scan from a rat AV-loop following Microfil – perfusion and explantation, demonstrating vascular sprouting originating from the AV-loop.
Further investigations toward 3D transplantable hepatocyte-islet devices for continuous in vivo stimulation were conducted to provide hepatotrophic stimulation. Hepatocyte and pancreatic islets were coseeded into polyvinyl alcohol matrices and supported hepatotrophic stimulation of bioartificial liver equivalents equally well, compared to a protocaval shunt operation. The AV loop model may therefore well serve as a platform for the development of highly vascularized in vivo tissue engineering-based liver support systems using these cotransplantation strategies. Successful engineering of cardiac muscles has been achieved by matrigel-suspended neonatal rat cardiomyocytes implanted inside the AV loop. The presence of gap junctions was shown between the cardiomyocytes at 4 and 10 weeks, and echocardiograms showed spontaneous contraction of the tissue in vivo. In vitro, positive chronotropy to norepinephrine and positive inotropy in response to calcium demonstrated biological function of the cardiac neo-organ.58 The AV loop separation chamber potential is not limited by the availability of local factors in its in situ tissue and organ engineering capacity. The regenerative potential of the established chamber construct may be even better exploited by chemotactic recruitment of supportive cells or signaling molecules. It was recently demonstrated that EPC that had been systemically injected before can be recruited from the circulation by CXCL12/stromal cell-derived factor 1,12 potentially enabling further modulation of the vascularization process. The isolation chamber setting with its potential to be monitored and manipulated may also be used as an investigation platform to investigate the phenomena and regulation of angiogenesis and antiangiogenesis in cancer research.5 Limitations to dimensions in small animal models have always been an argument against successful transfer of tissue engineering applications to clinical relevance. Consequently, a large animal model (sheep) was established where de novo formation of axially vascularized tissue was demonstrated for the first time ever by sequential computed tomography angiography (CTA) and magnetic resonance angiography (MRA) in vivo, as well as by postexplantational micro-computed tomography (Fig. 2) and histology.59 This constitutes further progress in approaching the first application of tissue engineering vascularized grafts with clinically relevant dimensions.59–61 In conclusion, the arteriovenous loop model has been investigated mainly as a powerful tool to create bioartificial tissues and organoids, but may also be very useful as a platform to investigate angiogenetic and antiangiogenetic phenomena in general, with particular emphasis on cancer therapy.5
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
Fiegel, H. C.; Pryymachuk, G.; Rath, S.; et al. J. Cell. Mol. Med. 2010, 14 (1–2), 267–274. Fröhlich, M.; Grayson, W. L.; Wan, L. Q.; Marolt, D.; Drobnic, M.; Vunjak-Novakovic, G. Curr. Stem Cell Res. Ther. 2008, 3 (4), 254–264. Stevens, M. M.; Marini, R. P.; Schaefer, D.; Aronson, J.; Langer, R.; Shastri, V. P. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (32), 11450–11455. Holt, G. E.; Halpern, J. L.; Dovan, T. T.; Hamming, D.; Schwartz, H. S. J. Orthop. Res. 2005, 23 (4), 916–923. Hutmacher, D. W.; Horch, R. E.; Loessner, D.; et al. J. Cell. Mol. Med. 2009, 13 (8A), 1417–1427. LaMattina, J. C.; Kumagai, N.; Barth, R. N.; et al. Transplantation 2002, 73 (5), 826–831. Cao, T.; Wang, X. J.; Roop, D. R. Hum. Gene Ther. 2000, 11 (16), 2297–2300. Hussey, A. J.; Winardi, M.; Han, X.; et al. Tissue Eng. A 2009, 15 (12), 3823–3833. Polykandriotis, E.; Arkudas, A.; Euler, S.; Beier, J. P.; Horch, R. E.; Kneser, U. Handchir. Mikrochir. Plast. Chir. 2006, 38 (4), 217–223. (Article in German). Discher, D. E.; Mooney, D. J.; Zandstra, P. W. Science 2009, 324 (5935), 1673–1677. Cheng, M.; Qin, G. Prog. Mol. Biol. Transl. Sci. 2012, 111, 243–264. Simcock, J. W.; Penington, A. J.; Morrison, W. A.; Thompson, E. W.; Mitchell, G. M. Tissue Eng. A 2009, 15 (3), 655–664.
350
13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61.
In Vivo Bioreactors
