Bone physiology can be examined on multiple length scales. Results of cell-level studies, typically carried out in vitro, are often extrapolated to attempt to understand tissue and organ physiology. Results of organ- or organism-level studies are often analyzed to deduce the state(s) of the cells within the larger system(s). Although phenomena on all of these scales—cell, tissue, organ, system, organism—are interlinked and contribute to the overall health and function of bone tissue, it is difficult to relate research among these scales. For example, groups of cells in an exogenous, in vitro environment that is well defined by the researcher would not be expected to function similarly to those in a dynamic, endogenous environment, dictated by systemic as well as organismal physiology. This review of the literature on bone cell culture describes potential causes and components of cell “culture shock,” i.e., behavioral variations associated with the transition from in vivo to in vitro environment, focusing on investigations of mechanotransduction and experimental approaches to mimic aspects of bone tissue on a macroscopic scale. The state of the art is reviewed, and new paradigms are suggested to begin bridging the gap between two-dimensional cell cultures in petri dishes and the three-dimensional environment of living bone tissue.
- tissue engineering
- mechanochemical transduction
- fluid flow
bone cells reside in a mechanically active, composite tissue comprising an “organic-mineral” solid phase and a dynamic aqueous phase that permeates spaces within the porous solid matrix and acts as a carrier for ions, proteins, and other nutrients. Bone's mechanobiological function modulates its organization and structure across length scales, e.g., from gross anatomical features at the organ level to the microarchitecture of trabeculae at the tissue level, to the anisotropic arrangement of the canaliculi defining the pericellular space, to the arrangement of extracellular matrix molecules within the pericellular space and extracellular matrix. More than a century of investigation provides a foundation for organ- and tissue-level understanding of bone mechanobiology and its role in bone physiology. In comparison, understanding of bone cell mechanobiology is in its nascency.
An understanding of bone cell mechanobiology is critical for understanding of bone physiology because the cells are the “micromachines” that actively model, maintain, and remodel the tissue (71, 72). Mesenchyme-derived bone-lining cells and osteoblasts lay down osteoid, an organic matrix, which is subsequently mineralized with a hydroxyapatite-like calcium phosphate. Leukocyte-derived osteoclasts dissolve mineral and digest matrix, providing space and potentially biochemical cues for osteoblasts to lay down new bone matrix (135). Osteoblasts that become embedded in new bone differentiate into a network of interconnected osteocytes that may play a key role in mechanotransduction (72). Remodeling provides a means for repair and adaptation of bone tissue in response to injury as well as changing mechanical and metabolic demands. However, though cellular activities associated with remodeling appear to be highly coordinated, the signaling and timing of interactions among osteocytes, osteoclasts, osteoblasts, and pluripotent cells are not clear. Although this has provided impetus for continued investigation, in vivo and in situ investigations of bone cells and their coordination during modeling and remodeling activity have been extremely limited.
One major barrier to understanding bone physiology at a cellular level is the lack of methodology to study bone cells in their native environment. Contributing to this difficulty are the fundamentally different experimental approaches used to examine bone physiology/biology on multiple scales. Bone can be approached as an organ within a system, where observations of bone function can then be extrapolated empirically to create informed hypotheses about the functions of the constituent cells. Alternatively, the cells can be investigated independently of their native environment, i.e., outside of the physiological system, organ, tissue, and perhaps even excluding the extracellular matrix. Macroscopic investigations may succeed in maintaining the native mechanobiological environment but by definition fall short in allowing for observation of cell-level responses. In contrast, a rich armamentarium of techniques has been developed to measure cell activity outcomes precisely in vitro. An inherent drawback of in vitro studies is that the cell is cultured in an exogenous setting. When a cell is removed from its native environment and reintroduced into an extracorporeal setting, it is expected that the cell will have to adapt to its new environment; the cell must recondition itself to respond to different cues. We suggest that that this “culture shock” (116) can be mitigated by better emulating physiological conditions in vitro.
