Joint and Muscle Dysfunction of the Temporomandibular Joint
Guest Editor
Arthur W. English, Atlanta, Ga.
42 figures, 4 in color, 2 tables, 2003
Basel 폷 Freiburg 폷 Paris 폷 London 폷 New York 폷 Bangalore 폷 Bangkok 폷 Singapore 폷 Tokyo 폷 Sydney
S. Karger Medical and Scientific Publishers Basel 폷 Freiburg 폷 Paris 폷 London New York 폷 Bangalore 폷 Bangkok Singapore 폷 Tokyo 폷 Sydney
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2003 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–7598–X
Vol. 174, No. 1–2, 2003
Contents
5 Preface English, A.W. (Atlanta, Ga.)
6 ‘Friction and Adhesive Forces’ – Possible Underlying Causes for
Temporomandibular Joint Internal Derangement Nitzan, D.W. (Jerusalem) 17 Subchondral Bone Resorption in Temporomandibular Joint Disorders Gruber, H.E. (Charlotte, N.C.); Gregg, J. (Blacksburg,Va.) 26 Nitric Oxide in Experimental Joint Inflammation. Benefit or Detriment? Wahl, S.M.; McCartney-Francis, N.; Chan, J.; Dionne, R.; Ta, L. (Bethesda, Md.); Orenstein, J.M. (Washington, D.C.) 34 Pathophysiological Mechanisms in Osteoarthritis Lead to Novel
Therapeutic Strategies Malemud, C.J.; Islam, N.; Haqqi, T.M. (Cleveland, Ohio) 49 Analgesic Effect of Elastoviscous Hyaluronan Solutions and the
Treatment of Arthritic Pain Balazs, E.A. (Ridgefield, N.J.) 63 Induction of Early Growth Response Gene Egr2 by Basic Calcium
Phosphate Crystals through a Calcium-Dependent Protein Kinase CIndependent p44/42 Mitogen-Activated Protein Kinase Pathway Zeng, X.R.; Sun, Y.; Wenger, L. (Miami, Fla.); Cheung, H.S. (Miami, Fla./ Coral Gables, Fla.) 73 Specialized Cranial Muscles: How Different Are They from Limb and
Abdominal Muscles? Sciote, J.J.; Horton, M.J. (Pittsburgh, Pa.); Rowlerson, A.M. (London); Link, J. (Pittsburgh, Pa.) 87 Sex Differences in Rabbit Masseter Muscle Function English, A.W. (Atlanta, Ga.); Widmer, C.G. (Gainesville, Fla.)
97 Author Index Vol. 174, No. 1–2, 2003 98 Subject Index Vol. 174, No. 1–2, 2003 99 Conference Calendar
© 2003 S. Karger AG, Basel Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/cto_issues
Cells Tissues Organs 2003;174:5 DOI: 10.1159/000070569
Preface
Temporomandibular (jaw) joint diseases and disorders are estimated to affect 10 million Americans every year with the majority of those seeking treatment being women in their childbearing years. These conditions commonly referred to as ‘TMJ’ represent a family of complex and poorly understood health problems manifested by pain in and around the jaw and associated muscles and limitations in the ability to make to normal movements of speech, facial expression, eating, chewing, and swallowing. Conditions that routinely affect other joints in the body, such as arthritis and trauma, also affect the TMJ. Researchers generally agree that the most common temporomandibular diseases and disorders fall into three main categories: arthritis, internal derangement of the joint, and/or myofascial pain. Currently there are no established treatments for TMJ conditions, so that the therapies patients receive generally reflect the personal views and practices of the provider, which, all too often, can lead to a worsening of their condition. The TMJ Association (TMJA) is a national nonprofit organization dedicated to advancing understanding and treatment of temporomandibular diseases and disorders. In its work to promote research and improve diagnosis, treatment, and prevention, the association has held two scientific meetings, aimed in part at broadening the base
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0005$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
of investigators studying TMJ problems. One result of the first meeting in May 2000 was the publication of a series of papers in Cells Tissues Organs (Vol. 169, No. 3, ISBN 3–8055–7229–8). That first meeting, called Moving TMJ Research into the 21st Century, provided a broad overview of TMJ disorders and elicited a number of fresh ideas and new perspectives. It was clear, however, that no single scientific meeting on TMJ disorders could encompass the many basic and clinical disciplines needed to advance the field. The organizers agreed that future meetings should focus on selected aspects of TMJ disorders, while retaining a commitment to an integrated, broadbased, interdisciplinary approach. Accordingly, the second meeting of the association, held in May 2002, was focused on the role that osteoarthritis and muscle mechanics play in TMJ diseases and disorders. This special topics issue of Cells Tissues Organs is a selection of papers presented at the second meeting. It is titled: Joint and Muscle Dysfunction of the Temporomandibular Joint. We are delighted to be able to publish these fine papers and hope that their circulation will continue to stimulate TMJ clinical and basic research as was the case with the publication of the earlier set of papers. Arthur W. English, Atlanta, Ga.