Asahara, T.; Takahashi, T.; Masuda, H.; et al. EMBO J. 1999, 18 (14), 3964–3972. Tepper, O. M.; Capla, J. M.; Galiano, R. D.; et al. Blood 2005, 105 (3), 1068–1077. Eriksson, E.; Vranckx, J. Am. J. Surg. 2004, 188 (1A Suppl.), S36–S41. Vranckx, J. J.; Slama, J.; Preuss, S.; et al. Plast. Reconstr. Surg. 2002, 110 (7), 1680–1687. Fenjves, E. S.; Gordon, D. A.; Pershing, L. K.; Williams, D. L.; Taichman, L. B. Proc. Natl. Acad. Sci. U.S.A. 1989, 86 (22), 8803–8807. Meng, X.; Sawamura, D.; Tamai, K.; Hanada, K.; Ishida, H.; Hashimoto, I. J. Clin. Invest. 1998, 101 (6), 1462–1467. Taichman, L. B. Proc. Assoc. Am. Phys. 1999, 111 (3), 206–210. Kneser, U.; Polykandriotis, E.; Ohnolz, J.; et al. Tissue Eng. 2006, 12 (7), 1721–1731. Kneser, U.; Schaefer, D. J.; Polykandriotis, E.; Horch, R. E. J. Cell. Mol. Med. 2006, 10 (1), 7–19. Cheng, M. H.; Brey, E. M.; Allori, A. C.; et al. Plast. Reconstr. Surg. 2009, 124 (3), 787–795. Miller, M. J.; Goldberg, D. P.; Yasko, A. W.; et al. Tissue Eng. 1996, 2 (1), 51–59. Tatara, A. M.; Kretlow, J. D.; Spicer, P. P.; et al. Tissue Eng. Part A 2015, 21 (9–10), 1520–1528. Guo, L.; Pribaz, J. J. Plast. Reconstr. Surg. 2009 Dec, 124 (6 Suppl.), e340–e350. Gill, D. R.; Ireland, D. C.; Hurley, J. V.; Morrison, W. A. J. Hand Surg. Am. 1998, 23 (2), 312–321. Mizumoto, S.; Inada, Y.; Weiland, A. J. J. Reconstr. Microsurg. 1993, 9 (6), 441–449. Halpern, J.; Lynch, C. C.; Fleming, J.; et al. Clin. Exp. Metastasis 2006, 23 (7–8), 345–356. Warnke, P. H.; Springer, I. N.; Wiltfang, J.; et al. Lancet 2004, 364 (9436), 766–770. Murphy, S. V.; Atala, A. Nat. Biotechnol. 2014, 32 (8), 773–785. Kang, H. W.; Lee, S. J.; Ko, I. K.; et al. Nat. Biotechnol. 2016, 34 (3), 312–319. Grikscheit, T. C.; Siddique, A.; Ochoa, E. R.; et al. Ann. Surg. 2004, 240 (5), 748–754. Duxbury, M. S.; Grikscheit, T. C.; Gardner-Thorpe, J.; et al. Transplantation 2004, 77 (8), 1162–1166. Perez, A.; Grikscheit, T. C.; Blumberg, R. S.; Ashley, S. W.; Vacanti, J. P.; Whang, E. E. Transplantation 2002, 74 (5), 619–623. Greenstein, J. L.; Sachs., D. H. Nat. Biotechnol. 1997, 15 (3), 235–238. Kappler, J. W.; Roehm, N.; Marrack, P. Cell 1987, 49 (2), 273–280. Sprent, J.; Paul, W. E. Fundamental Immunology, Vol. 3. Raven: New York, 1993; pp. 74–109. Yamada, K.; Shimizu, A.; Utsugi, R.; et al. J. Immunol. 2000, 164 (6), 3079–3086. Yamada, K.; Shimizu, A.; Ierino, F. L.; et al. Transplantation 1999, 68 (11), 1684–1692. Vallabhajosyula, P.; Griesemer, A.; Yamada, K.; Sachs, D. H. Cell Biochem. Biophys. 2007, 48 (2–3), 201–207. Gajanayake, T.; Olariu, R.; Leclère, F. M.; et al. Sci. Transl. Med. 2014, 13 (6), 249. Vemula, P. K.; Boilard, E.; Syed, A.; et al. J. Biomed. Mater. Res. A 2011, 97, 103–110. Kremer, T.; Giusti, G.; Friedrich, P. F.; et al. Microsurgery 2012, 32 (2), 118–127. Baumert, H.; Mansouri, D.; Fromont, G.; et al. Eur. Urol. 2007, 52 (5), 1492–1498. Baumert, H.; Simon, P.; Hekmati, M.; et al. Eur. Urol. 2007, 52 (3), 884–890. Folkman, J.; Hochberg, M. J. Exp. Med. 1973, 138 (4), 745–753. Erol, O. O.; Spira, M. Surg. Forum 1979, 30, 530–531. Mian, R.; Morrison, W. A.; Hurley, J. V.; et al. Tissue Eng. 2000, 6 (6), 595–603. Hofer, S. O.; Knight, K. M.; Cooper-White, J. J.; et al. Plast. Reconstr. Surg. 2003, 111 (3), 1186–1192; (discussion 1193–1194). Cassell, O. C.; Morrison, W. A.; Messina, A.; et al. Ann. N. Y. Acad. Sci. 2001, 944, 429–442. Weis, C.; Covi, J. M.; Hilgert, J. G.; et al. J. Microsc. 2015, 259 (3), 185–196. doi:10.1111/jmi.12252. Arkudas, A.; Pryymachuk, G.; Beier, J. P.; et al. Plast. Reconstr. Surg. 2012, 129 (1), 55e–65e. Arkudas, A.; Beier, J. P.; Heidner, K.; et al. Tissue Eng. 2007, 13 (7), 1549–1560. Arkudas, A.; Pryymachuk, G.; Hoereth, T.; et al. Tissue Eng. A 2009, 15 (9), 2501–2511. Bach, A. D.; Arkudas, A.; Tjiawi, J.; et al. J. Cell. Mol. Med. 2006, 10 (3), 716–726. Kaufmann, P. M.; Fiegel, H. C.; Kneser, U.; Pollok, J. M.; Kluth, D.; Rogiers, X. Tissue Eng. 1999, 5 (6), 583–596. Kaufmann, P. M.; Kneser, U.; Fiegel, H. C.; et al. Transplantation 1999, 68 (2), 272–279. Morritt, A. N.; Bortolotto, S. K.; Dilley, R. J.; et al. Circulation 2007, 115 (3), 353–360. Beier, J. P.; Horch, R. E.; Arkudas, A.; et al. Microsurgery 2009, 29 (1), 42–51. Weigand, A.; Beier, J. P.; Hess, A.; et al. Tissue Eng. Part A 2015, 21 (9–10), 1680–1694. Boos, A. M.; Loew, J. S.; Weigand, A.; et al. J. Tissue Eng. Regen. Med. 2013, 7 (8), 654–664.
