Extracellular matrix bioscaffolds in tissue remodeling and morphogenesis

Abstract

During normal morphogenesis the extracellular matrix (ECM) influences cell motility, proliferation, apoptosis, and differentiation. Tissue engineers have attempted to harness the cell signaling potential of ECM to promote the functional reconstruction, if not regeneration, of injured or missing adult tissues that otherwise heal by the formation of scar tissue. ECM bioscaffolds, derived from decellularized tissues, have been used to promote the formation of site appropriate, functional tissues in many clinical applications including skeletal muscle, fibrocartilage, lower urinary tract, and esophageal reconstruction, among others. These scaffolds function by the release or exposure of growth factors and cryptic peptides, modulation of the immune response, and recruitment of progenitor cells. Herein, we describe this process of ECM induced constructive remodeling and examine similarities to normal tissue morphogenesis.

Introduction

The ECM represents the secreted composite accumulation of structural and functional molecules that are sufficient and necessary for the organization, viability, and phenotypic plasticity of the resident cell population of every tissue and organ. In fact, the ECM is both responsible for and responsive to the continuously changing structural, mechanical, metabolic, and functional needs of these cells; a process often referred to as dynamic reciprocity (Bissell and Aggeler, 1987). These requirements and activities of the ECM are present during tissue development, growth, homeostasis, and in response to changing mechanical and physiologic states, and importantly, in response to injury.

The ECM can be isolated from tissues and organs by processes that remove the resident cell population (i.e. decellularization). The methods used to disrupt the cells and remove the resultant cellular debris determine the amount of damage to the ultrastructure, composition, and associated ligand landscape of the ECM. Such bioscaffolds have been used as interventional implants to provide both a macro-and microenvironmental niche for mechanical support and to promote a constructive and functional tissue repair. Many of the fundamental processes that occur during the ECM-mediated tissue repair process are similar or identical to those which occur during normal tissue development and growth, but would otherwise be absent or at least inhibited during the default repair processes which occur following tissue injury in postnatal mammals.

The matrix is composed of extracellular collagens, glycoproteins, glycosaminoglycans, proteoglycans, adhesion molecules, growth factors, chemokines and cytokines. Each of these components have important functions during embryonic development. Cell-ECM interactions are necessary for many developmental processes such as cell migration, branching morphogenesis, and cell fate specification. Proof of the critical role of ECM proteins during development is the fact that loss-of-function mutation of many ECM molecules such as fibronectin, laminin, or collagen are embryonic lethal (reviewed by (Rozario and DeSimone, 2010)). Fibronectin is just one example of an extracellular matrix protein that is necessary for development and morphogenesis of many tissues and organs throughout embryogenesis. During cardiac development, precursor cells require fibronectin to complete their migration and in the absence of this matrix molecule epithelial organization is disrupted (Matsui et al., 2007; Trinh and Stainier, 2004). In the mammary gland, lung, kidney, and salivary gland fibronectin deposition and remodeling is critical for branching morphogenesis (Harunaga et al., 2011; Liu et al., 2010; Mettouchi, 2012; Roman, 1997). Fibronectin also plays a role in directing cellular differentiation (Taylor-Weiner et al., 2013).

ECM components that are critical during development also play a key role in regeneration, as observed during the regeneration of amphibian limbs. In regenerating limbs, progenitor cells are recruited into a blastema that can completely and faithfully replace lost tissue (Simon and Tanaka, 2013). Components of the dynamic blastemal ECM, such as hyaluronic acid, tenascin C, and fibronectin, differentially affect cell behaviors such as DNA synthesis, cell migration, and myoblast fusion (Calve et al., 2010). The fact that amphibian and mammalian cells exhibit similar responses to ECM molecules suggests that the ability to sense and respond to regenerative signals is evolutionarily conserved.

The field of tissue engineering and regenerative medicine has attempted to leverage the morphogenetic properties of ECM for the repair of tissue injuries. Biologic scaffolds composed of ECM are created by removing the cellular components of tissues or organs using physical, chemical and enzymatic methods, leaving behind an intact meshwork of ECM components with tissue specific architecture (Crapo et al., 2011; Keane et al., 2015). These scaffolds provide both a structural framework and biochemical signaling platform for tissue reconstruction following injury. ECM scaffolds have been used successfully in pre-clinical animal models and in human clinical applications where they have been shown to act as an inductive template for the formation of new tissues (a partial list of commercially available ECM bioscaffolds is shown in Table 1). The generation of site-appropriate, functional tissue following injury has been termed “constructive remodeling” (Badylak et al., 2011). Herein, we highlight several examples of ECM scaffold induced constructive remodeling and discuss the potential mechanisms by which the ECM may elicit this response. A better understanding of the complex bidirectional communication between cells and ECM during development and regeneration will lead to improved tissue engineering and regenerative medicine approaches for restoration of functional tissues.

