Mari Tsuzaki

IL-1β induces COX2, MMP-1, -3 and -13, ADAMTS-4, IL-1β and IL-6 in human tendon cells

Overuse injuries and trauma in tendon often involve acute or chronic pain and eventual matrix destruction. Anti‐inflammatory drugs have been used as a treatment, however, the cellular and molecular mechanisms of the destructive processes in tendon are not clearly understood. It is thought that an inflammatory event may be involved as an initiating factor. Mediators of the inflammatory response include cytokines released from macrophages and monocytes. Interleukin‐1 beta (IL‐1β) is a candidate proinflammatory cytokine that is active in connective tissues such as bone and cartilage. We hypothesized that tendon cells would express receptors and respond to IL‐1β in an initial „molecular inflammation”︁ cascade, that is, connective tissue cell expression of cytokines that induce matrix destructive enzymes. This cascade results in expression of matrix metalloproteinases (MMPs) and aggrecanases that may lead to matrix destruction. Normal human tendon cells from six patients were isolated, grown to quiescence and treated with human recombinant IL‐1β in serum‐free medium for 16 h. Total RNA was isolated and mRNA expression assessed by semi‐quantitative RT‐PCR. IL‐1β (1 nM) induced mRNAs for cyclooxygenase 2 (COX2), MMP‐1, ‐3, ‐13 and aggrecanase‐1 as well as IL‐1β and IL‐6, whereas mRNAs for COX1 and MMP‐2 were expressed constitutively. The IL‐1β‐treated tendon cells released prostaglandin E2 (PGE2) in the medium, suggesting that the inducible COX2 catalyzed this synthesis. Induction of PGE2 was detectable at 10 pM IL‐1β. IL‐1β also stimulated MMP‐1 and ‐3 protein secretion. Induction of MMP‐1 and ‐3 was detectable at 10 pM IL‐1β. Post‐injury or after some other inciting events, exogenous IL‐1β released upon bleeding or as leakage of local capillaries may drive a proinflammatory response at the connective tissue cell level. The resulting induction of COX2, MMP‐1 and ‐3 may underscore a potential for nonlymphocyte‐mediated cytokine production of MMPs that causes matrix destruction and a loss of tendon biomechanical properties. Endogenous IL‐1β might contribute to the process through a positive feedback loop by stimulating expression and accumulation of MMPs in the tendon matrix.

PDGF-BB, IGF-I and mechanical load stimulate DNA synthesis in avian tendon fibroblasts in vitro

Resident cells in the surface epitenon and internal compartment of flexor tendons are subjected to cyclic mechanical load as muscle contracts to move limbs or digits. Tendons are largely tensile load bearing tissues and are highly matrix intensive with nondividing cells providing maintenance functions. However, when an injury occurs, tendon cells are stimulated to divide by activated endogenous growth factors and those from platelets and plasma. We hypothesize that tendon cells detect mechanical load signals but do not interpret such signals as mitogenic unless an active growth factor is present. We have used an in vitro mechanical load model, application of cyclic strain to cells cultured on flexible bottomed culture plates, to test the hypothesis that tendon cells require platelet-derived growth factor (PDGF-BB) and insulin-like growth factor-I (IGF-I) in addition to mechanical load to stimulate DNA synthesis. In addition, we demonstrate that in avian tendon cells, load and growth factors stimulate phosphorylation of tyrosine residues in multiple proteins, including pp60src, a protein kinase that phosphorylates receptor protein tyrosine kinases. A lack of mitogenic responsiveness to mechanical load alone by tendon cells may be a characteristic of a regulatory pathway that modulates cell division.