5.19.1 5.19.2 5.19.2.1 5.19.3 5.19.3.1 5.19.3.2 5.19.3.3 5.19.3.3.1 5.19.3.4 5.19.3.5 5.19.3.5.1 References
Introduction Elements of the In Vivo Bioreactor Stem Cells or Progenitor Cells Applications and Tissues Skin Bone Small Intestine Transplant tolerance induction and restoration of metabolic and organ function by cotransplantation of thymus with solid organs Local Immunosuppression in Vascularized Composite Tissue Allotransplantation (VCA) Urinary Tract The arteriovenous loop model: An “in vivo bioreactor” to generate vascularized tissues and organs
Abbreviations apoE Apolipoprotein E AV Arteriovenous bFGF Basic fibroblast growth factor BMP Bone morphogenetic protein CD Cluster of differentiation CSF Colony-stimulating factor CT Computed tomography CTA Computed tomography angiography ELISA Enzyme-linked immunosorbent assay EPC Endothelial progenitor cells FLC Fetal liver cells
5.19.1
343 344 344 345 345 345 346 347 347 347 348 349
IE Islet equivalents IK Islet kidneys MGH Massachusetts General Hospital MHC Major histocompatibility complex MLR Mixed lymphocyte reaction MRA Magnetic resonance angiography SBS Short bowel syndrome SIS Small intestinal mucosa TESI Tissue-engineered small intestine TIK Thymo-islet-kidneys VEGF Vascular endothelial growth factor
Introduction
The purpose of an in vivo bioreactor is to maximize the role of the patient’s own body in the regeneration of tissues and organs. The elements traditionally contributing to tissue engineering applications include cells that may be modified to serve specific purposes or stem cells, signal molecules, that is, growth factors and cytokines that exert specific effects on these cells, bioactive scaffolds that may replace or support form or function of impaired tissues or organ systems, and the extracellular environment that maintains protection and nutrition (Fig. 1). Many different approaches to combine cells, scaffolds, and signal molecules have been employed in tissue engineering and regenerative medicine, emphasizing either the in vitro or the in vivo approach to a varying extent. A bioreactor has thus far been defined as a device or system that supports a biologically active environment and has been extensively used in the past to optimize in vitro culture conditions, particularly in the field of liver1 and bone2 tissue engineering. On the contrary, the concept of an “in vivo bioreactor” is relatively new. To our knowledge, the term was first coined in 2005 by two independent research groups3,4 in a successful attempt to induce formation of tissue-engineered bone in vivo. Even though the animal models and the study design were entirely different from each other, the advantages of an in vivo approach toward tissue engineering were similar and continue to be so: availability and chemotaxis of autologous pluripotent stem or progenitor cells, eliminating the need for ex vivo cell manipulation and transplantation; autologous supply of growth factors as well as other
☆
Change History: October 2016. O. Bleiziffer and U. Kneser updated the text and references section.
This is an update of O. Bleiziffer, R.E. Horch and U. Kneser, 5.512 – In Vivo Bioreactors. In Comprehensive Biomaterials, edited by Paul Ducheyne, Elsevier, Oxford, 2011, pp. 169–175.
Comprehensive Biomaterials II, Volume 5
doi:10.1016/B978-0-12-803581-8.10225-5
343
344
In Vivo Bioreactors
Fig. 1 Principles and mechanisms involved in the in vivo bioreactor.
supportive stimuli; and potential for in situ creation of a vascular network supplying the tissue-engineered construct, including a main vessel axis, enabling a later transfer of the vascularized neo-tissue or organoid to the desired site. Our overview will highlight attempts toward generating “in vivo bioreactors” in different compartments of the body with a broad range of applications which include not only tissue regeneration but also the investigation of tumor and metastasis formation,5 tolerance induction in tissue and organ transplantation,6 and substitution of metabolic function.7,8 As insufficient oxygen supply has been one of the major obstacles in the generation of bioartificial tissues and organs, we will particularly emphasize vascularization strategies and present the application possibilities of an axially vascularized separation chamber that could be employed not only to create transplantable tissues or organoids with their own blood supply, based on a microsurgically created arteriovenous axis (reviewed in Polykandriotis et al.9) but may also be used as a platform for cancer research (reviewed in Hutmacher et al.5).
5.19.2 5.19.2.1
Elements of the In Vivo Bioreactor Stem Cells or Progenitor Cells
Stem cells or progenitor cells are considered among the most promising tools today for regenerative medicine applications, but despite extensive research, control over trafficking, survival, proliferation, and differentiation remains extremely challenging and the underlying mechanisms largely unknown (review in Discher et al.10). The interplay of growth factors, cell–cell contacts, and cell–matrix adhesions taking place in vivo, determines the response of these cells to a certain stimulus determining their action and effect, but is very difficult to reproduce under in vitro conditions. Drawbacks of this ex vivo approach lie in the fact that stem and progenitor cell populations may be very heterogeneous, making it difficult to obtain purity of the cell population, thereby limiting reproducibility. In addition, growth in two-dimensional (2D) culture systems may lead to changes in phenotype and differentiation and proliferation status, altering the cells' potential to exert their proper inherent function. Absence of physiological stimuli and factors such as the extracellular matrix, mechanical forces, and signaling molecules may all contribute to undesired changes in the cell's features and capacity for regeneration. The in vivo bioreactor makes use of the fact that potentially useful cell populations already exist in the body and attraction of these cells to a desired anatomic site may provide new therapeutic options. At the same time, complexity of manipulation and cost associated with in vitro bioreactors can be reduced. Stevens et al. were the first to coin the term “in vivo bioreactor” and demonstrated the potential of de novo bone growth solely based on the creation of an artificial space between bone and
In Vivo Bioreactors
345
periosteum, which was subsequently filled with a calcium alginate gel that cross-linked in situ and provided the scaffold for powerful generation of functional native bone.3 An increasing number of studies have been aimed at the mobilization of endothelial progenitor cells (EPC) to ischemic tissues with chemotactic agents, yielding disappointing results as far as retention of the transplanted cells is concerned. Most trial protocols have attempted to mobilize progenitor cells using injections of granulocyte colony-stimulating factor (G-CSF). G-CSF, however, induces irreversible cleavage of cell–surface adhesion molecules a4-integrin and CXCR4 by proteases, thereby reducing retention of progenitor cells in the ischemic target areas. Indeed, use of alternative mobilization agents appear to be promising in this regard.11 In vivo vascularization of a given space containing a matrix or a scaffold is gaining increasing attention to prebuild a vascularized structure and deliver either pluripotent or specialized cells to it as the blood vessel tree continues to grow.12 This approach avoids the typical inevitable problem of hypoxic periods encountered in the primarily in vitro approach where cells are combined with a matrix in a bioreactor followed by transplantation into an initially avascular recipient bed where the blood supply has yet to grow into the construct. As mentioned above, EPC have been under intense investigation in the past due to their angiogenic (either directly or via secondary cytokines such as VEGF13) and vasculogenic (via direct integration and subsequent de novo blood vessel formation).14 In an attempt to increase homing of transplanted human EPC to a growing prevascularized tissue-engineered organoid in a separation chamber in a nude rat model, Simcock et al. administered the chemokine CXC chemokine ligand 12 (CXCL12) and achieved enhanced homing of xenogenic EPC to an angiogenic site developing around an arteriovenous loop inserted in a tissue engineering chamber.12 Similar approaches will undoubtedly be developed and optimized to attract vascular progenitor cells and other precursor or stem cells to the in vivo tissue engineering site.