Table 1.

Partial list of commercially available biologic scaffolds composed of extracellular matrix.

Product	Company	Material
AlloDerm	Lifecell	Human dermis
CuffPatch™	Biomed Sports Medicine	Porcine small intestinal submucosa (SIS)
Oasis®	Cook Healthpoint	Porcine small intestinal submucosa (SIS)
Restore™	DePuy	Porcine small intestinal submucosa (SIS)
Surgisis®	Cook Medical	Porcine small intestinal submucosa (SIS)
MatriStem®	ACell Inc.	Porcine urinary bladder matrix (UBM)
CollaMend FM Implant	C.R.Bard	Porcine dermis
Pelvicol®	C.R.Bard	Porcine dermis
Permacol™	Covidien	Porcine dermis
XenMatrix Surgical Graft	C.R.Bard	Porcine dermis
Simmer Collagen Patch®	Tissue Science Laboratories	Porcine dermis
Dura-Gaurd®	Synovis Surgical	Bovine pericardium
Peri-Guard®	Synovis Surgical	Bovine pericardium
Vascu-Guard®	Synovis Surgical	Bovine pericardium
Veritas®	Synovis Surgical	Bovine pericardium
PriMatrix™	TEI Biosciences	Fetal bovine dermis
SurgiMend™	TEI Biosciences	Fetal bovine dermis
TissueMend®	TEI Biosciences	Fetal bovine dermis
OrthADAPT™	Synovis Life Technologies	Equine pericardium

Promotion of constructive remodeling by extracellular matrix

Skeletal Muscle

Skeletal muscle has a remarkable ability to regenerate and remodel in response to injury and stress, partially recapitulating the embryonic developmental program. Muscle regeneration occurs in five interrelated phases: degeneration (necrosis), inflammation, regeneration, remodeling, and maturation/functional repair (Musarò, 2014). Deposition and remodeling of ECM is essential for muscle regeneration. During the stage of regeneration, infiltrating inflammatory cells and activated resident cells such as fibroblasts and myoblasts deposit a new ECM that provides support for further proliferation and migration of these cells (Grounds, 2008; Serrano and Muñoz-Cánoves, 2010). Matrix remodeling by metalloproteases is also required for myoblast migration and maturation (Casar et al., 2004; Snyman and Niesler, 2015). However, trauma that results in the loss of a large volume (20% or greater) of skeletal muscle exceeds the endogenous ability to regenerate functional tissue. The default healing response results in the deposition of scar tissue. Current techniques for the repair of these defects include autologous tissue transfer of vascularized or free muscle flaps. These methods provide some cosmetic improvement and coverage; however, there is little restoration of muscle function. Therefore, clinicians and researchers have focused upon tissue engineering and regenerative medicine strategies to address this unmet clinical need.

Acellular bioscaffolds composed of ECM have been successfully utilized for the repair of large muscle defects in several pre-clinical models of VML (Chen and Walters, 2013; Merritt et al., 2010; Sicari et al., 2012; Turner et al., 2010; Valentin et al., 2010). The newly formed skeletal muscle was vascularized, innervated, and contractile. The ECM bioscaffolds were derived from homologous skeletal muscle tissue sources (Chen and Walters, 2013; Merritt et al., 2010) as well as heterologous tissue sources such as small intestinal submucosa (Sicari et al., 2012; Valentin et al., 2010). In a rat latissimus dorsi defect model where skeletal muscle derived ECM was used for repair, Chen et al. showed that muscle ECM was well-integrated with the host tissue by 8 weeks with regeneration of the microvasculature and formation of well-formed new muscle fibers across the defect region. Another study examined scaffolds composed of porcine small intestinal submucosa ECM implanted into a canine model of musculotendinous junction injury at 2, 4, and 6 months following injury (Turner et al., 2010). By 2 months, newly synthesized collagen was present along with multipotential stem cells, myoblasts, and disorganized myofibers. By 4 months mature muscle fiber bundles were present within the implant with disorganized orientation. By 6 months, newly formed muscle was similar to native, uninjured tissue. It should be noted however, that the amount of new muscle formed is typically 15–40% of the injured or missing muscle with the remaining tissue consisting of organized dense collagen. This study showed that ECM bioscaffolds can at least partially circumvent the default response to tissue injury and promote the formation of vascularized, functionally innervated skeletal muscle by 6 months post-implantation (Turner et al., 2010).

