A Brief History of Tendon and Ligament Bioreactors: Impact and Future Prospects

Abstract

Bioreactors are powerful tools with the potential to model tissue development and disease in vitro. For nearly four decades, bioreactors have been used to create tendon and ligament tissue-engineered constructs in order to define basic mechanisms of cell function, extracellular matrix deposition, tissue organization, injury, and tissue remodeling. This review provides a historical perspective of tendon and ligament bioreactors and their contributions to this advancing field. First, we demonstrate the need for bioreactors to improve understanding of tendon and ligament function and dysfunction. Next, we detail the history and evolution of bioreactor development and design from simple stretching of explants to fabrication and stimulation of 2- and 3-dimensional constructs. Then, we demonstrate how research using tendon and ligament bioreactors has led to pivotal basic science and tissue engineering discoveries. Finally, we provide guidance for new basic, applied, and clinical research utilizing these valuable systems, recognizing that fundamental knowledge of cell-cell and cell-matrix interactions combined with appropriate mechanical and chemical stimulation of constructs could ultimately lead to functional tendon and ligament repairs in the coming decades.

INTRODUCTION

For nearly four decades, bioreactors have catalyzed the discovery of mechanical and biological factors that regulate tendon and ligament formation, homeostasis, and degeneration. In addition to their utility for the study of tendon and ligament physiology and pathology, bioreactors have been integral to tissue engineering novel grafts for tendon and ligament repair and replacement, currently in pre-clinical development. In this review, we will first define the clinical and biological relevance of tendon and ligament bioreactor systems, highlighting breakthroughs in basic and applied science. We will next present a brief history of tendon and ligament bioreactor development by first examining the challenges these tissues offer and then describing the evolution and increasing complexity of bioreactors to meet these challenges (Fig. 1A). We will then critically review two of the primary research areas of bioreactors: 1) basic science research into the mechanobiological mechanisms regulating cells and extracellular matrix (ECM); and 2) tissue engineering/regenerative medicine (TERM) research to create tissue-engineered constructs (TECs) to improve tendon and ligament repair and replacement (Fig. 1B). Finally, we will summarize how tendon and ligament bioreactors have impacted orthopaedic research and our thoughts for their future use. While this is not an exhaustive review of every use case for bioreactors, we trust that this historical perspective will help guide future studies utilizing these powerful tools with the ultimate goal of improving patient care.

Fig. 1: Evolution of bioreactors and their impact on the tendon and ligament field.

A) Tendon and ligament bioreactors have evolved from simple clamp and stretch devices to multiple constructs being loaded at one time with perfusion systems to deliver nutrients. B) As a result, the field has evolved from studies examining basic science mechanisms to more basic and applied mechanisms to tissue engineering/regenerative medicine research for pre-clinical development.

THE CLINICAL IMPORTANCE OF TENDON AND LIGAMENT BIOREACTORS

Musculoskeletal injuries are both common and debilitating in human and veterinary patients, with tendon and ligament injuries directly causing pain and reduced function, and indirectly causing osteoarthritis as a consequence of joint instability.^1,2 There are two main types of tendon and ligament injury: acute rupture due to mechanical loading above the biomechanical limit of the tissue, and chronic degeneration associated with structural and biological changes that fail to be repaired or regenerated over time.³ Tendon and ligament ruptures and partial ruptures occur due to trauma, sport, or work, and can be preceded by chronic degeneration of the tissue. Tendon and ligament degeneration (tendinopathy and desmopathy) results from repetitive overuse, ECM microdamage, and a failure of cells to remodel damaged extracellular matrix and replace it with normal tissue.^4,5 Commonly injured tendons include Achilles, patellar, quadriceps, hamstring, supraspinatus, and hand and wrist flexor and extensor tendons. Ligaments such as the anterior and posterior cruciate ligaments and medial and lateral collateral ligaments of the knee and ankle are also frequently damaged. An epidemiological study of patients presenting to an orthopaedic trauma center identified tendon or ligament injuries in up to 8.7% of patients.⁶ In a whole-village survey to identify the prevalence of supraspinatus injury, 20.1% of the sampled population had complete tears.⁷ These same soft tissue injuries occur at a similar prevalence in veterinary patients; this is particularly impactful when working dogs, sport horses and racehorses are injured. Injury of the suspensory ligament or superficial digital flexor tendon (analogous to the Achilles tendon in humans) reportedly occurs in 14.7% of racehorses.⁸ Supraspinatus tendinopathy is frequently diagnosed in dogs, especially working dogs and dogs used for sport, and cruciate ligament ruptures are common and often bilateral among medium to large-sized dogs.^9,10 Similar to those in human patients, all of these injuries cause chronic pain and carry a poor prognosis without costly advanced treatments or surgery. These injuries can lead to retirement from work/athletics and possibly euthanasia in veterinary patients.^9,11,12

