Coupling Osteogenesis and Vasculogenesis in Engineered Orthopedic Tissues

Abstract

Inadequate vascularization of engineered tissue constructs is a main challenge in developing a clinically impactful therapy for large, complex, and recalcitrant bone defects. It is well established that bone and blood vessels form concomitantly during development, as well as during repair after injury. Endothelial cells (ECs) and mesenchymal stromal cells (MSCs) are known to be key players in orthopedic tissue regeneration and vascularization, and these cell types have been used widely in tissue engineering strategies to create vascularized bone. Coculture studies have demonstrated that there is crosstalk between ECs and MSCs that can lead to synergistic effects on tissue regeneration. At the same time, the complexity in fabricating, culturing, and characterizing engineered tissue constructs containing multiple cell types presents a challenge in creating multifunctional tissues. In particular, the timing, spatial distribution, and cell phenotypes that are most conducive to promoting concurrent bone and vessel formation are not well understood. This review describes the processes of bone and vascular development, and how these have been harnessed in tissue engineering strategies to create vascularized bone. There is an emphasis on interactions between ECs and MSCs, and the culture systems that can be used to understand and control these interactions within a single engineered construct. Developmental engineering strategies to mimic endochondral ossification are discussed as a means of generating vascularized orthopedic tissues. The field of tissue engineering has made impressive progress in creating tissue replacements. However, the development of larger, more complex, and multifunctional engineered orthopedic tissues will require a better understanding of how osteogenesis and vasculogenesis are coupled in tissue regeneration.

Impact statement

Vascularization of large engineered tissue volumes remains a challenge in developing new and more biologically functional bone grafts. A better understanding of how blood vessels develop during bone formation and regeneration is needed. This knowledge can then be applied to develop new strategies for promoting both osteogenesis and vasculogenesis during the creation of engineered orthopedic tissues. This article summarizes the processes of bone and blood vessel development, with a focus on how endothelial cells and mesenchymal stromal cells interact to form vascularized bone both during development and growth, as well as tissue healing. It is meant as a resource for tissue engineers who are interested in creating vascularized tissue, and in particular to those developing cell-based therapies for large, complex, and recalcitrant bone defects.

Introduction: Cellular Interactions in Bone Regeneration

The physiological process of bone regeneration involves a coordinated interplay of tissue development, maturation, and remodeling. Osteogenesis and vasculogenesis are two of the key elements in recreating stable and mature bone, and it is well established that the cells responsible for these processes interact during tissue development, regeneration, and homeostasis. The general anatomy of bone and associated blood vessels is shown schematically in Figure 1. The cells in bone are nourished by blood vessels, which regulate the transport of oxygen, nutrients, and immune components that maintain tissue homeostasis. Blood vessels are lined with endothelial cells (EC) that serve an antithrombotic function, and also regulate the permeability of vessels to cells and cytokines.¹ Mesenchymal stromal cells (MSCs) are multipotent stromal cells present in most adult connective tissues that serve as a reservoir of precursor cells to maintain and repair damaged tissues. Physiologically, MSCs have a close spatial relationship with ECs and numerous studies have shown that MSCs can be isolated from the perivascular space of multiple organs, including bone marrow, skeletal muscle, lung, pancreas, adipose, and placental tissues.^2,3 This close relationship has motivated the development of strategies that exploit EC–MSC interactions for the regeneration of a variety of tissues.

FIG. 1.

General anatomy of bone. A macroscopic-to-microscopic schematic shows how blood vessels are integrated into bone tissue to support the resident cells in both cancellous and cortical bone. ECs secrete osteogenic factors and osteoblasts secrete angiogenic factors during native bone regeneration. Figure adapted from Lopes et al.¹¹² and Grosso et al.¹⁷ ECs, endothelial cells. Color images are available online.

MSCs are an important cell type in bone regeneration applications because of their demonstrated ability to create mineralized tissue,⁴ and their involvement in bone formation.^5,6 It has also been shown that the incorporation of ECs into osteogenic constructs can improve their regenerative potential after transplantation.⁷ A number of studies have demonstrated that ECs can modulate the differentiated state of MSCs,⁸ including the potentiation of an osteogenic phenotype.⁹ Conversely, MSCs have been shown to play a role in new vessel formation by supporting the development and maturation of EC networks through proangiogenic cytokine secretion and pericyte-like interactions.^10,11 The development of blood vessels is critical for the success of tissue engineering bone constructs as a robust vascular network is essential for tissue survival after transplantation, especially for larger-sized constructs in which nutrient diffusion becomes a limiting factor for maintaining appropriate viability. The interactions between these cell types in the context of bone regeneration are not completely understood, but it is clear that both osteogenesis and vasculogenesis are required to form stable and mature tissues.

Despite an increase in research aimed at creating vascularized bone constructs, efforts to understand how EC–MSC interactions can be manipulated to enhance both osteogenesis and vasculogenesis concurrently are still lacking. The sections below focus on the physiological processes necessary for the formation of both bone and vasculature in development and after injury. Furthermore, this review describes tissue engineering strategies using ECs and MSCs to create vascularized bone constructs, with an emphasis on recent techniques in developmental engineering, including remaining challenges and future perspectives.