While the mechanism by which ECM promotes new muscle formation is not fully understood, several cell types have been proposed to contribute to this process. Chen et al. observed MyoD expressing cells throughout the ECM scaffold at 8 weeks, suggesting the presence of muscle progenitor cells. However, because the final time point in the study was 8 weeks, it was not determined whether the progenitor cells present could regenerate muscle if given a longer period of time. Turner et al. showed an association of CD133+ cells with new myofibers at 2 and 4 months that could play an important role in the formation of new muscle. Traditionally considered a marker for circulating progenitor and hematopoietic stem cells (Hristov and Weber, 2004; Miraglia et al., 1997), CD133 has more recently been described as a marker of a myogenic cell population capable of regenerating skeletal muscle in patients with Duchene muscular dystrophy (Torrente et al., 2004). In a murine model of VML, repair with urinary bladder ECM was associated with a population of perivascular stem cells (CD146+/Ng2+) that were located outside of their normal perivascular niche and scattered throughout the ECM scaffold material. The presence of these progenitor cell populations suggests a potential mechanism for new muscle formation by which these cells are recruited to the site of injury. Further research is necessary determine the contribution of these cell types to constructive remodeling and the role of ECM in cellular recruitment and/or differentiation.

Tendon

Tendon injuries are a common clinical problem with an estimated 300,000 tendon and ligament repair surgeries performed annually in the U.S. (Pennisi, 2002). Surgical treatments to repair or replace damaged tendon utilize autografts, allografts, or xenografts (Yang et al., 2013). However, the outcomes of surgical repair remain unsatisfactory due to high failure rates, donor site morbidity, risk of injury recurrence and limited long-term functional recovery. Despite surgical intervention, the natural healing process of tendon is slow due to their hypocellular and hypovascular nature (Yang et al., 2013). These limitations have served as motivation for the development of new tissue engineering strategies to create functional replacements or boost the innate healing of tendon defects.

ECM scaffolds have been used for the repair of musculotendinous tissue including the Achilles tendon and rotator cuff (Metcalf et al., 2002; Sclamberg et al., 2004). Several pre-clinical animal studies have demonstrated that ECM derived from porcine small intestinal submucosa can successfully remodel into tendon or ligament tissues of the rotator cuff (Dejardin et al., 2001), medial collateral ligament (Musahl et al., 2004), Achilles tendon (Badylak et al., 1995; Gilbert et al., 2007b; Hodde, 2006), and spinal ligaments (Ledet et al., 2002). These studies have demonstrated that the small intestinal submucosa ECM is degraded rapidly after implantation and replaced by of host tissue that is histologically similar to native tissue. In a canine model of rotator cuff repair, regular orientation of collagen fibers and minimal vasculature was observed at 6 months following treatment and were comparable to sham-operated and native tendons (Dejardin et al., 2001). Additionally, the failure mode of SIS-regenerated tendons mimicked those of sham-operated tendons, suggesting host tissue ingrowth and implant remodeling with solid integration of the re-generated tissue to muscular and bony interfaces. In the Achilles tendon application, degradation of the device was observed histologically within the first 8 weeks (Badylak et al., 1995). ECM scaffold degradation was quantitatively evaluated in an Achilles tendon model using small intestinal submucosa ECM labeled with ¹⁴C. Approximately 10% of the scaffold material was degraded and removed from the site of implantation by as early as 3 days after surgery. By 14 days, approximately 20% of the scaffold was degraded. The amount of ¹⁴C in the remodeled tissue was equal to background levels by 60 days, indicating complete scaffold degradation (Gilbert et al., 2007a). Immobilization following surgery slows the healing of SIS treated tendons and limits cellular infiltration and neovascularization of the ECM graft (Hodde et al., 2007); stated differently, mechanical forces affect remodeling just as such forces ae necessary for normal development.

In addition to the use of small intestinal submucosa ECM, decellularization protocols have been optimized to generate acellular tendon scaffolds i.e., tendon ECM (Pridgen et al., 2011). These scaffolds showed similar biomechanical properties compared to native tendon tissue and also demonstrated a crimp pattern characteristic of native tendon and retain ECM proteins and growth factors, suggesting potential bio-functionality (Pridgen et al., 2011). Decellularized tendon scaffolds retain the native tendon ECM microenvironment cues, including surface topography, biochemical composition, and stiffness that allow them to support the proliferation and tenogenic differentiation of stem cells (Ning et al., 2015). The tenogenic inductive properties of decellularized tendon further support the use of ECM as a promising material for tendon repair/reconstruction.