5.19.3 5.19.3.1
Applications and Tissues Skin
The skin is the largest organ in the body and primarily functions as a barrier between the organism and the environment. Major skin loss as a result of trauma and burns has been a major concern to physicians and lead to the development of a multitude of strategies to promote skin wound healing and restore surface integrity. Among them, skin tissue engineering has gained particular attention in recent years and has led to a wide variety of commercially available skin substitutes of different compositions. Significant drawbacks, however, include high costs or the need for extensive ex vivo manipulation. The Eriksson group made the first attempt to use the skin's accessibility and inherent regenerative capacities for healing by creating a standardized and isolated in vivo liquid wound environment using a transparent flexible chamber for sealing. This system allowed the delivery of analgesics, antibiotics, growth factors via direct or cell-based gene transfer, growth media, and cells into the chamber, creating an in vivo incubator-like environment. The group demonstrated healing of clean wounds as fast as or faster than any other method, with less scarring.15 The beneficial effects were also demonstrated in a clinical study.16 Easy access to the skin has induced interest beyond wound healing and skin regeneration. Keratinocytes secrete a wide variety of proteins that have a role in skin physiology, including proteins and proteinases for matrix remodeling, growth factors, and cytokines.7 In addition, it has been demonstrated that skin cells, namely keratinocytes, secrete proteins that can reach the bloodstream and gain systemic impact. This effect was initially demonstrated by Taichman and coworkers who grafted human keratinocytes onto athymic mice and rats. Human apolipoprotein E (apoE) was thereafter detected in the systemic circulation of the animals as long as the graft remained on the animals.17 Human apoE was reduced to undetectable levels within 24 h after graft removal. The potential of using genetically altered skin cells such as keratinocytes and fibroblasts, given that a skin-derived protein attained systemic distribution, has been exploited in numerous studies since then.7 Transgene products were not only detected in the serum, but also exhibited physiological effects. A study by Meng et al.,18 in which an interleukin-10 expressing vector was injected into the dorsal skin of rats, demonstrated that the transgene was detected in the bloodstream and led to an inhibition of contact hypersensitivity at distant areas of the skin. A serious drawback of methods of transient gene transfer to exploit the skin's accessibility as a delivery route for systemically effective agents has been the fact that efficiency is low and the effect only transient, likely due to promoter silencing of the transgene and immunological responses.7 Therefore, transgenic mouse models have been used for experimental evaluation of the skin as a bioreactor, as gene expression was shown to be higher and more ubiquitous, and simultaneously eliminating a host immune response. Targeting of the therapeutic gene to specific layers of the dermis was achieved by stratum-specific promoters such as keratin 5 and keratin 14. Gene delivery exploiting the skin’s potential as an in vivo bioreactor may provide therapeutic options for metabolic diseases such as growth hormone for dwarfism, coagulation factors for coagulopathies, and insulin for diabetes mellitus.19
5.19.3.2
Bone
Treatment of large bone defects still requires autologous transfer of vital bone. Due to the discrepancy between the need on the one hand and limited supply as well as significant donor site morbidity on the other hand, generation of bone by the principles of tissue engineering has become a subject of extensive research efforts.20,21 Most experimental setups employ traditional principles
346
In Vivo Bioreactors
of tissue engineering where harvested cells are cultured in vitro, combined with porous scaffolds, and stimulated with exogenous morphogens and mitogens in vitro and/or in vivo. In contrast to this, the Langer group designed a rabbit model where an artificial space between the tibia and the adjacent periosteum was created, which was referred to as an “in vivo bioreactor.”3 Owing to the exposition of this mesenchymal layer which is rich in pluripotent cells, the creation of significant volumes of bone was possible in a predictable manner, without the necessity for delivery of growth factors or cell transplantation. Biological functionality of the newly created bone was demonstrated by successful transplantation to contralateral tibia defects, resulting in complete integration without significant donor site morbidity. This study also showed that inhibition of angiogenesis in this setting and the creation of a more hypoxic environment promoted cartilage formation. A similar model has been described by Miller and coworkers who used a plastic chamber for periosteum-guided bone formation from the rib in a sheep model.22–24 The lack of a defined vascular axis, however, limits the size and potential to transplant ectopically grown neo-bone to the desired location. Therefore, the ideal bone graft would not only include osteoconductive scaffolds, osteoinductive proteins, and osteogenic cells, but also a vascular network based on a main vascular access constituted by an artery and a vein. Prefabrication25 is defined as “the process of neovascularization by the implantation of a vascular pedicle into tissues to be transferred later to a distant site either on the pedicle or as a free graft by microvascular anastomosis.”26 Many attempts focused on using the inferior epigastric pedicle as a vascular axis to supply the in vivo bone bioreactor created around a wide variety of scaffolds consisting of hydroxyapatite27 and coralline,4 among others, in a separation chamber. A certain degree of bone formation could be demonstrated after addition of BMP-2 and establishment of a vascular network. This in vivo bone bioreactor was also shown to be useful as a model of tumor–bone interaction.28 A clinical application of in situ-generated vascularized engineered bone was reported by Terheyden’s group to repair an extended mandibular defect. They created a custom-made construct based on a titanium mesh cage filled with bone mineral blocks and infiltrated with recombinant human bone morphogenetic protein-7 (BMP-7) and the patient’s bone marrow. Thus prepared, the transplant was implanted into the latissimus dorsi muscle and 7 weeks later, transplanted as a free bone–muscle flap to repair the mandibular defect. Follow-up by skeletal scintigraphy and CT showed bone remodeling and mineralization inside the mandibular transplant both before and after in vivo transplantation, thus providing radiological evidence of new bone formation.29 In recent years, 3-D bioprinting has emerged as a new and promising tool towards generation of a variety of tissues, including bone. As reviewed by Atala, it enables precise positioning of cells, biochemical materials under spatial control and layer-by-layer positioning. Principal approaches to 3D bioprinting are referred to as biomimicry, autonomous self-assembly and mini-tissue building blocks.30 The same author and his group recently introduced an integrated tissue–organ printer (IOP) which is expected to enable generation and engineering of stable, human-scale tissue constructs of a desired shape. As proof of principle, they generated a human-sized mandible fragment based on a computer-aided design model (CAD) using data from a CT scan of a human mandible defect. Human amnionic-fluid derived stem cells were employed in combination with a hydrogel consisting of a combination of PCL and tricalcium phosphate (TCP), and Pluronic F127.31 The CAD model was translated into a program that controls motion of printer nozzles which are responsible for 3D delivery and dispension of cells into preformed microchannels. This fashion of delivery circumvents the diffusion limit of 100–200 mm for cell survival in engineered tissues. After induction of osteogenic differentiation, 3 D mandible-shaped bone formation was successfully demonstrated.31 The principle of in vitro 3D printing may efficiently be combined with the in vivo bioreactor principle in the future.