Tendons and ligaments are indispensable for stability and motion of the musculoskeletal system, but our understanding of the biology of their development and diseases is incomplete. Tendons transmit the force of muscle to bone across a joint, and thereby facilitate movement and joint stabilization.¹³ Ligaments attach bone to bone across a joint and constrain joint movement within a normal range of motion.¹⁴ As such, ligaments experience more complex multi-directional loading, whereas tendons typically experience uniaxial loading except in regions where flexor or extensor tendons wrap around a pulley. Therefore, the mechanical properties of tendons and ligaments are critical to their function. Tendons and ligaments comprise dense connective tissues with abundant ECM. Partially influenced by the types of loads that they experience, tendons have more aligned ECM than ligaments, while ligaments have higher levels of non-collagenous ground substance (e.g., proteoglycans, glycoproteins, water, etc.). Tendon and ligament fibroblasts are arranged in linear arrays.^15-17 Their expression patterns differ depending on the tendon or ligament (i.e. intrasynovial vs. extrasynovial)¹⁸ and unique markers have still not been established for specific stages of tenogenic or ligamentous differentiation. However, a number of groups are currently using single-cell analyses to demonstrate heterogeneity within these cell populations [unpublished work].^19-21 Tendons and ligaments also have low vascularity. This low vascularity is thought to contribute to the poor innate healing of these tissues, especially within synovial compartments such as tendon sheaths (e.g., digital flexor tendons) and joints (e.g., cruciate ligaments). Changes in ECM content, hierarchical organization, crosslinking, bundling, and integrity are responsive to mechanical load and directly responsible for the mechanical properties of the tissues.^22,23 Critical biomechanical parameters for tendon and ligament function include strength (maximum load), elasticity (stiffness), toughness, and viscoelasticity.^24-26 Though tendons and ligaments are predominantly loaded in tension, the anatomic location provides specific mechanical requirements, especially as these tissues span curves resulting in combinations of tension, compression, torsion, shear, and friction to different regions of the tissue.^27-29 Similar to bone, microdamage of the tendon and ligament ECM can result from excessive strain or fatigue loading, and mechanical forces can be high enough to overload the structural integrity of tendon and ligament resulting in partial or total rupture. Unlike bone, however, tendons and ligaments heal poorly and slowly. While there is no regenerative mechanism for healing these ruptured tissues, it is clinically demonstrated that physical therapy can improve patient outcomes suggesting a role for mechanical forces in improved healing. Consequently, more work is needed to understand degeneration, the roles of exercise and aging on remodeling and homeostasis, and potential molecular and cellular targets for therapy.

Prior to the development of ex vivo tendon explants, TECs, and bioreactors, little was known about the essential relationship between mechanical stimulation and tissue development, maturation, injury and healing, particularly the molecular and cellular details of mechanotransduction in tendon and ligament.^30-33 In contrast, mechanobiological signal transduction pathways and regulation of bone and cartilage cellular responses have long been well characterized. For cartilage and bone, a combination of in vivo and ex vivo studies unlocked the relationship between force and cellular response;^34-37 this same combination is likely required for tendon and ligament. Bioreactors offer this possibility of investigating important basic science questions related to the mechanobiology of injury, healing, repair, and replacement.