Together, these data show that ECM scaffolds are capable of remodeling into tendon or ligament tissue with histologic and mechanical properties similar to that of native tissue. However, the biologic mechanisms through which this occurs have yet to be elucidated. Recently, specific biological criteria have been proposed for the functional tissue engineering of tendon (Breidenbach et al., 2014). These criteria, which are believed to be essential for normal tendon function are: (1) scleraxis-expressing cells, (2) well-organized and axially aligned collagen fibrils having a bimodal distribution, and (3) a specialized tendon-to-bone insertion site. Examination of these biologic criteria during the course of ECM mediated tendon repair could provide valuable insight into the tendon remodeling process.

Fibrocartilage

Fibrocartilage is a fibrous connective tissue that is found in menisci, intervertebral discs, tendon and ligament entheses, and the temporomandibular joint (TMJ). These tissues have very poor natural healing outcomes and are a focus of regenerative medicine approaches.

TMJ disorders affect an estimated 10–36 million Americans and encompass a wide spectrum of clinical conditions involving components of the TMJ (Farrar and McCarty, 1979; Oakley and Vieira, 2008; Rollman and Gillespie, 2000; Tanaka et al., 2008). For many of these patients, removal of the TMJ disc is the only treatment that will relieve pain and restore motion to the jaw (Hall et al., 2005; Krug et al., 2004; Nyberg et al., 2004; Vázquez-Delgado et al., 2004). However, while removal of the TMJ disc results in a reduction of pain, subsequent degenerative changes in the articular cartilage occur. (Krug et al., 2004; McKenna, 2001; Nyberg et al., 2004). Only short-term success has been reported for TMJ replacement with alloplastic materials of autogenous tissues and replacement of the disk has been associated with negative outcomes such as donor site morbidity, scar tissue formation, decreased range of motion and additional joint pathology (Alonso et al., 2009; Dimitroulis, 2005; Dolwick and Aufdemorte, 1985; Ferreira et al., 2008; Matukas and Lachner, 1990). Extracellular matrix may be an ideal biomaterial to create an off-the-shelf disc replacement that could integrate with the surrounding host tissue, act as a template for cellular growth, and eventually restore function and native morphology of the TMJ disc.

A recent study demonstrated that a device composed of decellularized porcine urinary bladder matrix functioned as an effective interpositional material and an inductive template for reconstruction of the TMJ disc in vivo (Brown et al., 2011). In that study, a device consisting of a powdered UBM pillow encapsulated within sheets of the same material was placed as an interpositional graft after discectomy in a canine model. The implanted device was observed to progressively remodel from 3 weeks to 6 months after implantation. Gross and histologic examination of the newly formed host tissues revealed that the remodeled tissue resembled the native fibrocartilage of the TMJ disc. A follow-up study demonstrated that the composition and mechanical properties of the remodeled tissue were also similar to that of the native disc (Brown et al., 2012a). Interestingly, the placement of the UBM device resulted in formation not only of fibrocartilage within the bulk of the implant, but also muscular and ligamentous attachments resembling those found at the periphery of native menisci.

Digit

The potential for limb regeneration in vertebrates has excited scientists for many years. Non-mammalian species such as newts and axolotls are capable of regeneration of complex tissues such as limbs and digits through a process known as blastemal-based epimorphic regeneration, in which genetic programs are activated and soluble factors recruit a population of multipotent stem cells preprogrammed to recapitulate a perfect phenocopy of the missing tissue (Kragl et al., 2009; Kumar et al., 2007; Lévesque et al., 2007; Monaghan et al., 2009; Neufeld and Zhao, 1995). Regrowth of amputated limbs of salamanders and axolotls is unique for several reasons; multiple tissue types are able to regenerate, regeneration occurs in the correct orientation, and there is ‘positional memory’ whereas regenerated structures recapitulate those that were lost in amputation.

Digital tip regeneration models have also been described in other mammalian species such as mice (Fernando et al., 2011), rats (Said et al., 2004), and monkeys (Singer et al., 1987). The regenerative power of mice is limited only to amputations distal to the distal interphalangeal joints (Neufeld and Zhao, 1995). The regenerated digit is grossly similar to the original, but is not a perfect replacement (Fernando et al., 2011). A blastema does not form at the site of injury in adult mammals; instead regeneration occurs through a wound-healing phase that is associated with an osteoclastic response from the stump bone. This process supplies stem cells from the bone marrow to form a blastoma-like structure, which will then enter into a regeneration phase and new bone is formed via intra-membranous ossification replacing the missing tissue.