The Physiological Development of Bone

Bone is a highly vascularized tissue and therefore blood vessel formation is a key component in bone development.¹² Bone-forming cells secrete proangiogenic factors, such as vascular endothelial growth factor (VEGF) that can stimulate signaling responses in multiple cell populations expressing VEGF receptors, including EC that make up blood vessels as well as chondrocytes, osteoblasts, and osteoclasts.^13–15 Conversely, vessel-forming cells present during bone regeneration release factors, such as bone morphogenetic protein (BMP)-2 and BMP-4 that act on chondrocytes and cells of the osteoblast lineage.^16–18 Given the intimate link between bone regeneration and vessel formation, understanding the cellular mechanisms behind these physiological processes is critical for the successful development of a vascularized bone construct.

Bone formation

The development of bone occurs through two distinct pathways: (1) the intramembranous pathway in which mesenchymal progenitor cells (MPCs) differentiate directly into bone-forming osteoblasts, or (2) the endochondral ossification pathway in which MPCs first differentiate into chondrocytes to form a cartilaginous template, followed by a transformation into calcified bone tissue. Generally speaking, intramembranous ossification is the process of development and growth of flat bones of the cranium and medial clavicles, whereas endochondral ossification occurs in the development of long bones.¹⁹ Each pathway involves a distinct set of biochemical cues, as outlined in Figure 2.

FIG. 2.

Pathways of bone formation during development. (A) Intramembranous ossification involves direct differentiation of mesenchymal progenitor cells into an osteogenic lineage. (B) Endochondral ossification involves formation of a cartilaginous template that undergoes hypertrophy, followed by transformation into bone. Figure adapted from Einhorn and Gerstenfeld⁶ and Almubarak et al.²⁹ Color images are available online.

Intramembranous ossification (Fig. 2A) is controlled through the coregulation of two critical classes of signaling proteins that induce MPC commitment directly toward an osteoblastic lineage: BMPs and wingless-related integration site (Wnt) proteins.⁶ These signaling cues direct cells toward an osteoblastic lineage through induction of the transcription factors runt-related transcription factor 2 (RUNX2), homeobox protein DLX5, and osterix (OSX).²⁰ As MPCs mature into bone-forming osteoblasts, they proliferate and condense into compact nodules. Osteoblasts then secrete a collagen/proteoglycan matrix, which is capable of binding calcium salts, ultimately causing mineralization of the bone matrix.²¹ Increased alkaline phosphatase (ALP) activity, and osteopontin (OPN) and osteocalcin (OCN) protein secretion are characteristic markers of this osteoblastic progression.²²

Endochondral ossification (Fig. 2B) is a multistage process beginning with the upregulation of key BMP signaling proteins necessary for chondrogenic lineage commitment.⁶ This stimulus leads to the condensation of MPCs and activation of SOX9, a transcription factor required to maintain chondrogenesis, promote proliferation, and induce collagen matrix deposition.^23–25 These immature chondrocytes then form a rudimentary cartilage template by secreting a matrix rich in collagen II and the proteoglycan aggrecan.²⁶ The next stage of maturation into prehypertrophic chondrocytes is marked by the expression of both parathyroid hormone 1 receptor and Indian hedgehog. These chondrocytes then cease proliferation, increase in size, and begin to mineralize the cartilaginous matrix.^27,28 The phenotype of hypertrophic chondrocytes is distinct from earlier maturation states and is characterized by the expression of VEGF-A, matrix metalloproteinase 13 (MMP-13), and secreted phosphoprotein 1 (also known as OPN/bone sialoprotein; SPP1).^26,29 This phenotype is also associated with markers of bone formation, including RUNX2, ALP, collagen I, OSX, OCN, and OPN.^30–33 It has been proposed that hypertrophic chondrocytes exit the cell cycle and represent a terminally differentiated cell state.^34,35 According to this model, hypertrophic chondrocytes undergo programmed cell death, and osteoblasts or progenitor cells are delivered through invading vasculature to form the bone matrix that replaces the cartilage.³⁶ However, an alternative mechanism has also been proposed, in which hypertrophic chondrocytes directly transition into osteoblasts and osteocytes. This differentiation pathway has been supported by several studies using lineage tracing techniques.^37–40

The repair of bone fractures occurs by either intramembranous or indirect fracture healing, which consists of both intramembranous and endochondral bone formation.^41,42 Indirect healing is the most common pathway as direct bone healing requires anatomical reduction and rigidly stable conditions, commonly only obtained by open reduction and internal fixation. However, when such conditions are achieved, the direct healing cascade permits immediate bone regeneration without any remodeling steps necessary. In all other nonstable conditions, bone healing follows a specific biological pathway, outlined schematically in Figure 3. This process begins with the formation of a hematoma at the defect site and infiltration of MSCs. Intramembranous ossification immediately occurs near the cortex and in regions of the periosteum. The external soft tissues stabilize the fracture by the formation of a callus. This callus subsequently undergoes chondrogenesis, and then a process highly similar to endochondral ossification. The phenotype of MSCs and the resulting bone-forming pathway is believed to be regulated by a variety of factors, including the relative distance to blood vessels, mechanical cues induced by fracture stability, and the spatial and temporal release profile of cytokines and growth factors.

FIG. 3.

Schematic of the regenerative niche following injury. The approximate timing and main stages of fracture healing. The process begins with an initial inflammatory phase and formation of a hematoma at the defect site. Surrounding vasculature invades the fibrous matrix and supplies cells responsible for subsequent callus formation. This cartilaginous callus becomes mineralized and eventually is remodeled, giving rise to new trabecular and cortical bone. Color images are available online.