Treating tendon and ligament injuries can be challenging and costly with or without surgery. Major surgery to repair, reconstruct, or engraft human tendon, ligament, and other soft connective tissues represents ~20% (~2.5 million) of all major musculoskeletal procedures in the US.³⁸ Surgical repair or reconstruction of tendon and ligament ruptures often results in complications and reduced function (e.g., anterior cruciate ligament) or repair failure (e.g., supraspinatus tendon).^39,40 Primary repairs are most common in tendon and can be successful depending upon location, severity of injury and time to surgery, and rehabilitation protocols.⁴¹ Delayed repair and extensive tissue loss result in gaps between tendon ends, which can require tendon grafting with autografts or allografts. Grafting of flexor and extensor tendons of the hand is especially challenging since restoration of function and dexterity require both a strong repair and effective gliding.⁴² Tendon autografts and allografts are also used in anterior cruciate ligament reconstruction, which is often more successful than direct repair. Autografts are preferred to allografts due to better tissue integration, recellularization and revascularization; however, donor site morbidity can be significant.^43-46 Developing custom-designed and engineered tendon and ligament grafts, namely TECs, conditioned in tendon/ligament bioreactors, would eliminate the issue of donor site morbidity, and allow for a superior match to recipient site requirements. While we recognize the unmet clinical needs and the early status of TEC and bioreactor development, our field must seek to employ these strategies (e.g., functional tissue engineering to mechanically stimulate tissue-engineered ligaments and tendons in bioreactors) and seek greater investment in basic science and preclinical research to improve soft tissue care in humans and animals alike.⁴⁷

THE HISTORY OF TENDON AND LIGAMENT BIOREACTORS

While a multitude of preclinical in vivo models have been used to study tendon and ligament biology, the complexity of performing these studies in vivo has highlighted the usefulness of bioreactor research. The pitfalls of tendon and ligament animal studies are many and include: high variation between subjects; limited technical replicates; inability to fully control mechanical input; difficulty in controlled local drug delivery; difficulty in real-time data collection; inefficient collection of outcome data; and inherent animal research pitfalls, including expense, labor, and animal welfare concerns. Bioreactors offer a number of advantages over in vivo studies in that mechanical or chemical inputs can be precisely controlled and the cells and/or tissue can be isolated from systemic factors in the body. Bioreactor designs have evolved to meet the needs of researchers and address key gaps in knowledge, from basic science questions of how mechanical forces alter cell behavior to the production of TECs to replace or augment tendon repairs in vivo. The following section will highlight these key advancements in bioreactor design.

Explant Bioreactors

Early bioreactors were “clamp and stretch” devices to study the effect of mechanical loading on bovine, canine, ovine, caprine, and avian tendons and ligaments (Figs. 1A and 2).^48-53 Over time, such ex vivo studies showed that static or cyclic mechanical loading regulate tendon and ligament mechanical properties, cell-cell communication, cell division, cell viability, gene expression, and ECM production and turnover.^48-50,54,55 The hallmarks of “clamp and stretch” bioreactors are the application of mechanical loads, often cyclic loads, while sensing deformation and/or forces in a physiologic environment. These systems allow examination of cell and tissue responses to physiologic loads compared with no load and supraphysiologic loads, sometimes while visualizing the tissue in real time.⁵⁶ They have been used to study both the biomechanical properties of tendon tissue, and the cellular responses to different loading regimens. These studies introduced the mechanobiological principles related to tendon homeostasis, healing, degeneration, and injury.

Fig. 2. An example of “clamp and stretch” explant bioreactors.