One study examined the administration of chemotactic ECM degradation products in a mouse model of digit amputation and investigated the effects on recruitment of multipotential cells to the site of injury (Agrawal et al., 2010; Agrawal et al., 2012). The results showed that ECM degradation products promoted the accumulation of a heterogeneous cell population at the amputation site 14 days post-injury. These cells expressed markers of multipotency including Sox2, Sca1, and Rex1 (Zfp42). Cells isolated from the site of amputation were capable of differentiation along multiple lineages, whereas cells isolated from control mice (not treated with ECM) were not capable of such differentiation. A proteomics approach to identifying the peptides responsible showed that a single subunit of Collagen III could promote the chemotaxis of multiple progenitor cell types in vitro and was able to recruit Sox2, Sca1 positive cells when injected in vivo (Agrawal et al., 2011). Further research is necessary to determine additional factors required for appropriate differentiation of multipotential cells, however, these results suggest that bioscaffolds composed of ECM can provide an inductive microenvironment which promotes site appropriate tissue formation in adult mammals.

Mechanisms of extracellular matrix influence on remodeling and morphogenesis

Modulation of the immune response

Due to their unique composition and surface topologies, naturally occurring extracellular matrix materials elicit a distinctly different surface response than those composed of synthetic material. Tissue remodeling following implantation of these materials is associated with a robust macrophage response beginning as early as two days post-implantation and continuing for several months, depending on the material and clinical application (Brown et al., 2009). An intense and long-term macrophage response to a biomaterial is typically associated with negative outcomes including chronic inflammation and encapsulation or scar formation (Anderson, 1988). However, the presence of macrophages is indispensable for the promotion of a constructive remodeling response (Badylak and Gilbert, 2008; Badylak et al., 2008; Brown et al., 2012b; Brown et al., 2009; Valentin et al., 2009). Ablation of circulating phagocytes with clodronate liposomes prevents the normal degradation and remodeling of ECM scaffolds, suggesting an important role for circulating mononuclear cells in the constructive remodeling process (Valentin et al., 2009).

The participation of macrophages is also necessary in regenerating species that can re-grow complete body structures as adults. Systemic macrophage depletion in the axolotl during the first 24 hours after limb amputation completely blocks limb regeneration (Godwin et al., 2013). Although the wound is able to close, extensive fibrosis and disregulation of ECM component gene expression are observed. In the absence of macrophages, markers of dedifferentiation were disregulated in the amputated limb, consistent with disruption of the blastema. This finding suggests that one mechanism by which macrophages permit regeneration in the axolotl is by promoting/allowing dedifferentiation and the formation of a progenitor cell pool. These results also suggest that macrophage-derived therapeutic molecules or modulation of the macrophage response could promote a regeneration permissive environment.

Further investigation of ECM induced constructive remodeling has shown that ECM scaffolds both promote the host innate immune response and modulate the phenotype of the cells involved in remodeling (Allman et al., 2001; Allman et al., 2002; Badylak and Gilbert, 2008; Badylak et al., 2008; Brown et al., 2012b; Brown et al., 2009; Palmer et al., 2002). Scaffold materials composed of ECM promote a switch from a predominantly M1-like macrophage (pro-inflammatory, cytotoxic) population immediately post implantation to a population enriched in M2-like macrophages (anti-inflammatory and prohealing) by 1 to 2 weeks following implantation. The mechanisms by which ECM-based scaffold materials modulate the balance between proinflammatory and anti-inflammatory phenotypes remain unknown. However, the downstream outcome associated with ECM scaffold intervention can be predicted based on the phenotypic profile of the immune cells that respond to these scaffolds at early time points (Brown et al., 2012b). Modification of scaffold materials with chemical cross-linking agents that delay or prevent macrophage-mediated degradation inhibits the favorable M2-like response, promotes the M1-like response, and results in downstream scar tissue formation (Badylak et al., 2008; Brown et al., 2012b; Cavallo et al., 2015). In fact, in silico analysis of in vitro macrophage protein secretion may be useful in predicting the host response to implanted biomaterials (Wolf et al., 2014). As degradation of ECM devices is critical for modulation of the macrophage phenotype, these results suggest that ECM degradation products may be, at least in part, responsible for constructive tissue remodeling.