5.19.3.3
Small Intestine
The management of patients with short bowel syndrome (SBS) has been unsatisfactory with conventional treatment strategies so far. Various studies from the Vacanti laboratory demonstrated improved recovery after massive small bowel resection (SBR) using tissue-engineered intestine.32 The tissue-engineered small intestine (TESI) was constructed applying “in vivo bioreactor” principles. First, segments of intestine harvested from neonatal rats were minced and full-thickness plugs of intestine created. These so-called organoid units were then seeded onto a tubular biodegradable scaffold polymer and then wrapped into the omentum of adult rats which served as the in vivo bioreactor. These organoid units, arranged on the biodegradable scaffold, became vascularized by the omentum to form a cyst, which histologically resembled the small bowel.33 Several biochemical features of digestive enterocyte function and polarity have been demonstrated in these cysts in later studies.34 As a next step, the mature, vascularized TESI cyst was placed in continuity with the gastrointestinal tract at the time of an 85% enterectomy in rats. TESI animals returned to preoperative weight more rapidly when compared with the animals undergoing SBR alone. On histologic analysis, the TESI cysts appeared to resemble normal adjacent intestine. Furthermore, the neointestine had preserved expression of green fluorescent protein that had been introduced into the organoid units prior to implantation. These marker studies strongly suggested that the neointestine was derived primarily from the neonatal-derived intestine, as opposed to ingrowth of adjacent recipient intestine. Additional studies showed lymphangiogenesis as detected by expression of the lymphatic endothelial marker vascular endothelial growth factor receptor (VEGFR)-3 using immunohistochemistry, and the lymphangiogenic growth factor VEGF-C was quantified by ELISA. Moreover, tubular structures, which resembled lymphatics architecturally, developed, and were distinct from CD34-positive blood vessels, and lacked luminal erythrocytes. This was the first demonstration of lymphatic vessels in an engineered tissue.35 Tissue-engineered intestine also had the capacity to develop a mucosal immune system with an immunocyte population similar to that of native small intestine depending on both exposure to luminal stimuli and the duration of this exposure.36 These results demonstrate the possibility of ectopic formation of a
In Vivo Bioreactors
347
biologically functional bioartificial organoid with access to blood and lymphatic circulation, which could even be transferred to an orthotopic site with subsequent proof of maintained biological function.
5.19.3.3.1
Transplant tolerance induction and restoration of metabolic and organ function by cotransplantation of thymus with solid organs
Organ transplantation to restore impaired or lost function is hampered by the short- and long-term complications of immunosuppression and the current worldwide shortage of donor organs.37 The induction of donor-specific tolerance in allogeneic and xenogeneic transplantation offers a promising solution to these problems. The thymus has been identified as critical for systemic central tolerance to self-antigens in which potentially autoreactive T cells are deleted or anergized by exposure to the appropriate self-antigens presented either by bone marrow-derived cells or thymic stromal cells.33,34 Based on these findings and their MGH miniature swine model, David Sachs and coworkers have developed composite allografts consisting of vascularized thymic tissue and the kidney as a solid organ to induce tolerance across major histocompatibility complex (MHC)-mismatched barriers.38 A new transplantable organ (thymokidney) was generated by autologous thymic grafts under the renal capsule. The thymic grafts exhibited normal macroscopic and microscopic structure, and normal thymocyte composition. To evaluate the thymic function of the graft, a composite thymokidney was transplanted into a recipient who had previously been thymectomized, had few circulating CD4-single positive cells, and had lost MLR reactivity. The number of CD4 þ /CD45RA þ cells in this animal increased steadily indicating functionality of the thymus within the composite thymokidney. In addition, MLR assays demonstrated that the recipient recovered immunocompetence.39 Additional work from the same laboratory demonstrated that transplantation of allogeneic thymic tissue as part of a composite vascularized graft was far more successful in terms of both engraftment and long-term survival than transplantation of thymic tissue or cells alone.40 This concept of the vascularized composite neo-organ was subsequently extended to transplantation of allogeneic islets, comparing survival of islet cell suspensions to that of vascularized composite islet-kidneys (IK), prepared by injection of autologous islets underneath the renal capsule 2–3 months prior to allogeneic transplantation of the composite organ. Composite IK were transplanted successfully across minor and full MHC mismatch barriers, using treatment regimens previously demonstrated to induce long-term tolerance of kidney allografts across these barriers. IK allografts containing Z5000 islet equivalents (IE) per kilogram recipient body weight were found capable of reversing surgically induced diabetes, while injection of comparable numbers of purified islets via the portal vein or under the renal capsule did not. An additional study was directed toward preparation of autologous “thymo-islet-kidneys” (TIK), for potential use as xenografts, in which the thymic component was intended to induce tolerance and the islets to reverse diabetic hyperglycemia. To evaluate the function of the transplanted islets, three animals bearing TIK and IK underwent total pancreatectomy 3 months following islet transplantation. All three animals maintained normoglycemia thereafter. In two of these animals, the IKs were removed 2 months after the pancreatectomy, and in both cases normoglycemia was maintained thereafter by the TIK. The combination of replacement of a lost metabolic function and vascular competence of the grafts, as well as the potential for immune tolerance induction, makes this model a particularly successful example of an unusual application of an in vivo bioreactor. Use of both types of composite organ transplants (IK and TIK) may eventually be applicable to the treatment of type I diabetic patients suffering from end-stage diabetic nephropathy.40
5.19.3.4
Local Immunosuppression in Vascularized Composite Tissue Allotransplantation (VCA)
In the previous chapter, tolerance induction with thymus contransplantation was demonstrated to have potential to improve transplant survival. Up to now, systemic immunosuppression, however, constitutes the standard treatment after allogenic solid organ transplantation, even though patients suffer from serious side effects. To circumvent these drawbacks, it would appear desirable to deliver immunosuppression locally. Recently, this strategy has been tested in a rat VCA hindlimb transplantation model. The immunosuppressive drug tacrolimus was loaded into a self-assembled hydrogel. During the post-transplantation inflammation process proteolytic enzymes are overexpressed, leading to degradation of the hydrogel and thereby inducing release of the drug. In their study, Gajanayake et al. reported significantly prolonged graft survival in a Brown Norway–to–Lewis rat hindlimb transplantation model after one-time local injection of the tacrolimus-laden hydrogel.41 This therapeutic principle requires the biological in vivo bioreactor-like environment within the allograft, where a stimulusresponsive sustained release of the immunosuppressive drug can take place.42 Since VCA, eg, of the upper extremity, is not a lifesaving procedure, new therapeutic strategies to improve long term survival of the graft whilst providing more safety and reduction of side effects are particularly crucial compared to solid organ transplantation. The principle of VCA has also been combined with the evaluation of surgical angiogenesis in an experimental rabbit model of knee joint transplantation.43