(A) A flexor digitorum profundus (FDP) tendon contacting the lower clamp face of the bottom set of jaws for (B) specimen holding for the tendon loading device. The bar across the tendon fits into the semicircular channel under the tendon and protrudes from each jaw edge so that it may be held by two arms that fit into the tendon loading device body. The other half of the jaw fits on top of the tendon and bar and is screwed together with two stainless steel screws. (C) The FDP tendon clamped in top and bottom jaws and inserted in a loading frame of the tendon loading device is shown. (D) The tendons loaded in the tendon loading device are shown but the tendon loading frame and tendon are immersed in culture medium. The cotton gauze at the top covers the tube top, is immobilized with a rubber band and prevents exposure to contaminated material. (Banes et al., 1999).⁴⁸

Two-dimensional (2D) Bioreactors

Two-dimensional (2D) bioreactors to study cellular response to mechanical loading have evolved rapidly since the mid-1980s, when Banes et al. helped lay the foundation for the tendon mechanobiology field.^57,58 Using tendon internal fibroblasts (TIFs) isolated from nearby collagen fibrils in avian flexor hallucis longus tendon, they found TIFs responded differently to applied strain than cells procured from surface paratenon and epitenon, involving time-dependent changes to the cytoskeleton.^48,58-64 Groups then used patterned surfaces or fibrous scaffolds (i.e., 2.5D environment) to study the effects of surface geometry, organization, and mechanical loading on cell behavior (Fig. 3).^58,65-70 While these advancements are important to our understanding of tendon cell biology, the three-dimensional (3D) organization of native tendon naturally led to the development of 3D bioreactors to better model in vivo conditions.

Fig. 3: Surface geometries dictate cell shape and behavior.

Studies using 2.5D bioreactor designs have determined the effects of surface topography on cell alignment of tendon fibroblasts. For instance, elongated grooves in the cell culture surface drive cells to align longitudinally with the grooves (A-C). Cells also align in a longitudinal pattern when cultured on aligned electrospun scaffolds, but not randomly arranged fibers (D-F). Panel A was adapted from Yang, et al., 2004,⁶⁸ panels B-C were adapted from Jones, et al., 2005,⁶⁵ and panels D-F were adapted from Schoenenberger, et al., 2018.⁷⁰

Three-Dimensional (3D) Tissue Engineered Construct (TEC) bioreactors

This “next” generation of bioreactors was designed to test constructs containing cells embedded in collagen or fibrin gel scaffolds and constrained geometrically so that cells could contract and compact the scaffold material into a uniaxial arrangement (Fig. 4). These constructs were variously named bioartificial tendons (BATs), tissue-engineered constructs (TECs), and ligament equivalents (LEs).^47,71-73 These systems included geometric constraints such as posts/anchors at ends of the gel or a tensioned suture running through the gel.^71-76 These 3D static-loaded systems not only induced longitudinally-arranged cells and ECM, but also increased ECM production and led to tendon- or ligament-like tissue depending on actual tissue type and application (Fig. 4A).^71-73 The arrangement of these geometric constraints was critical to controlling cell phenotype as aspect ratio could dictate matrix alignment (Fig. 4B)⁷⁷ and incorrect arrangements (e.g., tensioned suture through gel) could even lead to aberrant osteogenic differentiation with mineral deposition, highlighting the importance of geometric constraints and the resultant environment on cell phenotype.^78,79 Currently, researchers are even producing 3D tendon or ligament constructs without an initial matrix (i.e., scaffold-less constructs) by seeding cells at high densities into channels with high aspect ratios or cultured as cell sheets in monolayer prior to being rolled into a uniaxial construct.^80-85 Although these studies avoid the need for an initial scaffold that can impede the ability of the cells to assemble and organize their own matrix, the very large number of cells required limit scale-up to larger sizes compared to scaffold-based constructs. These static 3D bioreactors have been useful for rapid comparison of cell types for their tenogenic capacity, and for a systemic determination of which growth factors or growth factor combinations further induce tenogenesis.^74,86

Fig. 4: Geometric constraints are used in bioreactors to dictate construct shape and cellular response.