Degradation products and cryptic peptides

The ECM during development, regeneration, and repair is a highly dynamic environment. The degradation and remodeling of ECM is necessary for normal tissue homeostasis, response to injury, and developmental processes. Cleavage of ECM components and creation of cryptic peptides during ECM remodeling is necessary for regulating ECM abundance, composition and structure, as well as for releasing embedded bioactive molecules such as growth factors. Degradation of implanted ECM biomaterials is necessary for constructive tissue remodeling, modulation of the immune response, and recruitment of progenitor cells. Studies using ¹⁴C-labeled ECM scaffolds have demonstrated that most of these scaffolds are degraded rapidly in vivo (Carey et al., 2014; Gilbert et al., 2007a; Gilbert et al., 2007b; Record et al., 2001; Valentin et al., 2009). For example, in a canine model of Achilles tendon repair, a 10 layer ¹⁴C-labeled scaffold was 60% degraded by 1 month post implantation and 100% degraded by 3 months following surgery (Gilbert et al., 2007b). The dermal ECM scaffolds degrade more slowly, likely as a consequence of the dense organization of collagen. Histologic analysis of these labeled devices shows that as the scaffold degrades it is populated by host cells and site-specific, functional host tissue is formed. The mechanisms of degradation of ECM scaffolds include both cellular and enzymatic pathways. Inflammatory cells such as macrophages produce oxidants as well as proteolytic enzymes that aid in matrix degradation. Proteinases within the MMP (matrix metalloproteinase) and Adamalysin (ADAM, a disintegrin and metalloproteinase and ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs) families target a wide variety of ECM components (Birkedal-Hansen et al., 1993; Page-McCaw et al., 2007; Tang, 2001). So far, 23 MMPs, 22 ADAMs, and 19 ADAMTS have been identified in humans. The mechanisms by which these proteinases modify ECM have been reviewed elsewhere (Bonnans et al., 2014; Lu et al., 2011).

As host cells degrade the ECM, peptide fragments are generated and new recognition sites with potent bioactivity are exposed. A wide variety of cryptic peptides, termed matricryptins or matrikines, have been identified within the ECM that have been shown to elicit biologic responses that are distinct from those of the full length ECM component (Table 2). These peptides regulate a variety of processes including angiogenesis, anti-angiogenesis, migration, adhesion, differentiation, and antimicrobial activity (Davis, 2010; Davis et al., 2000; Maquart et al., 2005). Exposure of cryptic peptides can also play a role in ECM assembly and influence ECM multimerization and the formation of ECM-growth factor complexes (Davis et al., 2000). For example, the degradation of fibronectin leads to the formation of peptides that can affect multimerization, interactions with other ECM components and growth factors including VEGF (Hocking and Kowalski, 2002; Vakonakis et al., 2007). The Arg-Gly-Asp (RGD) peptide present within fibronectin, collagen, vitronectin, and osteopontin is one of the most well-known examples of a matricryptic peptide (Ruoslahti, 1996). The RGD sequence is the cell attachment site of a large number of ECM and cell surface proteins, including nearly half of the over 20 known integrins (Ruoslahti, 1996). Tissue engineers have also taken advantage of the RGD peptide to design polymers with enhanced cell adhesion properties (Furth et al., 2007). In addition to the formation of cryptic peptides, ECM degradation can also result in the release of growth factors that have been sequestered within the matrix. The presence of several growth factors has been observed in decellularized matrix including VEGF, bFGF, and TGFβ (Table 2) (Reing et al., 2010). These growth factors may promote angiogenesis and other cellular responses in the local tissue environment as the ECM scaffold is naturally degraded by endogenous proteinases.

Table 2.

List of selected signaling molecules within the ECM.

Signaling Molecule	Activity
Cryptic Peptides
Endostatin	Inhibits angiogenesis
Angiostatin	Inhibits angiogenesis
Canstatin	Apoptosis, inhibits chemotaxis and proliferation
Restin	Inhibits migration
Tumstatin	Inhibits angiogenesis, promotes apoptosis
RGD (Arg-Gly-Asp)	Promotes adhesion
Hyaluronic Acid Fragments	Promotes angiogenesis, Promotes MMP production
Growth Factors
VEGF	Promotes angiogenesis
bFGF	Promotes angiogenesis
TGFβ	Regulates cell growth and differentiation
NGF	Regulates neuron growth and proliferation