Blood vessel formation

Understanding the process of vascularization is fundamental to the field of tissue engineering since large, functional tissues require a blood supply to remain viable after implantation.⁴³ During embryonic development, blood vessel formation occurs through vasculogenesis, the de novo assembly of tubular structures by ECs, which establishes a primitive vascular plexus. These vessel networks continue to expand through sprouting from pre-existing vessels to form new capillaries, a process known as angiogenesis.⁴⁴ Vascular development is a dynamic process involving interactions between ECs and supporting pericytes to form a functional and mature vessel network, as illustrated schematically in Figure 4. In cases of tissue injury and repair, revascularization through angiogenesis is modulated by oxygen tension in the local microenvironment. When exposed to low oxygen conditions, cells in the surrounding tissue experience an increase in hypoxia-inducible factor-1α, which leads to downstream transcriptional cascades resulting in an upregulation of vascular protein secretion, such as VEGF. Gradients of VEGF and other angiogenic factors spatially and temporally control the invasion of sprouting vessels into the ischemic tissue to restore blood supply.^45–47

FIG. 4.

Cellular interactions between ECs and MSCs in blood vessel formation. ECs form the inner lumen of the vessel, while MSCs differentiate into pericyte-like cells and promote maturation and stabilization of vasculature. Figure adapted and used by permission from Melchiorri et al.⁴⁵ MSCs, mesenchymal stromal cells. Color images are available online.

Sprouting angiogenesis involves the proliferation and migration of ECs to form neovessels, which either regress or proceed to maturity through interactions with pericytes.⁴⁷ Although the identity of pericytes is not well defined, these cells are generally described as periendothelial stromal cells that play a role in vascular development and homeostasis.⁴⁸ It has been demonstrated that while not all MSCs exhibit pericyte behavior, specific subpopulations do interact with ECs and take on a pericyte-like role through paracrine and direct cell–cell signaling.⁴⁵ The initial, immature tubule structures formed by ECs are stabilized by VE-cadherin junctional complexes with neighboring ECs.⁴⁹ As these structures mature, ECs secrete insulin-like growth factor-1 (IGF-1), platelet-derived growth factor (PDGF), heparin-binding epidermal growth factor (HB-EGF), and stromal-derived factor 1α (SDF-1α). These growth factors bind to receptors on local pericytes and promote migration of and adhesion to the developing vessels.^45,50 Maturing ECs also secrete sphingosine-1 phosphate, which strengthens EC–pericyte interactions through the expression of N-cadherins.⁴⁷ The presence of pericytes upregulates transforming growth factor (TGF)-β secretion, which inhibits EC proliferation and promotes pericyte differentiation.⁵¹ Pericytes also aid in vessel maturation through extracellular matrix (ECM) deposition to establish the vascular basement membrane, which provides mechanical stability and serves as a scaffold for signaling proteins.⁵² The formation of a basement membrane is orchestrated by both ECs and pericytes; however, pericytes serve as the primary contributor as they deposit many critical ECM components, including collagen IV, fibronectin, laminins, and nidogens.⁵³

The vasculature of bone appears to be formed mainly through angiogenesis, during both tissue development and repair.¹² After injury, the ingrowth of blood vessels into damaged bone tissue is essential for the formation of the soft callus, which is then remodeled into new bone through endochondral ossification. Hypertrophic chondrocytes and other osteoprogenitor cells secrete proangiogenic signals, such as isoforms of VEGF and other factors, which initiate the ingrowth of vasculature and formation of an initial vascular network.^12,54,55 As bone tissue continues to develop and remodel, signals released by maturing and hypertrophic chondrocytes further promote vessel growth and ossification.

Tissue Engineering Strategies to Regenerate Vascularized Bone

While bone tissue does have an inherent capacity to regenerate after injury, the healing of large and complex bone wounds remains a clinical challenge.^5,6 The inherent limitations of current bone grafting techniques have motivated the development of alternative tissue engineering strategies for bone regeneration. Initial strategies applied a range of natural or synthetic materials to create tissue-compatible scaffolds and constructs, which were often loaded with active molecules to encourage host cell infiltration and subsequent ossification. These strategies led to the development of several Food and Drug Administration-approved products, but these are generally limited to applications involving only mild trauma and small bone defects.⁵⁶ To address larger and more complex bone defects, the field of orthobiologics has developed with the goal of enhancing regeneration by incorporating more functional growth factors and cells into materials and engineered constructs before implantation. Such cell-based approaches are of particular interest, especially when intrinsic bone regeneration responses are diminished by age, health, or other factors.⁵⁷

Despite the enormous potential of orthobiologic regeneration strategies, the lack of adequate and timely vasculature has been identified as a major challenge preventing its clinical translation. Vascularization is critical for the success of native bone regeneration, as blood vessels supply oxygen, nutrients, and osteoprogenitor cells necessary for bone healing and remodeling after injury.^58,59 When tissue-engineered constructs are implanted, the seeded cells have limited capacity to both uptake substrate molecules (oxygen, glucose, and amino acids) and to clear byproducts of metabolism (CO₂, lactate, and urea).⁶⁰ These limitations can impair cell viability and hinder the regenerative success of engineered constructs.^58,59 Therefore, the field of bone tissue engineering has actively pursued strategies that encourage not only bone matrix deposition and mineralization, but also concomitant vessel formation and maturation.