A) 3D bioreactor designs that harness inherent cell contraction of ECM when embedded in hydrogels anchored at two ends resulting in uniaxial cell and matrix alignment. The aspect ratio, as defined by the distance between the anchors divided by their width, has a significant effect on construct shape as seen in these two examples where the TEC with the polyethylene posts (top row) had a “racetrack” configuration compared to the uniaxial configuration with the silk suture anchors (bottom row). B) Aspect ratio also dictates collagen alignment within the TEC with high aspect ratio configurations leading to higher and more uniform alignment. Panel A was adapted from Huang, et al., 1993 and Kapacee, et al., 2008.^73,86 Panel B was adapted from Nirmalanandhan, et al., 2007.⁷⁷

Three-Dimensional (3D) Bioreactors with Mechanical Stimulation

Since tendons and ligaments experience cyclical loading during daily activities, the next generation in bioreactor design was to apply mechanical conditioning along a principal strain direction to drive cells within the construct to synthesize and better organize ECM for ultimate linear tendon or multi-axial ligament applications.^71,87-92 With applied tensile strain, cells aligned with their cell bodies and nuclei, and actin filaments oriented along the principal strain direction. At first, simple materials testing systems were adapted for the application of cyclic load to TECs. Then, more sophisticated systems were developed that a) allowed loading of more delicate scaffolds, b) collected load and deformation data in real-time, and c) incorporated the ability to image the TECs using confocal microscopy. Studies using these approaches revealed that tendon, ligament, and even bone marrow stromal cells in 3D could respond in a dose-dependent fashion to applied strain of a given magnitude and frequency, maintain a cell expression profile, and increase breaking strength.^88,92-97 These studies led to the tendon and ligament bioreactor field as we know it today.^98-100

Evolving Technologies in Bioreactor Design

Advances in bioreactor design are being made on several fronts (Fig. 5). The waveform of the mechanical stimulus is continually being refined, honing in on low (ligament) and higher (tendon) physiologic amplitudes and frequencies to elicit the desired response.^47,101,102 Application of multi-dimensional strains provided more physiological loading regimens, relevant for certain ligaments.⁹¹ By recording the forces within constructs during the culture period, investigators could measure how the constructs matured with time without stopping the experiment for a terminal response measure.^103,104 Other systems, in which multiple samples could be loaded at one time, increased throughput, resulting in more replicates and treatment groups for experiments and more TECs for tissue-engineered repairs.^91,105,106 Finally, constructs are now infused with additional biologics during the culture period to improve matrix production and the ability to make larger constructs needed for translation.^91,107 Many of the new bioreactor designs were commercialized into systems such as the Ligagen^® from BISS,¹⁰² the Tissue Train^® from Flexcell,⁷¹ and the BioDynamic 5200^® from TA Instruments, thus providing basic and clinician scientists access to elegantly engineered and sophisticated tendon and ligament bioreactors.

Fig. 5: Increasing complexities of tendon and ligament bioreactors.

(A) Newer bioreactor designs increased throughput by applying a consistent load to several samples at once, as seen in the FlexCell Tissue Train system with 24 samples. (B) Other sophisticated bioreactors were developed to include multi-dimensional force application, as well as perfusion of nutrients, which is critical to development of larger TECs. Panel A was adapted from Garvin, et al., 2003.⁷¹ Panel B was adapted from Altman, et al., 2002.⁹¹

THE IMPORTANCE OF BIOREACTORS IN STUDYING BASIC MECHANISMS OF TENDONS AND LIGAMENTS

The mechanisms that regulate proper tendon or ligament cell differentiation and subsequent formation of ECM capable of withstanding normal tissue forces during ADLs are not fully understood. Additionally, our knowledge of the mechanisms by which cells maintain the ECM during homeostasis, healing, repair, and disease are incomplete. While a multitude of preclinical in vivo models address these questions, there is a need to control the cells’ local environment therefore bioreactors were designed to fill this need. Described below is a sampling of the impact of bioreactor studies on our understanding of different facets of tendon and ligament biology.