5.19.3.5
Urinary Tract
Congenital and acquired ureteric defects cause significant morbidity and have been notoriously difficult to regenerate by tissue engineering approaches to create a neoureter in the past. Such attempts to generate a neoureter in animal models failed because of major inflammatory response. Avoidance of such inflammation requires a well-differentiated urothelium. Baumert et al. investigated whether omental maturation of a small intestinal submucosa (SIS) matrix seeded with previously cultured urothelial and
348
In Vivo Bioreactors
smooth muscle cells that had been derived from bladder biopsies in a pig model could achieve terminal differentiation of the urothelium to allow construction of a stricture-free neoureter. After 2 weeks of cell growth, the in vitro SIS-seeded construct was shaped around a silicone drain and wrapped by the omentum to obtain neoureters. These neoureters were left in the omentum without any contact with urine, and then harvested 3 weeks later for histologic and immunohistochemical studies. Before implantation, the in vitro constructs were composed of a mono- or bilayer of undifferentiated urothelium overlying a monolayer of smooth muscle cells. After 3 weeks of omental maturation, these constructs were vascularized and comprised a terminally differentiated multilayered urothelium with umbrella cells over connective tissue and smooth muscle cells, with no evidence of fibrosis or inflammation. This model of in vivo maturation using the omentum as an in vivo bioreactor enabled for the first time the generation of a mature neoureter composed of a well-differentiated multilayered urothelium.44 When looking at other parts of the genitourinary system, bladder regeneration is another challenging goal. One of the main problems is to limit the development of ischemic fibrosis during tissue maturation. Similar to their ureter tissue engineering, Baumert et al. described a model using the omentum as an in vivo bioreactor for a previously seeded scaffold. A highly vascularized tissue-engineered construct was generated that contracted in response to acetylcholine stimulation. The wall thickness was 4 mm, on average. Histologic and immunostaining analysis of the construct confirmed the presence of a multilayer urothelium on the luminal aspect and deeper fascicles of organized tissue composed of differentiated smooth muscle cells and mature fibroblasts without evidence of inflammation or necrosis. Large- and small-diameter vessels were clearly identified histologically in the tissue obtained. The omentum as an in vivo incubator permitted in vivo maturation of seeded scaffolds with the development of a dense vascularization that is anticipated to prevent fibrosis and loss of contractility. This in vivo maturation into the omentum could be the first step before in situ implantation of the construct.45
5.19.3.5.1
The arteriovenous loop model: An “in vivo bioreactor” to generate vascularized tissues and organs
The development of a functional and adequate blood vessel network is a prerequisite for successful application of engineered tissues and organs. When in vitro-engineered constructs containing cells are transferred in vivo, they lack a defined axial blood supply consisting of a main arterial inflow and a venous outflow, respectively, and therefore rely on blood vessel ingrowth from their surroundings, that is, interstitial fluid diffusion and blood perfusion.20,21 This limitation in blood supply constitutes the main drawback when transferring in vitro tissue engineering approaches to the in vivo environment. Cell support via diffusion occurs only up to 200 mm into the matrix,46 resulting in insufficient nutrition and subsequent loss of cells located in the center of such in vitroengineered constructs. For these reasons, induction of vascularization is essential in virtually all successful tissue engineering applications, and a defined vascular axis is desirable whenever the bioengineered tissue or organ is yet to be transplanted to a different site within the host. In order to enable standardized transfer techniques of neo-tissues and organs to a different site, reconstructive surgeons have therefore aimed to generate axially vascularized tissues and organoids that are supplied by a defined arteriovenous axis and could hence be transferred to the desired site using microsurgical techniques of vascular anastomosis. As opposed to other techniques, these tissues are immediately connected to the systemic blood circulation upon implantation as free flaps are. The arteriovenous loop model based on the femoral vessels, including a venous graft from the contralateral side in a rat separation chamber, fulfills our criteria to constitute an in vivo bioreactor. The studies presented in this chapter are based on the initial description by Erol and Spira47 who successfully generated a prefabricated skin flap by creating an arteriovenous vessel loop using either arterial or venous grafts in a rat model. They observed the outgrowth of a dense 3D neovessel tree originating from the entire AV loop. This model was later further developed by isolation of the AV loop into isolation chambers and induction of vascularization in polymer matrices.48–50 Based on these works, a custom-made processed bovine cancellous bone matrix was successfully vascularized by means of the AV loop.20,21 A detailed analysis of the vascularization was established using invasive techniques such as corrosion casting as well as noninvasive methods such as micro-CT scanning.51 Using histology-based morphometric analysis, vascularization was meticulously quantified and it was demonstrated that more than 90% of the scaffold vascularized 8 weeks following implantation. Prevascularization of these matrices significantly improved survival of osteoblasts engrafted into the matrix, emphasizing the importance of an intrinsic blood supply for cell vitality within the construct.51 The experimental model has also been successfully modified to allow for combined intrinsic with extrinsic vascularization52 which was shown to significantly accelerate axial vascularization of a matrix within the isolation chamber. The AV loop model to vascularize tissue-engineered constructs is not limited to hard matrices and the attempt to create bone, but may be extended by using other matrices and cell populations. A fibrin gel approved for clinical use can be employed to generate a fibrin matrix, which was successfully vascularized and may be used as a platform to create soft tissues or organoids. The vascularization within this fibrin matrix was shown to respond to proangiogenic stimuli by accelerating the vascular outgrowth when the recombinant growth factors VEGF and bFGF were suspended in the matrix.53,54 Transplantation of myoblasts into a previously vascularized fibrin matrix containing the AV loop enhanced cell survival and maintenance of the myogenic phenotype compared to fibrin matrices without AV loops, even though no formation of skeletal muscle-like tissue could be observed, possibly due to the lack of potent myogenic stimuli.55 The vascularized matrix embedded in the AV loop may also be employed to create organoids or replace organ functions. As a new approach to hepatic tissue engineering, fetal liver cells (FLC) were suspended in the fibrin matrix, resulting in neo-tissue formation containing viable fetal donor cells in the highly vascularized fibrin matrix after 14 days.1 Previous work demonstrated that continuous hepatotrophic stimulation is necessary in tissue-engineered liver support systems, and it has been demonstrated that cocultivation with pancreatic islets may have stimulatory effects on hepatocytes and that they are mediated by soluble factors, and are dependent on insulin and glucagon.56,57
In Vivo Bioreactors
349
Fig. 2 Micro-CT scan from a rat AV-loop following Microfil – perfusion and explantation, demonstrating vascular sprouting originating from the AV-loop.