MSCs are used widely in the field of bone tissue engineering because of their central role in native bone regeneration, ease of isolation and in vitro expansion, and well-documented capacity to undergo controllable osteogenesis both in vitro and in vivo. MSCs are also widely used as supportive pericytes in vascular tissue engineering and play an instrumental role in the process of vessel formation by secreting angiogenic factors and directly interacting with ECs to stabilize vasculature. This dual potential suggests that MSCs are an ideal cell type to create both osseous and vascular components within a single engineered tissue construct. However, to successfully exploit this potential, MSCs must receive the appropriate morphogenic cues to separately but concurrently guide both osteogenesis and vessel formation, in conjunction with ECs. Control of this complex, multicellular system is an engineering challenge that requires a thorough understanding of the induction of both osteogenesis and vasculogenesis, and then applying this knowledge to generate a dual-phase system with concomitant bone and blood vessel formation.

MSC-mediated bone regeneration

Cell-based bone tissue engineering strategies often apply MSCs in combination with a synthetic or natural biomaterial scaffold designed to direct cellular migration, proliferation, and differentiation upon implantation.⁶¹ Generally speaking, scaffolds should promote spatially uniform MSC seeding and appropriate cell loading for clinical application.⁶² Cell seeding efficiency can be regulated by modifying the porosity, interconnectivity, and surface treatment of carrier materials.^63–65 Proliferation after seeding is influenced by contact inhibition between adjacent cells, which is largely determined by initial seeding density.⁶⁶ Successful MSC-based bone regeneration therapies require transplantation of sufficient cell numbers to potentiate the healing process, although the optimal cell number for a given application is not clear.⁵⁹ Too low a seeding density may compromise cellular contact and paracrine signaling, whereas too high a density may result in limited nutrient transport and insufficient waste removal.⁶⁷ The local conditions at the site of implantation will also affect the number of cells that can be initially supported, and how those cells will engraft, proliferate, and function.

The most appropriate phenotype for transplanted cells is also a matter of consideration. It has been suggested that undifferentiated MSCs are not strongly osteogenic when implanted, as compared with MSCs that have undergone in vitro preculturing to induce appropriate lineage commitment.^68,69 It has been shown that undifferentiated MSCs elicit an inhibitory effect on osteoblast differentiation and proliferation in vitro,⁷⁰ including through the secretion of proinflammatory factors, which are shown to repress osteogenesis.^71,72 In contrast, preconditioning of MSCs in culture to promote lineage-specific differentiation may reduce these inhibitory effects and promote osteogenic function. Osteogenic lineage commitment is commonly enhanced using standard medium supplements, including beta-glycerophosphate, ascorbic acid, and dexamethasone. Preculture with these biochemical factors has been shown to promote the bone-forming capacity of MSCs in several in vivo bone defect models.^73–75 Figure 5 shows results from a study in which osteogenically induced MSCs were compared with controls in a subcutaneous implant in a heterotypic nude mouse model, demonstrating that predifferentiation can significantly potentiate bone formation.⁷⁵ The degree and length of osteogenic preculture is also an important variable, which again depends on the therapeutic approach and intended application.⁵⁹ Short-term osteogenic induction may be insufficient for achieving adequate cell differentiation to a preosteoblastic phenotype, whereas longer durations have been shown to decrease bone-forming potential, perhaps due to a decrease in cell proliferation capacity as cells become more strongly differentiated.⁷⁶

FIG. 5.

Osteogenic predifferentiation of MSCs improves bone-forming potential in vivo. BMSCs were expanded in either basal or osteoinductive medium for 3 weeks before seeding on beta-tricalcium phosphate scaffolds and implantation subcutaneously in a heterotypic nude mouse model for up to 20 weeks. Freshly isolated BMNCs were used as a control. Constructs seeded with osteo-induced BMSCs produced more robust bone regeneration, relative to both noninduced BMSC- and BMNC-seeded constructs, as shown by greater implant density quantified from microcomputed tomography images. Scale bar = 5 mm. *p < 0.05 compared with both the noninduced BMSCs and BMNCs. Figure adapted and used by permission from Ye et al.⁷⁵ BMSC, bone marrow-derived MSC; BMNC, bone marrow mononucleated cell. Color images are available online.

Vascularization strategies using ECs and MSCs

A variety of models of angiogenesis and vasculogenesis have been used to study blood vessel formation in vitro and in vivo, in an effort to understand the cell types and factors that can be used to promote perfusion of engineered tissues.^43,44 The interactions between ECs and various support cell types have been of particular interest, since pericyte identity and phenotype influence the rate and quality of vessel formation.^77,78 Multiple cell types have been shown to induce vascularization in tissue engineering constructs, including fibroblasts, smooth muscle cells (SMCs), and MSCs. However, harvesting of mature SMCs from the walls of large vessels presents a clinical challenge,⁷⁹ and fibroblasts may not produce vessels of the same quality as other support cell types.⁷⁷ The use of MSCs is therefore of particular interest in tissue engineering, since these cells are abundantly available, secrete proangiogenic factors,^11,80,81 and have the ability to differentiate toward an SMC lineage.^82,83