Stretch and interleukin‐1β induce matrix metalloproteinases in rabbit tendon cells in vitro

Little is known about the factors that initiate and propagate tendon overuse injuries, but chronic inflammation and matrix destruction have been implicated. The purpose of this study was to evaluate the production of cyclooxygenase II (COX‐2) and matrix metalloproteinases (MMPs) by tendon cells exposed to cyclic strain and inflammatory cytokines in vitro. Rabbit Achilles tendon cells were subjected to a stretching protocol with 5% elongation at 0.33 Hz for 6 h, or treated with 1000 pM interleukin‐1β (IL‐1β), or exposed to IL‐1β and stretching together. Gene expression was evaluated by RT‐PCR and production of stromelysin was quantified with an ELISA. IL‐1β induced the expression of the collagenase‐1 and stromelysin‐1 genes. Production of stromelysin proenzyme by cells stimulated with IL‐1β was 17 times higher than production by control cells. Cells exposed to IL‐1β and stretching produced 20 times more stromelysin than control cells. Cells subjected to stretching alone did not produce more stromelysin than control cells. The synergistic effect of IL‐1β and stretching was observed at doses of IL‐1β ranging from 10 to 1000 pM. These data suggest that mechanical load and inflammatory cytokines can initiate a matrix destructive pathway in tendon that is more pronounced than with mechanical loading or inflammation alone.

ATP modulates load-inducible IL-1β, COX 2, and MMP-3 gene expression in human tendon cells

Tendon cells receive mechanical signals from the load bearing matrices. The response to mechanical stimulation is crucial for tendon function. However, overloading tendon cells may deteriorate extracellular matrix integrity by activating intrinsic factors such as matrix metalloproteinases (MMPs) that trigger matrix destruction. We hypothesized that mechanical loading might induce interleukin‐1beta (IL‐1β) in tendon cells, which can induce MMPs, and that extracellular ATP might inhibit the load‐inducible gene expression. Human tendon cells isolated from flexor digitorum profundus tendons (FDPs) of four patients were made quiescent and treated with ATP (10 or 100 μM) for 5 min, then stretched equibiaxially (1 Hz, 3.5% elongation) for 2 h followed by an 18‐h‐rest period. Stretching induced IL‐1β, cyclooxygenase 2 (COX 2), and MMP‐3 genes but not MMP‐1. ATP reduced the load‐inducible gene expression but had no effect alone. A medium change caused tendon cells to secrete ATP into the medium, as did exogenous UTP. The data demonstrate that mechanical loading induces ATP release in tendon cells and stimulates expression of IL‐1β, COX 2, and MMP‐3. Load‐induced endogenous IL‐1β may trigger matrix remodeling or a more destructive pathway(s) involving IL‐1β, COX 2, and MMP‐3. Concomitant autocrine and paracrine release of ATP may serve as a negative feedback mechanism to limit activation of such an injurious pathway. Attenuation or failure of this negative feedback mechanism may result in the progression to tendinosis.

Gap junctions regulate responses of tendon cells ex vivo to mechanical loading.

Avian digital flexor tendons were used with a device to apply load ex vivo to study the effects on deoxyribonucleic acid and collagen synthesis when cell to cell communication is blocked. Flexor digitorum profundus tendons from the middle toe of 52-day-old White Leghorn chickens were excised and used as nonloaded controls, or clamped in the jaws of a displacement controlled tissue loading device and mechanically loaded for 3 days at a nominal 0.65% elongation at 1 Hz for 8 hours per day with 16 hours rest. Tendon samples were radiolabeled during the last 16 hours with 3H-thymidine to monitor deoxyribonucleic acid synthesis or with 3H-proline to radiolabel newly synthesized collagen. Cyclic loading of whole avian flexor tendons stimulated deoxyribonucleic acid and collagen synthesis, which could be blocked with octanol, a reversible gap junction blocker. Cells from human digital flexor tendon were used to populate a rectangular, three-dimensional, porous, polyester foam that could be deformed cyclically in vitro. Together, these results support the hypothesis that tendon cells must communicate to sustain growth and matrix expression and that an engineered three-dimensional construct can be used to study responses to mechanical load in vitro.