Some of the distinct advantages of the cell culture methodology include controlled definition of culture environment conditions, including pH, temperature, osmolarity, and solute concentration. Control of these parameters allows for the design of experiments that would not be feasible in situ or in vivo. This is due to the fact that the biochemical and mechanobiological milieu of bone cells is challenging to observe in situ and in vivo. Nonetheless, specific, basic environmental parameters, including pH, osmolarity, and solute concentration, have not been measured extensively in vivo. For example, it is known that bone extracellular fluid is distinctly different from that of plasma or interstitial fluid, e.g., in other inner organs, particularly considering its high content of K+ (71). In the same vein, it is assumed, although not yet measured in situ, that mechanical loads imparted through bone as a natural consequence of physiological activity influence the local mechanobiological environment of the cell through both direct deformation of the pericellular environment and indirect effects of load-induced fluid movement, including drag and shear stresses at the cell surface, as well as transport of nutrients, chemokines, and other osteotropic factors (71). Furthermore, although a number of technologies are being developed to optimize cell culture environments, no single approach has been developed to bridge the gap in understanding between single-cell physiology, the physiology of cells cultured in groups (e.g., to confluency), and the physiology of cellular networks in tissues and organs. Several multidisciplinary approaches have been applied recently. These include the use of knockout mice with musculoskeletal modifications (58) as well as the identification of genes responsible for regulation of osteogenic cytokines (13) and genes giving rise to various bone disorders (38, 84). Furthermore, animal models have been developed and refined to correlate bone morphology and damage with apoptosis (20, 92) to study fracture repair (93), microgravity effects (21), and pathological conditions (62, 91, 119).
Whereas the local environment of a cell may modulate its activity at any given time, cell-level communication and transport are coordinated across the tissue and at the organ level (74). Few studies integrate findings from organ-, tissue-, and cell-level studies, which hinders translation of new fundamental insights to clinical modalities. Hence, the purpose of this compendium is to examine current approaches to bone cell culture to reduce “cellular culture shock” through improvement of methodologies emulating physiological conditions. Identifying commonalities as well as implications of bone mechanobiology at the organ, tissue, and cellular levels will advance the ultimate goal of developing new strategies to prevent, cure, and/or replace diseased bone tissue in the future.
BONE CELL LINES AND CULTURE
An inherent strength of cell culture is that the cells are derived from a common source that can be passaged as many as 50 times while maintaining the basic phenotypic characteristics of the line. This aids in study design because a “cohort of study subjects” is readily available, and it allows for comparison of data from a given cell line that is cultured according to standardized protocols. However, it should be noted that there can be significant variation among cells isolated from a single source; in fact, this is exploited to produce subclones of cell lines with specifically desired characteristics (95, 113, 139). Furthermore, comparison of data derived from studies using different cell lines may not be appropriate, as it is analogous to comparing outcome measures (e.g., bone remodeling dynamics) between different cohorts of patients or different species.
According to current practice, bone cells are either isolated from primary human or animal tissue or obtained from an immortalized or otherwise transformed cell line. Immortalized cell lines take advantage of genetic instabilities in cells that have been cultured from pathological source tissue, selected from late-generation primary cultures, or transfected with specific genes. Osteoblast cell lines are commonly derived from either animal or human osteosarcoma. A recently isolated osteocyte-like cell line, murine long bone osteocyte Y-4 (MLO-Y4), exhibits high levels of osteocalcin and low levels of alkaline phosphatase production as well as expression of gap junctional proteins such as connexin 43 (Cx43), similar to primary osteocytes and distinct from primary osteoblasts (66); hence, this cell line is used almost exclusively as an osteocyte model (Fig. 1, refer to Table 1 for an overview of bone cell lines reported in the literature).
Alternatively, primary osteogenic cell cultures can be obtained directly from healthy donor tissue and are, as such, the “real cells,” transplanted from the endogenous to an exogenous environment. Major drawbacks of this approach are that considerable variation can exist between cells from different tissue sources, cell populations obtained from a single tissue source can be heterogeneic depending on isolation and purification procedures, and these cell cultures can typically be maintained for only three to four generations before phenotypic drift becomes evident. In fact, significant changes in behavior (contractile force) have been reported between cells of successive generations within only three passages (154). A variety of methods are employed to isolate primary cells. One common method involves digestion of the minced bone explanted from sites such as long bones, (fetal) calvaria, mandibles, or the iliac crest with collagenase or similar enzymes, followed by subculture of the resulting cells (17, 83, 143). A variation on this technique entails culture of cells that grow out of undigested bone chips (41). A potential shortcoming of this approach is that it does not allow selection or exclusion of particular cell types; however, the use of improved culture substrates with better bone fixation (137) and the inclusion of differentiation-inducing osteogenic factors such as dexamethasone, ascorbate, and phosphates in the culture media generate cells that are distinctly osteogenic (63). Osteogenic precursor cells are otherwise routinely isolated from bone marrow and can become mature osteogenic cells by the inclusion of osteogenic factors in the culture medium (50, 134). Many current efforts to improve the homogeneity of marrow-derived cells include utilization of protocols that take advantage of antibody-based detection of select membrane receptors that indicate predisposition to specific cell types (107, 125). Primary osteoclast culture is also possible via isolation from bone tissue (143) and, more efficiently, from the hematopoietic stem cells in bone marrow via exposure to intercellular signaling factors, including osteoclast growth factor (30), specific interleukins (44), macrophage-stimulating factor (128), colony-stimulating factors (76, 120, 126, 147), receptor activator of nuclear factor-κB ligand (RANKL) (127, 148), tumor necrosis factor (81), osteoprotegerin (121), and B lymphocytes (57). Coculture with osteoblasts and osteocytes known to express these factors also induces osteoclastogenesis, which can be enhanced, in turn, by a wide variety of stimulating factors (6a, 29, 67).