MSCs can indirectly influence the formation of neovessels by secreting a variety of cytokines and growth factors. Specifically, MSCs have been shown to secrete VEGF, basic fibroblast growth factor (FGF), placental growth factor, monocyte chemotactic protein 1, and interleukin (IL)-6,⁸⁴ all of which play important roles in vessel assembly by independently or synergistically influencing EC proliferation and migration, tubule formation, and vascular stability.^11,51 Recent studies have further explored culture methods that can be used to enhance the proangiogenic potential of MSCs. Maintenance of MSCs in hypoxic conditions has been applied to enhance the secretion of proangiogenic cytokines and overall exosome production.^{45,81,85–87} It has also been demonstrated that culture of MSCs as spheroids can enhance secretion of proangiogenic factors.^81,87–89

Direct heterotypic gap junction communication between ECs and MSCs also plays an important role in the formation of new blood vessels. It has been shown that the direct coculture of MSCs with ECs can result in the upregulation of common pericyte markers, including CD146, neural/glial antigen 2, and alpha-smooth muscle actin (α-SMA),⁸³ suggesting that MSCs can assume a pericyte-like phenotype. Figure 6 illustrates this effect in a study in which human MSCs were directly cultured with human ECs, resulting in upregulation of SMC-specific markers (Fig. 6A) and collagen gene expression (Fig. 6B). Interestingly, when MSCs were subsequently removed from cocultures, the expression of these markers decreased over time, while markers of MSC stemness increased (Fig. 6C).⁸² It has also been shown that MSCs will migrate toward ECs in culture and will align with and stabilize nascent endothelial networks. Only upon direct contact of MSCs with ECs is an upregulation of α-SMA evident.⁹⁰ Taken together, these studies show the propensity of MSCs to differentiate toward an SMC phenotype when cultured with ECs, thereby enhancing their ability to act as pericytes and support the formation of stable vascular networks.

FIG. 6.

Direct coculture of MSCs with ECs promotes smooth muscle cell differentiation of MSCs. HMSCs cocultured with HUVECs exhibit increased (A) expression of smooth muscle cell differentiation markers and (B) collagen genes. Removal of MSCs from coculture resulted in decreased expression of these genes over time, and (C) an increase in mesenchymal stemness markers. *p < 0.05 relative to control without cocultured ECs. Figure adapted and used by permission from Lin and Lilly.⁸² HMSC, human bone marrow-derived MSC; HUVEC, human umbilical vein endothelial cell.

Coupling orthopedic and vascular tissue engineering

The coordinated development of mineralized bone and nourishing blood vessels is a key element in regenerating large volumes of bone tissue. Without a perfused blood vessel network, transplanted cells rely on diffusion for the delivery of nutrients and removal of metabolic byproducts. However, this transport mechanism is only effective over relatively short distances, typically a few hundred microns.^41,91 This diffusional constraint represents an important challenge preventing the clinical translation of bone tissue engineering strategies,^58,60 and suggests that establishment of a stable, perfusable vessel network in engineered constructs before implantation could improve their regenerative potential.

The established bidirectional and complementary interactions between ECs and MSCs in forming new bone and vascular tissues make these cells promising candidates for development of cell-based bone constructs that can support both osteogenesis and vasculogenesis. In coculture, ECs and MSCs exhibit increased osteogenic and angiogenic potential as evidenced by amplified gene and protein expression of ALP, BMP-2, VEGF, and platelet endothelial cell adhesion molecule.⁹² EC paracrine signaling likely contributes to the observed behavior through secretion of factors shown to increase MSC recruitment, proliferation, and osteogenesis, including BMP2, vasoconstrictor endothelin-1, IGF, and TGF-β.^18,93,94 A systematic review of 22 preclinical studies in which osteogenic and vasculogenic cells were cocultured and cotransplanted suggested that such systems can enhance bone regeneration in particular craniofacial applications, potentially through enhancing the regenerative phenotype of the transplanted cells.⁷

Despite promising results suggesting that EC–MSC cocultures can enhance the osteogenic and vasculogenic potential of both cell types through synergistic interactions, achieving concurrent osteogenic differentiation and stable vasculature within a single construct before in vivo transplantation has proven elusive. A main challenge is the need to tailor the culture environment to promote both osteogenesis and vasculogenesis without compromising either function. A range of studies using ECs and MSCs have examined and illuminated this issue.^95,96 In particular, it has been shown that vascular structures formed within osteogenically primed constructs are rudimentary and lack important pericyte-like associations, highlighting the challenge of promoting osteogenic lineage commitment without compromising vessel formation.⁹⁷ The development of more complex culturing platforms that better recapitulate the myriad of cues present during native bone regeneration has been explored as a means to enable the coordinated development of osseous tissue and vasculature.⁹⁸ However, the development of multiphase constructs suitable for clinical translation has yet to be achieved. The need to maintain separate functional phenotypes within a construct remains an important challenge in creating multiphase and multifunctional engineered tissues.