Cells Tissues Organs 2003;174:6–16 DOI: 10.1159/000070570
‘Friction and Adhesive Forces’ – Possible Underlying Causes for Temporomandibular Joint Internal Derangement D.W. Nitzan Department of Oral and Maxillofacial Surgery, The Hebrew University Hadassah School of Dental Medicine, Jerusalem, Israel
Key Words Friction W Adhesive forces W Temporomandibular joint W Internal derangement W Open lock W Anchored disc phenomenon W Osteoarthritis
Abstract Since normal temporomandibular joint (TMJ) movements depend primarily on the disc freely sliding down the slope of the eminence, understanding how aberrations in the lubrication system contribute to TMJ dysfunction is clearly critically important. It provides a possible explanation for the genesis of disc displacement and helps make us familiar with the clinical appearance and ways of treating limited mouth opening caused by the anchored disc phenomenon (ADP) versus disc displacement without reduction, TMJ open lock versus dislocation and osteoarthritis. This understanding clarifies the efficiency of procedures such as joint hydraulic pump, arthrocentesis and arthroscopic lavage and lysis particularly in ADP, open lock and osteoarthritis. Copyright © 2003 S. Karger AG, Basel
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0006$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
Introduction
Temporomandibular disorder refers to various dental and medical conditions affecting the temporomandibular joint (TMJ), masticatory muscles and contiguous tissue components. The wide use of temporomandibular disorder by clinicians and researchers avoiding a precise diagnosis has led to a great deal of meaningless research and indiscriminate treatment. The common phenomenon of multiple TMJ surgery clearly suggests a lack of basic understanding of TMJ internal derangement. With the aim of reducing iatrogenic TMJ insults, an increasing number of studies are directed toward the intra-articular, biomechanical and biochemical events underlying TMJ internal derangement.
Abbreviations used in this paper
ADP HA PL PLA2 SAPLs SF TMJ
anchored disc phenomenon hyaluronic acid phospholipid phospholipase A2 surface-active phospholipids synovial fluid temporomandibular joint
D.W. Nitzan Department of Oral and Maxillofacial Surgery Faculty of Dental Medicine, PO 12272 Jerusalem 91120 (Israel) Tel. +972 2 6776148, Fax +972 51 874537, E-Mail
[email protected]
Displacement of the articular disc in the TMJ from its normal position was described and identified as a potential clinical problem over 100 years ago [1]. In the 1950s, Ireland [1] pointed to this as the first stage in the sequence of events leading to osteoarthritis. Disc displacement is the postulated cause of joint pain, limited mandibular movement, joint sounds and osteoarthrotic changes in the TMJ [3–13]. Although the damage caused by disc displacement has been given much attention, very little effort has been made to explain the factor(s) behind this phenomenon [3–7]. While surgical procedures to restore normal TMJ anatomy and function [14, 15] which ignore the causes of disc displacement have led to ostensibly successful outcomes, there have been many cases of multiple surgical interventions, sometimes involving severe complications [16]. Such iatrogenic insults have stimulated rethinking of the role played by disc displacement in TMJ complaints [17, 18]. Disc displacement was diagnosed in normal individuals. It is often not associated with joint pain, and diagnosing the origin of pain is important for developing a proper treatment plan [19, 20]. In addition, lavage of the upper joint compartment using arthroscopy [21] or arthrocentesis [22], neither of which change the disc position [23, 24], were found to markedly improve function and alleviate pain in disorders such as severe closed lock. As a result, the focus has shifted from studies on the ‘position of the disc’ to the search for the intra-articular biomechanical and biochemical events underlying pain and disc displacement. This review considers the TMJ from a different angle. Since normal TMJ movements depend primarily on free sliding of the disc down the slope of the eminence, understanding how aberrations in the lubrication system contribute to TMJ internal derangement is clearly critically important.