The objective of several bioreactor studies has been to better understand the tenogenic and ligamentous cell differentiation processes. Since the first report of resident tendon stem/progenitor cells (TSPCs),¹⁰⁸ it is now accepted that these populations exist in tendons and ligaments, although further identification of individual markers that specifically define this population and the stimuli that regulate their differentiation are still needed.^109,110 Nonetheless, tendon/ligament resident stem/progenitor cells may be a more suitable cell source for tissue-engineered repairs, compared to previously used bone marrow-derived stromal cells. Bioreactors are a pivotal platform to test the effect of different stimuli on the differentiation of these cells in a controlled environment. Whether starting with endogenous (i.e., tendon stem/progenitor cells, TSPCs) or exogenous (i.e., bone marrow- or adipose-derived) stem/progenitor cell populations, the goal of these studies remains to define genetic, biochemical, and/or mechanical stimuli that drive differentiation to mature tendon or ligament fibroblasts.^{100,102,104,111} In vivo, transforming growth factor-β (TGF-β) signaling was found to be a pivotal pathway in specification and differentiation of tendon progenitor cells during growth and development.^112,113 Stimulation of cells with TGF-β in bioreactors was shown to increase expression of tenogenic markers; matrix production and organization; and mechanical properties.^114,115 Another TGF-β superfamily member, bone morphogenic protein 12 (BMP-12), was shown to increase bone marrow stromal cell expression of the tendon marker scleraxis in a 3D bioreactor and to improve longitudinal alignment of cells when compared with TGF-β, IGF-1, or fibroblast growth factor 2 (FGF-2).¹¹⁶ These bioreactor studies are being done in parallel with in vivo studies to better define markers of the cell lineage and stimuli that regulate differentiation. Bioreactor studies will be pivotal to validate in vivo findings and fine-tune the local environmental cues that are unique and critical for teno- and ligamento-genesis.

Bioreactors offer a versatile tool to elucidate the interplay among mechanical and/or chemical stimulation, cellular responses, matrix turnover, and tissue structure and function. Mechanical stimulation influences collagen intra- and intermolecular cross-link profile,^117-119 collagen fibril diameter,¹²⁰ fiber orientation, density, and length,^53,121 which ultimately impact the tissue’s functional properties, including its tensile strength, stiffness, and viscoelasticity.^122-124 As insufficient or excessive mechanical stimulation can induce matrix metalloproteinase (MMP) expression and other pathologic changes in tendon, many bioreactor studies have helped to reveal the relationships between mechanical stimulus and tendon homeostasis.^125,126 Loading amplitudes distinctly affect cell and tissue properties with i) no mechanical stimulation resulting in increased MMP expression, cell apoptosis, and loss of mechanical strength, ii) physiologic loads leading to an overall anabolic response and increased strength, and iii) high-amplitude loads leading again to a catabolic response.^{50,54,55,127-130} Growth factors are also important to tendon homeostasis, and bioreactor studies have demonstrated a synergistic relationship between mechanical load and growth factors on tendon cell function.¹³¹ Additionally, cell communication is critical to these processes as inhibition of gap junctions leads to reduced cell division and collagen production⁴⁸ while primary cilia mediate force transduction.⁵⁶ The rapid advancement in our understanding of mechanical and/or chemical cues on tendon and ligament homeostasis were certainly aided by pivotal bioreactor studies.