A main practical consideration when attempting to create engineered constructs with both osteogenic and vasculogenic components is the choice of culture media during fabrication and maintenance of the constructs. The ideal medium would maintain viability of both cell types while also promoting the desired functional outcomes. MSCs require a phosphate source and other additives to induce osteogenic differentiation and mineralization, typically ascorbic acid and dexamethasone.⁹⁹ Similarly, ECs require particular angiogenic factors (typically VEGF, IGF, FGF, and others) to promote survival and vessel assembly in vitro. Osteoinductive supplements have been shown to negatively influence EC elongation and viability in vitro,^100,101 as illustrated in Figure 7A–C. These factors are also known to compromise the function of vascular pericytes, causing them to undergo osteogenesis through a mechanism similar to aortic calcification.^102,103 It has also been shown that vasculogenic medium decreases the osteogenic potential of MSCs even when supplemented with osteoinductive factors (Fig. 7D, E),^104,105 and vasculogenic medium components have been found to directly interfere with MSC osteogenic differentiation.^106–108

FIG. 7.

Effect of culture medium on vessel development and osteogenic differentiation. EC elongation over 2 weeks in culture is clearly affected by the nutrient medium: (A) osteogenic medium, (B) vasculogenic medium, (C) combination of osteogenic and vasculogenic media. CD31 staining (green) demonstrated the presence of vessel-like structures and DAPI staining (blue) identifies all cell nuclei present in cultures. Similarly, culture medium affects the expression of osteogenic markers by MSC over 2 weeks in culture: (D) Alizarin Red, (E) ALP activity. Scale bar = 150 μm. *p < 0.05 compared with other groups. Figure adapted and used by permission from Kolbe et al.¹⁰⁴ DAPI, 4′,6-diamidino-2-phenylindole; ALP, alkaline phosphatase; ODM, osteogenic differentiation medium; EGM2, vasculogenic medium; ODM-S_EC, osteogenic medium with endothelial supplements; EGM2-βGly, vasculogenic medium supplemented with β-glycerophosphate; EGM2-S_ODM, vasculogenic medium supplemented with β-glycerophosphate, ascorbic acid, and dexamethasone. Color images are available online.

Mimicking Nature: Developmental Engineering

The challenge presented by maintaining multiple cell phenotypes within a single engineered tissue construct has led to alternate approaches to creating vascularized bone tissue. Some strategies have harnessed advances in developmental biology and regenerative medicine to inform what has been termed “developmental engineering.”^109–111 In the context of bone tissue engineering, there have been efforts to mimic the sequential induction of biochemical factors governing cell differentiation and matrix production.¹¹² In particular, there has been an emphasis on recapitulating the spatial and temporal profile of the native endochondral ossification pathway since this developmental process naturally couples bone and vasculature through a cartilaginous intermediate, as illustrated schematically in Figure 8. New tissue formation is guided by coordinated biochemical signals that result in chondrogenesis, angiogenesis, and ultimately osteogenesis. For this reason, tissue engineers have investigated ways to recreate the key steps of endochondral ossification to manipulate EC–MSC cocultures. Since the mineralization process is relatively well understood, there has been an emphasis on better understanding and controlling (1) chondrogenic priming of MSCs and (2) prevascularization using ECs and MSCs, with the overall goal of combining these processes to regenerate vascularized bone tissue.

FIG. 8.

The process of endochondral ossification. Initial MSC differentiation into chondrocytes and condensation is followed by a hypertrophic phase with vascular invasion. The intermediate cartilaginous template undergoes remodeling and mineralization, and further remodeling to produce mature bone. Figure adapted and used by permission from Freeman and McNamara.⁵⁵ Color images are available online.

Chondrogenic priming

The formation of a cartilaginous template is an important step in endochondral ossification, and therefore the field of development engineering has applied chondrogenic priming of MSCs to initiate bone regeneration. Chondrogenic differentiation is commonly induced by culture of MSCs in the presence of specific biochemical factors, either in a hydrogel scaffold or in pellet culture. Standard chondrogenic differentiation medium typically contains TGF-β, ascorbic acid, linoleic acid, dexamethasone, and insulin–transferrin–selenium.^113,114 This combination of differentiation factors has been found to effectively recreate critical elements of the endochondral pathway, as shown by the in vitro formation of a cartilage-like matrix rich in collagen type II and X.¹¹⁵ Furthermore, these cartilaginous matrices are capable of becoming mineralized through subsequent treatment of a phosphate source.^110,116 Implantation of chondrogenically primed MSCs has confirmed that this phenotype can positively influence bone formation in both ectopic^{110,117–119} and critical-sized orthotopic defect models.^120,121 It has been suggested that the formation of bone in chondrogenically primed implants proceeds through a process similar in both temporal and spatial progression to limb bone development, as demonstrated by initial formation of hypertrophic cartilage, followed by osteoclastic resorption of the cartilage template and appearance of hematopoietic tissue components.¹¹⁰

Interestingly, chondrogenic priming of MSCs can lead to enhanced mineral deposition relative to osteogenic priming,^117,122 suggesting that the endochondral-based approach can promote the osteogenic potential of MSCs without the need for exogenous osteogenic supplements. This is particularly attractive for vascular bone tissue engineering strategies since osteogenic supplements may compromise the development of stable vasculature. A number of studies have shown that a chondrogenic priming strategy can result in improved bone formation, relative to osteogenically primed constructs.^117,123,124 Figure 9 shows an example in which MSC constructs were primed to follow either the intramembranous (osteogenic) or endochondral (chondrogenic) pathway to bone formation.¹²² Chondrogenically primed constructs revealed more advanced bone formation as shown by enhanced characteristics of cortical bone with embedded osteocytes and lamellar-like structures. Furthermore, the highest degree of bone formation was obtained with chondrogenic-primed MSCs embedded within a chondrogenic matrix suggesting that both endochondral-based priming and scaffold composition play significant roles in the healing response. The mechanisms by which chondrogenic priming leads to increased bone formation by MSCs are not fully understood. The effect may be related to the availability of the cartilaginous matrix to be mineralized, and there is also evidence that chondrogenically primed MSCs can further differentiate into osteoblasts to form new bone tissue.^37–40

FIG. 9.