There is abundant literature on lubrication of large synovial joints. Various biomechanical, biochemical and animal studies suggest that two major constituents are responsible for free joint movement: surface-active phospholipids (SAPLs) and hyaluronic acid (HA). SAPLs, the major boundary lubricants protecting articular surfaces, are highly effective [25–27]. They consist of a polar component bound to the articular surface, thereby orienting their nonpolar moieties outward. The latter form a hydrophobic surface with a relatively low
surface energy, which is much less conducive to friction than exposed articular surfaces (without SAPL) [25]. Good cohesion, an essential factor in load bearing, is provided by hydrogen bonding among phospholipid (PL) molecules [25–27]. Under high load, SAPL may attain very low friction coefficients. Electron microscopy of sheep knee joint articular surfaces, using a special fixation technique (essentially substituting tannic acid for glutaraldehyde), revealed multilamellar structures, 200–300 nm in diameter, typical of SAPL [25, 27]. Schwarz and Hills [27] showed that lamellar bodies, which are present in type B synoviocytes, are the probable source of surfactant. Lubricin and proteolipid, which have been isolated from the synovial fluid (SF), seem to facilitate SAPL deposition, providing efficient boundary lubrication at articular surfaces [28]. Various biomechanical studies suggest that PLs play a role in joint lubrication [29–32]. Biochemical analysis of articular surfaces and SF, revealing the presence of PLs, primarily phosphatidylcholine, support the EM findings. Traces of phosphatidylethanolamine and sphingomyelin have also been found [25, 33–36]. To further confirm their proposed lubrication role, the friction in the joint was determined following SAPL removal from the SF. Thus, lipid solvent applied to the joint articular surface increased the friction by 150% [25]. Potent lipid solvents were shown to adsorb SAPL, thereby reducing hydrophobicity [26]. Addition to SF of phospholipase A2 (PLA2, an enzyme secreted into the SF by the synoviocytes, chondrocytes and osteoblasts), which acts specifically on PLs, significantly increased friction [37]. As opposed to neutral lipids, Rabinowits et al. [35] and Punzi et al. [36] reported a dramatic decrease in PL levels, concomitant with increased PLA2 activity, in the traumatized joint. These findings provide further strong support for SAPLs as active elements in the joint lubrication system, contributing to protection of articular surfaces. Of special interest is the role assigned to HA in the current view of the joint lubrication system. A high-molecular-weight mucopolysaccharide, HA, forms a ‘full fluid film’, which keeps the articular surfaces apart and prevents generation of friction. An in vitro study revealed adherence of HA to PL membranes (liposomes), thereby protecting it from lysis by PLA2 [38]. Thus, HA might play an important indirect role in joint lubrication by adhering to SAPLs, which are thereby protected against uncontrolled lysis by PLA2 in the SF [38]. A similar defense mechanism has been described at the cellular level [39].
‘Friction and Adhesive Forces’
Cells Tissues Organs 2003;174:6–16
Synovial Joint Lubrication
7
TMJ Lubrication
An efficient lubrication system in the TMJ is mandatory for the disc to slide down the slope of the eminence. Lipids, in general, and PLs, in particular, have not been considered in most studies dealing with TMJ articular surface and SF composition [40–43]. Marchetti et al. [44] identified an electron-dense layer which was not discernible in disordered joints. They postulated that this covering layer maintains proper joint function and prevents conditions conducive to adherence [44]. Osmiophilic layers with embedded vesicular structures have been demonstrated in the TMJ, which are similar to the ‘surface coat’ described by others [45–47]. The conventional fixation technique for TEM showed the presence of PL multilamellar structures with diameters of up to 290 nm, similar to those reported by Hills et al. [25]. The osmiophilic droplet cluster in centrifuged SF, as described previously, is degraded following exposure to PLA2, which specifically acts on PL. Biochemical studies on the SF of normally functioning TMJs revealed the presence of neutral lipids, such as triglycerides, cholesterol and cholesterol esters. Among the polar lipids, phosphatidylcholine was dominant with much lower levels of phosphatidylethanolamine and sphingomyelin [48, 49]. Further in-depth studies on the lubrication system might eventually lead to the development of appropriate tools for the prevention and treatment of TMJ disorders as well as improved intra-articular injection materials.