Although the use of certain specific cell lines, such as MC3T3-E1 subclones for osteoblast-like cells and MLO-Y4 cells for osteocyte-like cells (2, 110, 132, 149), is becoming more commonplace, there remains no standard cell line or preferred isolation method for bone cell study and no standardized maintenance protocols following culture. Because these methods tend to be selected for ease of use, for known responses to particular stimuli, and to model specific stages of the osteocyte lineage, it is unlikely that standards will be established without a significant shift in the current paradigm. However, more routine use of multiple cell lines within similar experimental contexts may better define the differences among the cell lines and provide a more complete understanding of the implications of their use in mechanobiological research (for further review of bone cell and tissue culture methods, see Ref. 85a).
ROLE OF THE “THIRD DIMENSION” IN EMULATING THE CELLULAR ENVIRONMENT
Regardless of their source, bone cells removed from their native setting occupy an exogenous environment. Although the environment is not alien enough to kill the cell, the effects of such environmental changes have not been thoroughly investigated. For example, two-dimensional culture models of osteocytes may not adequately mimic the behavior of native osteocytes that reside naturally in a disparate three-dimensional network. The complex mechanochemical environment of the periosteal and endosteal surfaces, cutting cones, lacunae within bone, or the marrow cavity is often replaced by culture on flat plastic surfaces, within a nutrient medium bereft of all but the most essential ions, peptides, and proteins. Response to these abrupt and substantial changes, or culture shock, may play a significant role in altering the response of cultured mammalian cells and in inducing cellular senescence (116, 144).
INFLUENCES OF CULTURE MEDIUM ON CELL FUNCTIONS
Cell lines derived from murine tissue, a common culture source, are prone to react poorly to inadequate in vitro conditions. One of the principal possible causes of this is excess oxidative stress found in culture environments; standard culture conditions (95% air-5% CO2, 37°C) lead to formation of reactive oxidative species that are normally limited in vivo (45). Interestingly, although cell biologists have developed and optimized culture media so that cells remain viable and functional, the chemical composition of bone fluids are not yet well characterized (71). It is not clear how differences between the well-defined exogenous medium and the poorly defined endogenous medium could affect cell behavior. Nonetheless, definition of standards for in vitro study allows exploitation of a cell's sensitivity to changes in medium composition.
One such technique, varying the normal 285–310 mosmol/l (42, 133, 146) osmolarity of the medium by altering the concentration of ionic species (including K+, Ca2+, and Na+, as well as nonionic species such as complex sugars) yields a number of effects. Decreasing the osmolarity exerts swelling pressure on cells; increasing the osmolarity causes the cells to shrink initially (Fig. 2). Swelling induces membrane stretch, which can be employed to manipulate the cytoskeleton (42). This phenomenon has also been used extensively to investigate the active regulation of volume by osteoblasts and other cells via regulation of ion flux, useful in defining and exploiting Ca2+-, K+-, and Na+-regulated signaling pathways (42, 133, 146). The electrical activity resulting from ion flows can be measured using patch-clamp techniques in which microelectrodes are inserted into cells (46). This approach has been used to characterize ion channel activity in bone cells (104, 153).