THE IMPORTANCE OF BIOREACTORS IN TISSUE ENGINEERING AND REGENERATIVE MEDICINE

Over the past 25 years, researchers have created and mechanically and/or chemically “preconditioned” many different cell-scaffold tissue-engineered constructs (TECs) in bioreactors. The level of preconditioning needed to produce a TEC that is capable of improving the repair outcome when implanted likely depends on the local conditions (i.e., biological and mechanical environment) of the tendon or ligament being repaired. It also likely depends on the cell source as TSPCs might not require as much preconditioning as bone marrow- or adipose-derived stem/progenitor cells. The functional tissue engineering (FTE) paradigm was developed to established benchmarks for these repairs to guide the course of stimuli and regimen (frequency, duration and magnitude) of preconditioning required to produce an efficacious tissue engineered repair.⁴⁷

The in vitro studies described previously have focused on establishing how cellular factors (e.g., cell type, density, time in culture), scaffold conditions (e.g., biomaterial type, density, alignment, time in culture), and mechanical and/or electrical (e.g., amplitude, duration, frequency, microamps)⁹² and/or chemical (e.g., growth factor density, time of introduction) and perfusion stimuli influence TEC maturation before implantation at surgery. These bioreactors, with or without mechanical, chemical, electrical, or shear stress stimulation, have been able to induce cells cultured in geometrically defined constructs (e.g., linear, circular, off-axis), to compact a scaffold and align, increase collagen gene expression, and even cross-link newly formed ECM.^{47,71,74,76,78,88,90,132-137} However, the sheer numbers of cellular, scaffold, mechanical, and chemical combinations is overwhelming, especially since the result depends on the specific wound site conditions in vivo. Consequently, human in vivo studies have still not been permitted using TECs to treat tendon and ligament injuries. Instead, preclinical models have been employed to discover combinations leading to more rapid and successful soft tissue repairs.⁴⁷

To overcome some of these challenges, research teams have been employing in vitro as well as in vitro-to-in vivo strategies to improve TEC performance and repair outcomes.

In vitro strategies.

Optimization algorithms like Response Surface Methodology (RSM) have identified promising in vitro parameters for stimulating the TEC in culture so as to better predict in vivo performance.⁹⁶ An investigator might study how three densities of bone marrow-derived progenitor cells (high, medium, low) and three alignments or lengths of collagen sponge (0, 45 and 90 degrees to loading axis) might affect a TEC’s linear stiffness over time in culture.^{77,92,96,138-140} The effects of mechanical strain rate (static, low and high frequencies) and growth factor augmentation (low, medium, high density of TGF-β) can also be examined efficiently in culture.^90,95 Choosing relevant levels of the input variables and appropriate outcome measures offer the possibility of better optimizing the TEC characteristics before creating and delivering functional constructs to the surgeon.

In vitro-to-in vivo strategies.

The questions here are how will the TEC perform: 1) when handled by the surgeon; 2) when inserted into a specific tendon or ligament wound site at surgery; and 3) over time for the patient’s expected activity levels. One attractive strategy for achieving some of these goals has been the use of Functional Tissue Engineering and design parameters.^{97,137,141-146} Beginning with “the end in mind,” investigators used in vivo force transducers^29,147 to first measure or model maximum in vivo forces acting on specific ligaments and tendons in several animal models for different ADLs.^{24-29,72,105,106,143,148-150} Ligaments were found to normally transmit only 7–10% of the tissue’s maximum force while tendons could sustain up to 40% of max force. Knowing the force-deformation properties of these tissues up to failure, relevant mechanical properties could then be established in the “in vivo” range to serve as benchmarks and design criteria for judging the TEC at the time of surgery and for the TEC-based repair post-surgery. This strategy resulted in studies where MSC-collagen sponge constructs were mechanically preconditioned in the bioreactor up to peak in vivo strains at expected in vivo frequencies and then implanted into rabbit full-length central patellar tendon defects. At 3 months after surgery, repairs were found to match native tissue stiffness and withstand up to 150% of expected in vivo forces.¹⁰⁵ These studies demonstrated great value in tissue engineering approaches to improve tendon repair. However, they were still limited to cases where the TEC was only used to augment the repair as these constructs could not resist native forces experienced by the tendon at the time of surgery.