Rabbit tendon cells produce MMP-3 in response to fluid flow without significant calcium transients

Forces applied to tendon during movement cause cellular deformation, as well as fluid movement. The goal of this study was to test the hypothesis that rabbit tendon fibroblasts detect and respond to fluid-induced shear stress. Cells were isolated from the paratenon of the rabbit Achilles tendon and then subjected to fluid flow at 1 dyn/cm(2) for 6h in a specially designed multi-slide flow device. The application of fluid flow led to an increased expression of the collagenase-1 (MMP-1), stromelysin-1 (MMP-3), cyclooxygenase II (COX-2) and interleukin-1beta (IL-1beta) genes. The release of proMMP-3 into the medium exhibited a dose-response with the level of fluid shear stress. However, not all cells aligned in the direction of flow. In other experiments, the same cells were incubated with the calcium-reactive dye FURA-2 AM, then subjected to laminar fluid flow in a parallel plate flow chamber. The cells did not significantly increase intracellular calcium concentration when exposed to fluid shear stress levels of up to 25 dyn/cm(2). These results show that gene expression in rabbit tendon cells is sensitive to fluid flow, but that signal transduction is not dependent on intracellular calcium transients. The upregulation of the MMP-1, MMP-3 and COX-2 genes shows that fluid flow could be an important mechanical stimulus for tendon remodelling or injury.

IL-1β sensitizes intervertebral disc annulus cells to fluid-induced shear stress

Chronic inflammation and altered mechanical loading are implicated as contributors to intervertebral disc degeneration. Biomechanical and biochemical factors play a role in disc degeneration but have received limited study. Mechanically, intervertebral discs are sheared during bending or twisting of the trunk. Biochemically, IL‐1β, detected in degenerative discs, promotes metalloproteinase expression. We hypothesized that disc cells might respond to shear stress and IL‐1β in a calcium signaling response. We measured the effect of single and combined stimuli on intracellular calcium concentration ([Ca2+]ic) and signaling. Cells were isolated from annulus tissue, cultured to quiescence, plated on collagen‐bonded Culture Slips® and incubated with Fura‐2AM. Cells then were incubated in IL‐1β. Cell response to the effects of fluid flow was tested using FlexFlo™, a laminar flow device. Human annulus (hAN) cells responded to laminar fluid flow with a one to three‐fold increase in [Ca2+]ic. IL‐1β alone produced a small, transient stimulation. hAN cells pretreated with IL‐1β responded to shear with a more dramatic and sustained increase in [Ca2+]ic, six to ten‐fold over basal level, when compared to shear then IL‐1β or shear and IL‐1β alone (P < 0.001 for all comparisons). This is the first study documenting synergism of a signaling response to biomechanical and biochemical stimuli in human disc cells. IL‐1β treatment appeared to “sensitize” annulus cells to mechanical load. This increased responsiveness to mechanical load in the face of inflammatory cytokines may imply that the sensitivity of annulus cells to shear increases during inflammation and may affect initiation and progression of disc degeneration. J. Cell. Biochem. 82: 290–298, 2001.

Vibratory loading decreases extracellular matrix and matrix metalloproteinase gene expression in rabbit annulus cells.

BACKGROUND CONTEXT Whole body vibration is an important factor contributing to low back and radicular pain. Vibratory loading as a mechanical stimulus is transferred to connective tissues as energy from ground reaction forces, as well as a direct input from the use of motorized tools and vehicles. Extracellular matrix degradation parallels increased age and mechanical stimuli resulting in disc degeneration and eventual spinal deformity. PURPOSE The objective of this study was to investigate the relationship between vibratory loading and extracellular matrix expression in cultured rabbit annulus cells. STUDY DESIGN/SETTING An in vitro rabbit model using cultured annulus fibrosis cells isolated from normal intervertebral disc was used to study matrix and metalloproteinase expression in response to vibration. METHODS Annulus fibrosis cells were isolated by collagenase digestion from New Zealand White rabbits. Vibratory stimulation was applied to annulus cells in vitro, using an oscillating platform to deliver 0.1 x gravity load at 6 Hz for 2, 4, 6 or 8 hours. Gene expression was assessed by reverse transcriptase polymerase chain reaction. RESULTS Aggrecan, collagen Type III and matrix metalloproteinase-3 gene expression was suppressed by vibratory loading in rabbit annulus cells. Suppression of the aggrecan gene might lead to a decrease in proteoglycan synthesis. CONCLUSIONS These data suggest that vibratory load may play an important role in extracellular matrix metabolism of intervertebral disc cells, especially in the gene expression pathway of proteoglycans. It has been proposed that vibratory loading increases production of matrix-degrading, proteolytic enzymes. We have demonstrated that gene expression for key matrix messages and matrix metalloproteinase is decreased by vibration. In conclusion, we believe that study of the roles between extracellular matrix gene suppression and mechanical stress may clarify the pathomechanism of disc degeneration, such as disc herniation or degeneration.