MECHANICAL MANIPULATION OF CELLS: SUBSTRATE DEFORMATION AND FLUID SHEAR
Application of mechanical stress through gross environmental changes is relatively easy to achieve, but other methods offer better control and provide for more straightforward observation and quantification of cellular response. Traditional in vitro methods typically apply unidirectional, biaxial, radial, and/or bending loads to the culture substrate. In addition, these methods often presume uniform deformation of cells, and allow for quantification of changes in cell physiology such as Ca2+ ion flux (136), regulatory effects of parathyroid hormone (PTH) (22), integrin expression (23), prostaglandin E2 (PGE2) production (108), collagen production (48), insulin-like growth factor (151), and others. However, it is becoming apparent that direct mechanical strain at physiological levels in bone tissue may not stimulate a response in cells (14, 97, 151) and that traditional in vitro methods are not as demonstrative as more physiologically relevant models could be. Even a seemingly simple experimental manipulation may have complex repercussions and result in confounding stimuli. For example, substrate deformation in aqueous medium exposes the attached cell monolayer to an unknown degree of fluid shear (16), limiting the precise control of culture environment and complicating the interpretation of results, leading to ambiguity in determining whether the substrate deformation or the associated fluid flow is the critical stimulus.
Interstitial fluid flow is increasingly implicated as a direct mediator of mechanical forces both throughout bone and to the cells that inhabit bone. Examination of the effects of flow on bone physiology have likewise become of interest to physiologists and biologists. Because of practical difficulties in studying bone fluid flow in situ during normal physiological activity, cell perfusion chambers have been developed to simulate physiological fluid flow (as seen in locations such as vascular endothelium) to observe cellular responses in vitro. In particular, the pressure-driven parallel-plate perfusion chamber design has been implemented (36, 59, 61) and optimized (15, 27, 112) for application of steady, oscillatory, or pulsatile fluid shear stresses to cell monolayers, allowing correlation of fluid shear stresses and nutrient transport with cell activity and adaptation (Fig. 3).
Variations of the parallel-plate flow chamber design have become commonplace tools in cell biology research and provide a basis for current in vitro modeling of many physiological flow regimes. Though developed to simulate better understood biological flows, this technology has been embraced by bone cell experimentalists. Flow regimes believed to be relevant to bone have been studied extensively. Flow chamber parameters are versatile and have allowed investigation of many distinct phenomena, including PGE2 upregulation (69, 105), NO· synthesis (69), Ca2+ flux within monolayers (142), and others. In addition to these early bone cell-related applications (59), fluid shear is currently used to investigate phenomena including mechanical regulation of integrin expression (100), actin cytoskeleton reorganization (25), gap junctional protein expression (132), and the relationship between short-term PGE2 production and mineralization (89) as well as to probe the underlying cellular mechanisms of transduction of fluid shear (19).
This approach has obvious advantages for investigating effects of fluid shear in many biomedical arenas, but it is not known how well these in vitro flow chambers perform, e.g., in achieving a desired stress at the cell level and in emulating physiological flow regimes (4, 73). The type of flow (steady, pulsatile, oscillatory, turbulent, or combinations thereof) plays a significant role in eliciting a cellular response in vitro that may or may not be predictive of cellular response in vivo, but this is often ignored (150). Oscillatory flow might be expected in a physiological setting, but cell monolayers can often be more responsive to flow with continuous components (61). Furthermore, local variations in the flow profile due to uneven topography of the monolayer are generally not accounted for (16), and few studies to date have reported how flow regimes induced through common and commercially available flow chamber designs compare (4). In addition, although these chambers have been used to examine chemokinetic modulation of shear response (5), they provide little control over chemokinetic gradients. However, recent advances in microfabrication techniques (80) could potentially allow incorporation of independently driven microchannels capable of maintaining complex chemical gradients into parallel-plate flow chambers, to provide more a more physiological environment.
FLOW-REGULATED CELLULAR MECHANISMS OF MECHANOTRANSDUCTION
Use of in vitro techniques, and perfusion chambers in particular, has provided much insight into bone physiology when taken in context of the better understood processes of primary and secondary osteogenesis. As osteoblasts lay down new osteoid, many become embedded in the matrix and differentiate into osteocytes that occupy lacunae in bone. As bone forms, individual osteocytes retain communication with one another (as well as with osteoblasts and bone lining cells) via processes occupying canaliculi that extend throughout the bone. This results in a syncytium of spatially remote cells that remain in contact with one another. Fluid movement through the lacunocanalicular pericellular spaces has been postulated as a key mediator of mechanotransduction via direct fluid shear effects as well as the ancillary transport of chemokinetic factors (71), but the active response of the osteocyte has not been thoroughly investigated until recently.