Further investigations toward 3D transplantable hepatocyte-islet devices for continuous in vivo stimulation were conducted to provide hepatotrophic stimulation. Hepatocyte and pancreatic islets were coseeded into polyvinyl alcohol matrices and supported hepatotrophic stimulation of bioartificial liver equivalents equally well, compared to a protocaval shunt operation. The AV loop model may therefore well serve as a platform for the development of highly vascularized in vivo tissue engineering-based liver support systems using these cotransplantation strategies. Successful engineering of cardiac muscles has been achieved by matrigel-suspended neonatal rat cardiomyocytes implanted inside the AV loop. The presence of gap junctions was shown between the cardiomyocytes at 4 and 10 weeks, and echocardiograms showed spontaneous contraction of the tissue in vivo. In vitro, positive chronotropy to norepinephrine and positive inotropy in response to calcium demonstrated biological function of the cardiac neo-organ.58 The AV loop separation chamber potential is not limited by the availability of local factors in its in situ tissue and organ engineering capacity. The regenerative potential of the established chamber construct may be even better exploited by chemotactic recruitment of supportive cells or signaling molecules. It was recently demonstrated that EPC that had been systemically injected before can be recruited from the circulation by CXCL12/stromal cell-derived factor 1,12 potentially enabling further modulation of the vascularization process. The isolation chamber setting with its potential to be monitored and manipulated may also be used as an investigation platform to investigate the phenomena and regulation of angiogenesis and antiangiogenesis in cancer research.5 Limitations to dimensions in small animal models have always been an argument against successful transfer of tissue engineering applications to clinical relevance. Consequently, a large animal model (sheep) was established where de novo formation of axially vascularized tissue was demonstrated for the first time ever by sequential computed tomography angiography (CTA) and magnetic resonance angiography (MRA) in vivo, as well as by postexplantational micro-computed tomography (Fig. 2) and histology.59 This constitutes further progress in approaching the first application of tissue engineering vascularized grafts with clinically relevant dimensions.59–61 In conclusion, the arteriovenous loop model has been investigated mainly as a powerful tool to create bioartificial tissues and organoids, but may also be very useful as a platform to investigate angiogenetic and antiangiogenetic phenomena in general, with particular emphasis on cancer therapy.5
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
Fiegel, H. C.; Pryymachuk, G.; Rath, S.; et al. J. Cell. Mol. Med. 2010, 14 (1–2), 267–274. Fröhlich, M.; Grayson, W. L.; Wan, L. Q.; Marolt, D.; Drobnic, M.; Vunjak-Novakovic, G. Curr. Stem Cell Res. Ther. 2008, 3 (4), 254–264. Stevens, M. M.; Marini, R. P.; Schaefer, D.; Aronson, J.; Langer, R.; Shastri, V. P. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (32), 11450–11455. Holt, G. E.; Halpern, J. L.; Dovan, T. T.; Hamming, D.; Schwartz, H. S. J. Orthop. Res. 2005, 23 (4), 916–923. Hutmacher, D. W.; Horch, R. E.; Loessner, D.; et al. J. Cell. Mol. Med. 2009, 13 (8A), 1417–1427. LaMattina, J. C.; Kumagai, N.; Barth, R. N.; et al. Transplantation 2002, 73 (5), 826–831. Cao, T.; Wang, X. J.; Roop, D. R. Hum. Gene Ther. 2000, 11 (16), 2297–2300. Hussey, A. J.; Winardi, M.; Han, X.; et al. Tissue Eng. A 2009, 15 (12), 3823–3833. Polykandriotis, E.; Arkudas, A.; Euler, S.; Beier, J. P.; Horch, R. E.; Kneser, U. Handchir. Mikrochir. Plast. Chir. 2006, 38 (4), 217–223. (Article in German). Discher, D. E.; Mooney, D. J.; Zandstra, P. W. Science 2009, 324 (5935), 1673–1677. Cheng, M.; Qin, G. Prog. Mol. Biol. Transl. Sci. 2012, 111, 243–264. Simcock, J. W.; Penington, A. J.; Morrison, W. A.; Thompson, E. W.; Mitchell, G. M. Tissue Eng. A 2009, 15 (3), 655–664.
350
13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61.