Tendon collagens: Extracellular matrix composition in shear stress and tensile components of flexor tendons

Outer synovial tissues were separated from the remainder of avian flexor tendon and the collagens characterized biochemically and compared with those of the internal portion of tendon and sheath. The collagen content of tendon synovium was 23%, whereas that of tendon and sheath were 78% and 73%, respectively, based on dry weight. Four genetic types of collagen were found in the pepsin solubilized matrices: in the synovium, types I (78%) and III (19%) predominated; types V and possibly VI were present as minor collagens. Purified synovial type V collagen was a heterotrimer, with chain composition [alpha 1(V)]2 alpha 2(V). In contrast, the internal portion of tendon and sheath were comprised of only type I collagen. There was a large amount (41%) of ethanol extractable, noncollagenous material present in synovium, a part of which was proteoglycans. In addition, collagen cross-links of these tissues were quantified: the internal tendon had an abundant concentration of pyridinoline; synovium exhibited high amounts of labile, reducible cross-links, particularly dihydroxylysinonorleucine. In the case of sheath, lysine aldehyde-derived cross-links appeared to be predominant. These results indicate that each tissue has its own collagen type distribution as well as cross-linking pattern reflecting their maturational and functional differences.

Annulus cells release ATP in response to vibratory loading in vitro

Mechanical forces regulate the developmental path and phenotype of a variety of tissues and cultured cells. Vibratory loading as a mechanical stimulus occurs in connective tissues due to energy returned from ground reaction forces, as well as a mechanical input from use of motorized tools and vehicles. Structures in the spine may be particularly at risk when exposed to destructive vibratory stimuli. Cells from many tissues respond to mechanical stimuli, such as fluid flow, by increasing intracellular calcium concentration ([Ca2+]ic) and releasing adenosine 5′‐triphosphate (ATP), extracellularly, as a mediator to activate signaling pathways. Therefore, we examined whether ATP is released from rabbit (rAN) and human (hAN) intervertebral disc annulus cells in response to vibratory loading. ATP release from annulus cells by vibratory stimulation as well as in control cells was quantitated using a firefly luciferin‐luciferase assay. Cultured hAN and rAN cells had a basal level of extracellular ATP ([ATP]ec) in the range of 1–1.5 nM. Vibratory loading of hAN cells stimulated ATP release, reaching a net maximum [ATP] within 10 min of continuous vibration, and shortly thereafter, [ATP] declined and returned to below baseline level. [ATP] in the supernatant fluid of hAN cells was significantly reduced compared to the control level when the cells received vibration for longer than 15 min. In rAN cells, [ATP] was increased in response to vibratory loading, attaining a level significantly greater than that of the control after 30 min of continuous vibration. Results of the current study show that resting annulus cells secrete ATP and maintain a basal [ATP]ec. Annulus cells may use this nucleotide as a signaling messenger in an autocrine/paracrine fashion in response to vibratory loading. Rapid degradation of ATP to ADP may alternatively modulate cellular responses. It is hypothesized that exposure to repetitive, complex vibration regimens may activate signaling pathways that regulate matrix destruction in the disc. As in tendon cells, ATP may block subsequent responses to load and modulate the vibration response. Rabbit annulus cells were used as a readily obtainable source of cells in development of an animal model for testing effects of vibration on the disc. Human cells obtained from discarded surgical specimens were used to correlate responses of animal to human cells.