Although the manner in which osteocytes sense fluid flow has not yet been elucidated through use of common in vitro modeling techniques, the osteocyte's interaction with the contents of the pericellular space has been identified as a putative factor for transduction of fluid shear. Ca2+ concentration, important to many bone cell functions, appears to influence fluid movement directly through osteoblast monolayers (55). The glycocalyx, or “cell coat” believed to be composed of hyaluronic acid chains with glycosaminoglycan side chains that anchor the cell to the lacuna wall, plays a significant, but not exclusive, role. In cells exposed to fluid flow, after removal of the hyaluronic acid from the cell surface, prostaglandin expression is reduced whereas Ca2+ flux is not (110). Another mechanism must exist for sensing flow, whether through another unknown component of the glycocalyx or, possibly, via a more direct contact with the membrane itself. Other kinds of membrane deformation in bone cells provoke Ca2+ flux, possibly via cytoskeletal deformation (42, 133, 146), which has been shown to occur rapidly and which may redistribute intracellular forces (53). Fluid shear disruption of integrin/ECM ligand interactions, which causes a dynamic unbinding/binding situation that stimulates production of signaling proteins, has been identified as a mechanotransduction mechanism in the vascular endothelium (62). It may be possible that a similar effect is at work in bone; integrin expression and matrix interaction have already been linked to osteoblast differentiation (7) as well as to Ca2+ channel kinetics of osteoblasts (96) and osteoclasts (33). A new paradigm in which some of these key in vitro experimental concepts are adapted and applied to an in vivo or ex vivo system may provide a crucial next step to decipher mechanotransduction mechanisms in situ through observation of native osteocytes with intact glycocalyx components, other extracellular bindings, and, presumably, different cytoskeletal conformations reacting to shear from flow through the lacunocanalicular system.
Modulation of bone cell responses to fluid shear is better understood than are the underlying mechanisms of bone mechanotransduction. Osteocytes communicate with one another as well as with osteoblasts and bone lining cells via gap junctions (31), characterized by the presence of Cx43 hemichannel subunits (26, 59, 149) capable of transmitting survival signals such as ERK (101) and NO· (86). Communication via Cx43 channels is modulated by fluid shear stresses, particularly in the presence of prostaglandins (64), indicating a clearly defined system by which fluid flow around one osteocyte regulates communication between distant cells. Interestingly, PGE2 production by osteocytes is dependent on the actin cytoskeleton (3) and can alter conformation of the actin fibers (145) in osteocytes, possibly indicating a cytoskeletal basis for a mechanotransduction pathway as has been postulated elsewhere (18, 114, 152). Osteocytes produce transforming growth factor-β (TGF-β), which can be regulated by Cx43 in other cell types (56) and which inhibits bone resorption by osteoclasts (52). This suggests at least one pathway by which fluid flow generated by applied loads can be sensed by cells. The cells, in turn, can regulate processes crucial to bone remodeling in ways that are consistent with fracture repair, functional adaptation, and theories indicating a cellular basis for various pathologies (71). A reduction in local fluid flow stimulus to osteocytes, whether due to inactivity, microgravity effects, damage, or a change in sensory network, can subsequently decrease intercellular communication, which would downregulate the inhibition of osteoclasts, increasing net resorption. Similar effects might also be expected if reduction of fluid flow stemmed nutrient transport to osteocytes, making the osteocytic syncytium less viable and therefore less functional.
POTENTIAL OF THREE-DIMENSIONAL MODELS
Intercellular communication can be examined by two-dimensional planar flow schemes that allow formation of intercellular junctions, but their ability to simulate other phenomena that are highly dependent on the porous microarchitecture of bone, such as flow-related nutrient transport, is necessarily limited. Several new bioreactor technologies are designed to overcome this limitation by employing three-dimensional convective fluid flow strategies. Suspension cultures, in which cells are cultured on microcarrier substrates, can be maintained in a dynamic fluid environment to apply shear stresses and altered nutrient gradients (12, 105). Several methods apply fluid flow through transudant structures. One of the simplest is the spinner flask, in which a scaffold is seeded with cells (Fig. 4) and submerged in culture medium that is continuously stirred by a magnetic bar (138). Though effective for cartilage tissue engineering, this strategy does not appear to provide internal nutrient gradients conducive to colonization of scaffold interiors by osteogenic cells (120). Similar approaches in which the culture vessel wall rotates can yield poor-quality engineered tissues as well (39). One promising technology, the flow perfusion bioreactor, directly applies fluid pressure to a variety of engineered scaffolds. The scaffold construct is constrained within a tightly fit cassette. Fluid is pumped through the top, moves through the entire porous structure, and exits through the bottom. This design effects a more homogenous distribution of nutrients, and osteogenic cell penetration is similarly distributed (6). Although only steady flow regimes have been used to date, the design should allow for potentially more pulsatile and oscillatory effects to be studied as well (Fig. 5).