In Vivo Bioreactors
Asahara, T.; Takahashi, T.; Masuda, H.; et al. EMBO J. 1999, 18 (14), 3964–3972. Tepper, O. M.; Capla, J. M.; Galiano, R. D.; et al. Blood 2005, 105 (3), 1068–1077. Eriksson, E.; Vranckx, J. Am. J. Surg. 2004, 188 (1A Suppl.), S36–S41. Vranckx, J. J.; Slama, J.; Preuss, S.; et al. Plast. Reconstr. Surg. 2002, 110 (7), 1680–1687. Fenjves, E. S.; Gordon, D. A.; Pershing, L. K.; Williams, D. L.; Taichman, L. B. Proc. Natl. Acad. Sci. U.S.A. 1989, 86 (22), 8803–8807. Meng, X.; Sawamura, D.; Tamai, K.; Hanada, K.; Ishida, H.; Hashimoto, I. J. Clin. Invest. 1998, 101 (6), 1462–1467. Taichman, L. B. Proc. Assoc. Am. Phys. 1999, 111 (3), 206–210. Kneser, U.; Polykandriotis, E.; Ohnolz, J.; et al. Tissue Eng. 2006, 12 (7), 1721–1731. Kneser, U.; Schaefer, D. J.; Polykandriotis, E.; Horch, R. E. J. Cell. Mol. Med. 2006, 10 (1), 7–19. Cheng, M. H.; Brey, E. M.; Allori, A. C.; et al. Plast. Reconstr. Surg. 2009, 124 (3), 787–795. Miller, M. J.; Goldberg, D. P.; Yasko, A. W.; et al. Tissue Eng. 1996, 2 (1), 51–59. Tatara, A. M.; Kretlow, J. D.; Spicer, P. P.; et al. Tissue Eng. Part A 2015, 21 (9–10), 1520–1528. Guo, L.; Pribaz, J. J. Plast. Reconstr. Surg. 2009 Dec, 124 (6 Suppl.), e340–e350. Gill, D. R.; Ireland, D. C.; Hurley, J. V.; Morrison, W. A. J. Hand Surg. Am. 1998, 23 (2), 312–321. Mizumoto, S.; Inada, Y.; Weiland, A. J. J. Reconstr. Microsurg. 1993, 9 (6), 441–449. Halpern, J.; Lynch, C. C.; Fleming, J.; et al. Clin. Exp. Metastasis 2006, 23 (7–8), 345–356. Warnke, P. H.; Springer, I. N.; Wiltfang, J.; et al. Lancet 2004, 364 (9436), 766–770. Murphy, S. V.; Atala, A. Nat. Biotechnol. 2014, 32 (8), 773–785. Kang, H. W.; Lee, S. J.; Ko, I. K.; et al. Nat. Biotechnol. 2016, 34 (3), 312–319. Grikscheit, T. C.; Siddique, A.; Ochoa, E. R.; et al. Ann. Surg. 2004, 240 (5), 748–754. Duxbury, M. S.; Grikscheit, T. C.; Gardner-Thorpe, J.; et al. Transplantation 2004, 77 (8), 1162–1166. Perez, A.; Grikscheit, T. C.; Blumberg, R. S.; Ashley, S. W.; Vacanti, J. P.; Whang, E. E. Transplantation 2002, 74 (5), 619–623. Greenstein, J. L.; Sachs., D. H. Nat. Biotechnol. 1997, 15 (3), 235–238. Kappler, J. W.; Roehm, N.; Marrack, P. Cell 1987, 49 (2), 273–280. Sprent, J.; Paul, W. E. Fundamental Immunology, Vol. 3. Raven: New York, 1993; pp. 74–109. Yamada, K.; Shimizu, A.; Utsugi, R.; et al. J. Immunol. 2000, 164 (6), 3079–3086. Yamada, K.; Shimizu, A.; Ierino, F. L.; et al. Transplantation 1999, 68 (11), 1684–1692. Vallabhajosyula, P.; Griesemer, A.; Yamada, K.; Sachs, D. H. Cell Biochem. Biophys. 2007, 48 (2–3), 201–207. Gajanayake, T.; Olariu, R.; Leclère, F. M.; et al. Sci. Transl. Med. 2014, 13 (6), 249. Vemula, P. K.; Boilard, E.; Syed, A.; et al. J. Biomed. Mater. Res. A 2011, 97, 103–110. Kremer, T.; Giusti, G.; Friedrich, P. F.; et al. Microsurgery 2012, 32 (2), 118–127. Baumert, H.; Mansouri, D.; Fromont, G.; et al. Eur. Urol. 2007, 52 (5), 1492–1498. Baumert, H.; Simon, P.; Hekmati, M.; et al. Eur. Urol. 2007, 52 (3), 884–890. Folkman, J.; Hochberg, M. J. Exp. Med. 1973, 138 (4), 745–753. Erol, O. O.; Spira, M. Surg. Forum 1979, 30, 530–531. Mian, R.; Morrison, W. A.; Hurley, J. V.; et al. Tissue Eng. 2000, 6 (6), 595–603. Hofer, S. O.; Knight, K. M.; Cooper-White, J. J.; et al. Plast. Reconstr. Surg. 2003, 111 (3), 1186–1192; (discussion 1193–1194). Cassell, O. C.; Morrison, W. A.; Messina, A.; et al. Ann. N. Y. Acad. Sci. 2001, 944, 429–442. Weis, C.; Covi, J. M.; Hilgert, J. G.; et al. J. Microsc. 2015, 259 (3), 185–196. doi:10.1111/jmi.12252. Arkudas, A.; Pryymachuk, G.; Beier, J. P.; et al. Plast. Reconstr. Surg. 2012, 129 (1), 55e–65e. Arkudas, A.; Beier, J. P.; Heidner, K.; et al. Tissue Eng. 2007, 13 (7), 1549–1560. Arkudas, A.; Pryymachuk, G.; Hoereth, T.; et al. Tissue Eng. A 2009, 15 (9), 2501–2511. Bach, A. D.; Arkudas, A.; Tjiawi, J.; et al. J. Cell. Mol. Med. 2006, 10 (3), 716–726. Kaufmann, P. M.; Fiegel, H. C.; Kneser, U.; Pollok, J. M.; Kluth, D.; Rogiers, X. Tissue Eng. 1999, 5 (6), 583–596. Kaufmann, P. M.; Kneser, U.; Fiegel, H. C.; et al. Transplantation 1999, 68 (2), 272–279. Morritt, A. N.; Bortolotto, S. K.; Dilley, R. J.; et al. Circulation 2007, 115 (3), 353–360. Beier, J. P.; Horch, R. E.; Arkudas, A.; et al. Microsurgery 2009, 29 (1), 42–51. Weigand, A.; Beier, J. P.; Hess, A.; et al. Tissue Eng. Part A 2015, 21 (9–10), 1680–1694. Boos, A. M.; Loew, J. S.; Weigand, A.; et al. J. Tissue Eng. Regen. Med. 2013, 7 (8), 654–664.