Cellular recruitment

ECM molecules can orchestrate the migration of cells during development, regeneration, and wound healing. For example, during amphibian limb regeneration, multipotent progenitor cells are recruited from nearby tissues to form a blastema that rebuilds nerve, muscle, tendon, and bone at the site of injury (Hay and Fischman, 1961; Namenwirth, 1974). In vitro studies have shown that different ECM components such as hyaluronic acid, tenascin, and laminin each have unique effects on the migratory behaviors of muscle progenitor cells (Calve et al., 2010). Reing et al. examined the effects of UBM digestion products on progenitor cells and differentiated endothelial cells and showed distinct differences in progenitor cell versus differentiated cell response (Reing et al., 2009). In these experiments, enzymatically digested UBM was used to represent the peptide products that would be generated by in vivo degradation of ECM. Human multilineage progenitor cells, MRL day 11 blastema cells, and human bladder and aortic endotheilial cells were exposed to the ECM degradation products. They showed that digested UBM was chemotactic for the progenitor cell populations that were tested, and inhibitory for endothelial cell migration. UBM degradation products increased proliferation of the MRL blastema cells and inhibited endothelial cell proliferation in vitro.

The ability to recruit and differentiate stem and/or progenitor cell populations is considered a key aspect of regenerative medicine strategies. In some applications, cells are delivered within a scaffold-based material into an injury site; in other applications stem or progenitor cells are recruited from endogenous sources through a number of mechanisms including the release of growth factors and cryptic peptides. One of the constructive biologic effects of ECM scaffold degradation is the recruitment of host cells to the site of degradation. In vivo studies have shown multiple progenitor cells types are recruited to the site of bioscaffold implantation, including MyoD+ muscle progenitors (Chen and Walters, 2013), CD133+ progenitor cells (Turner et al., 2010), CD146+/Ng2+ perivascular stem cells (Sicari et al., 2012), and Sox2+/Sca1+ multipotent progenitor cells (Agrawal et al., 2010; Agrawal et al., 2012). In addition to recruitment of cells from adjacent tissues, ECM scaffolds have been shown to recruit bone marrow-derived cells to the site of implantation, which participate in the long-term remodeling response and contribute to newly formed tissue (Nieponice et al., 2013; Yoder et al., 1996; Zantop et al., 2006). In vitro assays have demonstrated that matrix remodeling includes a cell recruitment component and that inhibition of collagen degradation by addition of protease inhibitors prevents cell migration (Mauney et al., 2010). The peptide fragments of ECM that are generated by in vitro scaffold degradation have been shown to have chemoattractant properties for several cell types in vitro, including progenitor cells (Beattie et al., 2009; Brennan et al., 2008; Li et al., 2004; Reing et al., 2009).

Stiffness and geometry

Not only is the composition of the ECM scaffold critical to its function, scaffold stiffness and topological landscape can affect cellular response to the material. The mechanical properties of ECM have been shown to be important in the developing embryo where tissue stiffness affects organ development. In the case of the developing lung, stiff matrix causes endothelial cells to upregulate genes involved in angiogenesis (Mammoto et al., 2013). This stiffness is modulated by the ECM crosslinking protein lysyl oxidase (LOX). Inhibition of LOX disrupts ECM structures, downregulates angiogenic genes and inhibits lung development.

Studies of synthetic materials, in which substrate mechanics can be examined independently from biochemical signals, have demonstrated that subtle changes in stiffness can influence cellular behavior (Nemir and West, 2010). Substrate stiffness influences a variety of cellular properties including cell spreading, adhesion, proliferation, apoptosis, and differentiation. The differentiation of stem cells to adipogenic, myogenic, neurogenic, tenogenic, and osteogenic lineages is regulated by substrate stiffness (Engler et al., 2006; Gilbert et al., 2010; Guvendiren and Burdick, 2012; Sharma and Snedeker, 2010). During the decellularization of tissues for generatation of ECM bioscaffolds, much effort is placed on removing cellular debris while maintaining the stiffness and architecture of the native tissue.

In addition to substrate stiffness, ECM geometry can dictate the host response to biomaterials. The mode of cellular migration is strongly influenced by scaffold geometry, as cells can use both proteolytic mechanisms to digest their surrounding matrix and non-proteolytic mechanisms to squeeze through the ECM (Wolf and Friedl, 2011). Naturally derived ECM scaffolds can exhibit several types of geometries that influence cell migration in unique ways. Relatively dense layers of ECM, such as basement membranes, consist of laminins, crosslinked collagen type IV, and proteoglycans and act as a formidable barrier to migrating cells (Voisin et al., 2010). Loose fibrillar collagen networks present in most mesenchymal connective tissues consist of collagen fibers or bundles and interfibrillar gaps and pores of variable shapes and spatial arrangement. This type of ECM allows for cell migration through elongated gaps and tunnels within the structure (Wolf et al., 2009). Lastly, mineralized matrix such as bone provides a ridged ECM landscape that lacks interstitial spaces and allows for migration along the surface while prohibiting migration into the material.