Another device simulates physiology to a greater degree than the perfusion bioreactor by relying on mechanical loading to induce flow. Tissue scaffolds are analogous to stiff, fluid-filled sponges. Compression of fluid-containing spaces induces flow out of the scaffold, and relaxation draws flow back into the scaffold. Applications of this poroelasticity theory have predicted load-induced flow in bone (9, 101), which has recently been demonstrated experimentally (73). The aforementioned device capitalizes on this concept and utilizes/harnesses load-induced flow as a means for enhanced nutrient transport: cyclical four-point bending is applied to a three-dimensional beam in an aqueous culture environment for extended periods of time. It can accommodate explanted tissue, synthetic resorbable polymer foams, and biopolymer constructs. Intended for use with rigid, porous structures similar to bone, it can potentially be used to study induced penetration of fluid into structures and the enhancement of nutrient transport into structures under various mechanical loading regimes. It has already been shown to alter penetration and long-term viability of osteogenic cells in a scaffold interior, as well as altering their matrix mineralization behavior (118) (Fig. 6).
A number of the aforementioned three-dimensional cell culture approaches lend themselves to application with explanted bone to study ex vivo effects, potentially a powerful strategy for better understanding of bone physiology and biology on multiple scales. Taking this one step further, tissue engineering necessitates the combination of cell culture with complex three-dimensional structures (scaffolds), requiring the further development of new analytical techniques to investigate cellular function in situ. It is expected that as these techniques are developed, they could be adapted readily to study bone per se, either ex vivo or in vivo. Furthermore, new imaging and immunohistochemical tools that have been developed to probe microfluidic movement and transport in bone (71) can be applied to an engineered tissue. Finally, as new fabrication processes provide greater control over the microarchitecture of tissue engineering scaffolds (115, 141), the potential exists to develop new, more accurate in vitro models that 1) incorporate structural elements common to various bone tissue pathologies; 2) provide an experimental test bed to evaluate analytical models that predict structural properties of bonelike mechanochemical transduction (123), hydraulic conductance (88), or streaming potential (47); and 3) integrate more complex fluid sources that imitate flow into and out of vasculature or mimic material aspects of bone, such as piezoelectric effects (49). With the increased availability of technology to observe biophysical phenomena at the cellular level comes an increasing impetus to challenge and, in some cases, discard past assumptions. This should, in turn, improve the understanding of bone mechanobiology and physiology from the cellular to the organ level and vice versa.
Cells removed from their native environments and maintained in vitro by current methods experience culture shock that is poorly understood. Cell types of many different and distinct lineages are incorporated into in vitro osteogenic models and yield conflicting results. Directed differentiation of cells in culture to yield osteoblast-like cells is a commonly used and well-understood technique; native differentiation into, for example, the osteocyte phenotype, which accounts for 90% of cells in healthy bone (72), is less well understood. Accordingly, a majority of in vitro research has focused on the investigation of fluid flow-related mechanobiology of osteoblast-like cells, even though the osteocytic syncytium has been indicated as the primary mechanosensory element in bone. However, new insights into mechanotransductive pathways are paving the way to new experimental methods that more accurately represent the in vivo environment. An evolving understanding of mechanobiological aspects of bone formation and functional adaptation on multiple scales—cell, tissue, and organ—is providing a basis for future research able to advance fundamental science while improving clinical translation.
This work was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant R21 AR-049351-01 (to M. L. Knothe Tate).
We gratefully acknowledge Beth Halasz, MFA, and the Dept. of Medical Illustration, The Cleveland Clinic Foundation, for all illustrations, published with the permission of The Cleveland Clinic Foundation.
- Copyright © 2004 the American Physiological Society