What pre-requisites must bioreactors possess in order to meet these fundamental as well as translational strategies? Fundamental studies in culture offer the possibility of high throughput in multi-well plates (e.g., in the Flexcell Tissue Train system) to investigate how many treatments (mechanical and chemical) affect numerous outcome measures in culture. While samples are usually small and sometimes fragile, they are likely sufficient since these biologics would not be implanted in vivo. Temperature, pH, oxygen concentration, and nutrient composition and delivery must be carefully controlled in each chamber. The need to establish success criteria for the chosen in vitro measures is critical. The translational studies from in vitro-to-in vivo application pose even greater challenges. TECs must be made suturable, larger, stiffer, and stronger to tolerate: 1) potentially rough handling by the surgeon plus time constraints in the OR; 2) substantial in vivo forces early post-op; and 3) even larger, unknown forces once patients resume normal ADLs. Larger systems like the BISS LigaGen and TA Instruments Biodynamic 5200 could possibly tolerate these higher loads (still in the range of 50 pounds force, however) plus larger displacements and sample lengths. However, only 1 to 4 samples can be conditioned at a time and each sample would tie up the system during conditioning. So important in this applied/clinical research are the need to: a) establish or at least estimate force limits for each tissue application; and b) develop in vitro predictors of in vivo outcome (e.g., TEC stiffness as predictor of in vivo repair stiffness).^{47,137,151,152} Moreover, controlling costs by maintaining high throughput and minimizing time in culture could limit the choice of bioreactor. Understanding and overcoming regulatory hurdles could also reduce the cell and scaffold options. A single bioreactor bridging the gap between basic science studies that condition a larger number of small samples and truly safe, efficacious and cost-effective cell therapies in patients remains a formidable task. Although these clinical studies are still possibly decades away, they remain a primary goal of the TERM community along with the need to better mimic true tendon and ligament development and mechanobiology.^137,153,154

CONCLUSION AND FUTURE DIRECTIONS

Multiple decades of bioreactor research have taught us that cells comprising a tendon or ligament are not only matrix-making machines, but also capable of communicating with each other through complex signaling channels. Tenocytes and ligament cells express not only growth factor receptors, but also adrenergic and purinergic receptors. They also express gap junction proteins, are connected in a syncytium, signal each other with ligands and load, and can be anabolic or catabolic in nature.^155-158 Tenocytes are more than “generic” fibroblasts, instead they are quite sophisticated. Additionally, we are just beginning to scratch the surface into the dynamics of resident stem/progenitor cells and their contribution to mature fibroblasts during development, homeostasis, and repair. No doubt mechanical and biological factors will need to be added temporally to drive cell differentiation and development of a strong matrix that integrates into an enthesis, tendon, ligament or appropriate muscle where applicable (or all structures in sequence in the bioreactor). Bioreactors for developing tendon or ligament constructs that can be practically integrated into host tissue are still in the preclinical study-model phase rather than in clinical trials. There have been natural and synthetic materials applied to the problems of tendon and ligament reconstruction without clear success and scale-up to broad use. While autografts and allografts remain the “Gold Standards” for surgical repairs, hopefully bioreactors will in the coming decades be used to create constructs that can be implanted in a patient and function therapeutically for an acceptable period.

In summary, mechanical and chemical stimulation of tendon and ligament constructs in bioreactors can advance our understanding of basic science mechanisms as well as advance the development of novel reparative strategies to mature engineered tissues that follow functional tissue engineering principles.^47,141 These include: 1) mechanical benchmarks such as achieving tensile strength exceeding peak in vivo forces typically experienced by the tissue being replaced as well as matching the healthy tissue’s tangent stiffness up to these peak forces with safety factors built in; and 2) biological benchmarks such as inducing the appropriate cellular phenotype of scleraxis-expressing cells for tendon vs. ligament applications, achieving the native matrix collagen fibrillar organization and parallel (tendon) vs. multi-directional (ligament) alignment, and ensuring the zonal fibrocartilaginous entheses in sites of boney insertion. Equally important, mechanical and chemical stimulation “tuned” in bioreactors must become indispensable tools to advance basic and translatable knowledge to elucidate the mechanobiology of healthy and injured tendons and ligaments.