Stiffness and geometry are carefully considered during the preparation of ECM biomaterials. An important goal of tissue decellularization is to remove cellular debris while maintaining native tissue properties such as stiffness and architecture as well as composition and ultrastructure. Maintenance of these properties is essential for promoting a constructive remodeling response as illustrated by the failure of remodeling with ECM scaffolds that have been altered by chemical crosslinking or other harsh methods that disrupt the structure of the material.

Conclusions and future directions

The functional importance of ECM is clearly demonstrated by the wide range of tissue defects or embryonic lethality that is caused by loss-of-function mutations in genes that encode ECM components. The ECM provides both a structural framework for cells and also participates in complex bidirectional signaling to promote cellular proliferation, migration, homeostasis, and differentiation. Dynamic ECM degradation and remodeling is a key feature of regeneration in animals with the ability to re-grow amputated body parts. In the adult mammal, many regenerative or repair processes such as skeletal fracture repair and wound healing recapitulate developmental mechanisms and require ECM remodeling to achieve normal healing. Harnessing the power of naturally derived ECM to promote the constructive and functional remodeling of tissues that cannot fully heal through endogenous mechanisms shows great promise. The examples outlined herein demonstrate that ECM can modulate the inflammatory response, recruit stem cells to the site of injury, and act as an inductive scaffold for the formation of new site-appropriate and functional tissues.

Many opportunities remain for further investigation into the use of ECM in regenerative medicine and for clinical translation of ECM therapies. While it is clear that progenitor cells are recruited to the site of ECM device implantation, the extent of their contribution to remodeling has not been determined. Several cell types, including myoblasts, perivascular stem cells, and bone marrow derived mesenchymal stem cells, have been implicated in the constructive remodeling of skeletal muscle tissue (Chen and Walters, 2013; Sicari et al., 2012; Turner et al., 2010). There is potential for the use of lineage tracing studies or cellular ablation experiments to further our understanding of the relative contribution of these cell types to the formation of new tissue during the repair, reconstructive and regenerative processes. This information could provide additional insight into the similarities between ECM bioscaffold induced remodeling and tissue development and healing.

Recent work in the field of regenerative medicine is attempting to reach beyond the reconstruction of relatively simple planar tissues and engineer complete organs for transplantation. This approach provides an attractive option as the shortage of viable organs for transplantation becomes more severe (Saidi, 2012). The whole organ engineering concept involves decellularization of a donor organ (allogeneic or xenogeneic) followed by the recellularization of the resultant 3-dimensional scaffold with autologous stem and progenitor cells with the goal of assembling and organizing cells and scaffold into a functional unit. To date, decellularized whole organ scaffolds have been prepared for liver, lung, kidney, and pancreas (Faulk et al., 2014; Keane et al., 2015). Recellularization of these scaffolds has been performed with varying degrees of success. Continuing work aims to optimize cell type, reseeding methods, and culture conditions to obtain transplantable, engineered organs.

The use of naturally derived ECM bioscaffolds has proven clinically successful for a variety of tissue engineering applications. A more detailed understanding of the complex organization of structural and functional molecules within ECM will provide a roadmap for the production of scaffold materials with the ability to stimulate and direct in vivo tissue regeneration. Towards this end, a number of groups have attempted to produce ‘smart’ biomaterials that incorporate biologic factors and synthetic polymers to achieve inductive or stimulating effects on surrounding cells and tissues. One approach is to link polymers to proteins and thereby tailor the surfaces that directly interact with biological environments. Biomaterial engineering with ECM-mimicking proteins has been extensively studied (Chua et al., 2008; Gomes et al., 2012; von der Mark et al., 2010; Xu et al., 2010; Zhu, 2010). For example biopolymers have been modified with collagen, fibronectin, or laminin to improve cell attachment and proliferation (Cao et al., 2011; Gümüşderelioğlu et al., 2011). However, due to the highly specialized composite structure of native ECM, generation of biomaterials to replicate the complex ECM/cell interactions remains a significant challenge. Understanding the embryonic genesis and patterning of the ECM and the mechanisms of bidirectional signaling during morphogenesis will provide critical information to direct regenerative medicine strategies that attempt to recapitulate the signaling mechanisms employed by ECM during healing and regeneration.