Chondrogenic and osteogenic priming of engineered bone constructs. MSC seeded on collagen/hydroxyapatite scaffolds were primed for 5 weeks in osteogenic or chondrogenic medium before implantation in a rat critical-sized calvarial bone defect model. Chondrogenically primed constructs showed more mature bone formation [(A–C), areas of newly formed bone are highlighted by blue arrows], and enhanced blood vessel formation (D). *p < 0.05 statistically significant difference between control and other groups at 8 weeks, **p < 0.05 statistically significant difference between 4 and 8 weeks of chondrogenically primed group. Figure adapted and used by permission from Thompson et al.¹²⁴ Color images are available online.

Chondrogenic priming has been shown to influence both the chondrogenic and osteogenic potential of MSCs in vitro and in vivo. However, the degree to which MSCs need to be chondrogenically induced to achieve maximal bone formation has not been well defined. It has been demonstrated that chondrogenic preculture time does affect the capacity of MSCs to adopt an osteogenic phenotype, with longer periods (14–21 days) resulting in enhanced mineral deposition.¹²² Longer preculture times are more likely to result in the hypertrophic chondrocyte phenotype, which is associated with endochondral ossification and forms the template for later vascularization and mineralization.^110,116 Priming MSCs to hypertrophy occurs either through long-term chondrogenic culture, or through controlled presentation of specific biochemical factors, for example TGF-β withdrawal, dexamethasone reduction, addition of thyroid hormone,¹²⁵ or addition of IL-1β.¹¹⁸ Induction of hypertrophy in MSCs using these methods has been found to produce mature, ossified tissue with evidence of host vasculature infiltration after 4 weeks of implantation in a subcutaneous model.¹¹⁰

Collectively, these studies support the use of chondrogenic priming of MSCs as a means to stimulate bone formation through the endochondral pathway. The priming duration and resulting MSC phenotype at implantation appears to influence the in vivo bone regeneration potential. However, the optimal in vitro culture time and associated degree of chondrogenic differentiation is still a matter in need of further investigation. The fate of chondrogenically primed MSCs must also be better understood in this tissue engineering context, as it remains unclear whether these cells serve as inductive mediators for host progenitors or if they themselves undergo osteogenic differentiation. In addition, it has been observed that chondrogenically primed tissue constructs can result in an uneven distribution of bone mineral¹¹⁰ and eventual core degradation in vivo.¹¹⁷ These effects may be the consequence of insufficient vascularization, since vessel infiltration is required to induce mineralization and promote bone remodeling.

Prevascularization

The transition of the initial hypertrophic cartilaginous template to a vascularized tissue is a key stage in the process of endochondral ossification. A stable and perfused vascular network is needed to provide nutrients and cellular components as the tissue mineralizes to form bone. Prevascularization strategies have been used to address the uneven mineral distribution and core degradation observed in chondrogenically primed constructs. Figure 10 illustrates this approach in a study where ECs and MSCs were seeded onto chondrogenically primed MSC aggregates.¹²⁶ Cells attached, proliferated, and infiltrated into primed aggregates, and over 3 weeks resulted in the formation of a rudimentary vasculature that was integrated within the aggregates. It was also demonstrated that coupling chondrogenic priming with prevascularization in vitro enhanced the osteogenic potential of MSCs compared with chondrogenic priming alone, even in the absence of osteogenic supplements.¹²⁶ Other studies have confirmed that chondrogenically primed MSCs produce factors that are important for the induction and infiltration of vasculature, including VEGF and MMPs, suggesting that the endochondral approach may intrinsically promote vessel formation within cartilaginous constructs.¹¹⁵

FIG. 10.

Chondrogenic priming, media, and coculture of MSCs and ECs influence osteogenic and vasculogenic potential of engineered constructs. MSC pellets were primed in chondrogenic medium (CP) for 3 weeks and were then continued as monocultures (−HUVECs), or were seeded with ECs (+HUVECs) or both ECs and MSCs (+HUVECs:MSCs) before being cultured in vasculogenic medium alone (−OM) or in combination with osteogenic supplements (+OM) for an additional 3 weeks. (A) CD31 staining (green) demonstrated the presence of integrated blood vessels in the group seeded with both ECs and MSCs, but not in the other groups. All cell nuclei were counterstained with either DAPI (blue) or Propidium Iodide (red). White boxes denote examples of the rudimentary vessels present within the aggregates. Quantification of osteogenic markers reveals enhanced (B) ALP activity and (C) calcium content in chondrogenic cultures that were seeded with HUVECs and MSCs. ^p < 0.05 versus CP21 + HUVECs group, ^ap < 0.05 versus CP21 − HUVECs, ^bp < 0.05 versus Osteo alone. Figure adapted and used with permission from Freeman et al.¹²⁶ CP, chondrogenically primed; −OM, no osteogenic supplements. Color images are available online.

Implantation of prevascularized, chondrogenically primed MSC aggregates in an ectopic subcutaneous site showed that after 4 weeks, constructs contained mature blood vessels within the center of the aggregates, as indicated by α-SMA-positive cells in the vessel walls and the presence of erythrocytes in the lumens.¹²⁷ Interestingly, vessels formed in engineered constructs in vitro are typically immature and lack pericyte support. However, in this study, mature vessels were evident in vivo, suggesting that the initial rudimentary vessel network may be augmented and remodeled upon implantation and interaction with the host. Prevascularization has also been shown to enhance the formation of mineralized nodules within the constructs, highlighting the potential of combined chondrogenic and vascular priming for the potentiation of bone regeneration that has also been observed in other studies.^126,128 The ability to create engineered constructs that concurrently exhibit both osteoprogenitor and pericyte functions remains elusive. However, the approach of harnessing developmental pathways to create engineered bone tissues suggests that mature cell phenotypes and tissue structures are not necessary to achieve bone regeneration in vivo. Taken together, these studies support the beneficial role of mimicking the endochondral ossification pathway by combining chondrogenic priming and prevascularization processes to generate a vascularized bone construct.

Summary and Perspectives

Cell-based tissue engineering strategies have compelling potential to address the limitations of current bone grafting options, particularly in large and complex defects. Over the past three decades, the field of bone tissue engineering has made important progress in developing new materials and approaches to enable cell-based therapies in orthopedics. However, achieving adequate vascularization of engineered constructs, both before and after implantation, has remained a significant roadblock that has prevented the broader success of these approaches.⁵⁴ This important problem has motivated the development of a variety of vascularization strategies, many of which harness ECs and the multifaceted capabilities of MSCs to regenerate tissue.

The natural process of bone formation provides insight into how multiphase, vascularized orthopedic tissues can be developed. The processes of osteogenesis and vasculogenesis are clearly coupled in vivo, through the action and interaction of progenitor cells. However, recapitulating and controlling these interactions in vitro to create multifunctional engineered tissues has proven very challenging. Cocultures of ECs and MSCs have been applied widely for this purpose, since these cell types can act synergistically to regenerate a variety of tissues. In particular, ECs are known to enhance the osteogenic function of MSCs to form bone, while MSCs can also support blood vessel formation by acting as supportive pericytes that modulate EC function. The challenge of creating single constructs using both cell types is the requirement of different environmental cues to carry out their differentiated function. Therefore, creation and maintenance of constructs that combine osteogenesis and vasculogenesis in vitro will require a better understanding of the culture conditions and supplements that are needed to support these dual functions.

The concept of developmental engineering has been applied to bone in an effort to mimic key stages in orthopedic tissue regeneration. The process of endochondral ossification is of particular interest since it couples both bone and blood vessel formation through a series of spatially and temporally regulated events. The work to date in this area has focused on two main elements of this process: (1) chondrogenic priming of MSCs to generate a hypertrophic cartilage template, and (2) vascularization of this tissue template through the coordinated action of ECs and MSCs. An important feature of this approach is that experimental evidence suggests that appropriately primed and cultured constructs can undergo ossification even in the absence of exogenous osteogenic cues. Similarly, there is evidence that appropriate priming can result in the secretion of vasculogenic factors that promote cellular infiltration and the formation of a vascular network, and that vascularization of the cartilaginous template further enhances matrix mineralization and osteogenic differentiation of MSCs. The control of the degree and timing of chondrogenic priming and subsequent vascularization allows multiple cell phenotypes to coexist in a single construct, and further enables a synergistic relationship between these phases. Therefore, temporally defined induction of these two key stages of the endochondral ossification pathway is a promising means to couple osteogenesis and blood vessel formation in the creation of vascularized bone constructs.

The coupling of osteogenesis and vasculogenesis in engineered tissue constructs represents both a challenge in the field of bone tissue engineering, as well as an opportunity to create more complex and functionally complete tissue replacements. An impressive amount of progress has been made in understanding how these processes are linked and how the key cellular components interact. At the same time, our knowledge about the timing, spatial distribution, and cell phenotypes that are most conducive to promoting concurrent bone and vessel formation is not complete. There is a need for advanced cell and tissue culture systems and characterization techniques to better understand the separate and interactive roles of the relevant cell types. Bioreactor systems and media formulations that can support the development of multiphase tissues are also needed, both for scientific study as well as to generate engineered tissues for therapeutic use. In addition, there is a need for improved animal models that truly mimic the large, complex, and recalcitrant bone defects that are the main targets of cell-based orthobiologic tissue engineering approaches. Taken together, these areas of study provide a rich opportunity to better understand how osteogenesis and vasculogenesis are coupled, and to apply this knowledge to creating vascularized orthopedic tissues.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Some of the figures in this article were created using Servier Medical Art templates, which are licensed under a Creative Commons Attribution 3.0 Unported License (smart.servier.com).

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported in part by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01-AR062636, to J.P.S.) and the National Institute of Dental and Craniofacial Research (R01-DE026630, to J.P.S.). N.G.S. is partially supported by the Cellular Biotechnology Training Program (T32-GM008353) at the University of Michigan. N.E.F. is partially supported by Tissue Engineering and Regeneration Training Grant (T32-DE007057) at the University of Michigan.