Joint Overloading and Impaired Lubrication
Joint function remains normal as long as its adaptive capacity is not compromised [11, 49–56]. Parafunction such as clenching is a good example of repetitive jaw motion associated with possible high TMJ impact loading that leads to the conversion of shearing stresses into compressive stresses [52, 54]. Overloading is the major cause of collapse of the lubrication system, which has various effects on synovial joints [51, 57–59], e.g. blood supply interruption [60–63]. When overloaded, the joint intraarticular pressure increases and, if it exceeds the capillary perfusion pressure, causes temporary hypoxia, which is compensated for by reoxygenation on cessation of overloading. This hypoxic-reperfusion cycle has been reported to evoke nonenzymatic release of ROS (i.e., superoxide and hydroxyl anions) [60–63]. The highly reactive ROS may enter into rapid chemical reactions in various tissues
8
Cells Tissues Organs 2003;174:6–16
or destroy important molecules. Among other effects of ROS on the joints are the inhibition of HA biosynthesis and degradation, giving rise to a marked decrease in SF viscosity [63]. As previously mentioned, reduced viscosity does not affect the SF lubrication efficiency [64, 65]. However, as demonstrated in an in vitro study, the degraded form of HA has an indirect effect on joint lubrication: by not inhibiting PLA2 activity, it fails to protect continuity of the SAPL layer [38]. Thus, PLA2 secreted into the SF is free to lyse SAPL [38, 39], disrupting the continuity of the boundary surfactant layer on the articular surfaces. Notwithstanding the accumulating body of knowledge on the mechanism of joint lubrication and its reciprocal relation to joint loading, whether and how the absence of SAPL affects the sliding ability of articular surfaces deserves further discussion.
Joint Performance in the Absence of Lubrication
The main factors determining the friction between uncovered articular surfaces are relative contact area, surface energy, shear and surface rupture potential, surface elasticity and applied load [57, 58, 66–70]. The articular surfaces are smooth planes [71], implying large contact areas which, together with the surface elasticity (such as at the disc) and the high surface energy [57, 58, 66–68], lead to increased friction in the absence of lubrication. It seems reasonable to assume that increased friction eventually leads to shear and rupture of articular surfaces, especially the softer planes [48, 49, 57, 72, 73]. There may also be ‘adhesion forces’ between the flexible disc and the fossa. If the fluid film is extremely thin (subboundary lubrication) between the naked surfaces, the resulting high adhesion forces may cause anchoring of the disc to the fossa. Such anchorage can be promptly released by lavage of the adhered surfaces.
Control of TMJ Overloading
Very little is known about the actual forces involved in TMJ function in situ, but weights of up to 17.7 kg have been recorded in Macaca monkeys [55, 56]. Chen and Xu [56] found contact stress in human TMJ reached a maximum of 1,650 lb/inch2, which is similar to that in the hip and knee joints. In weanling pigs, intra-articular pressure in the superior TMJ compartment was as high as 20 mm
Nitzan
Hg during masticatory movement [74]. In awake humans, intra-articular pressure ranging from 8 to 200 mm Hg (mean 63.9 B 52.25 mm Hg) has been documented during clenching [52]. Pressure of over 40 mm Hg, which may occur in clenching, exceeds the capillary perfusion pressure. This may elicit temporary hypoxia, followed by reoxygenation on cessation of the activity [60–63]. Rest, which is beneficial for other symptomatic synovial joints, is not indicated for the TMJ, since parafunction is essentially an uncontrollable behavior occurring during the night. For example, an interocclusal appliance along the lines of a lever III system, which reduces the force on the TMJ, is a most efficient and immediate way to control the forces generated in the TMJ [52]. Indeed, such appliances can reduce the pressure generated in the joint during clenching by 81.2% [52]. This is not unlike the effect of different bandaging techniques as an adjunct to orthopedic management.
TMJ Arthrocentesis
Arthrocentesis, a treatment modality for various TMJ disorders [75–77], consisting of saline injection and lavage of the upper compartment in arthrocentesis, forces apart the disc from the fossa and then washes away degraded particles and inflammatory components, including ROS and PLA2. An arthrocentesis outflow of 100 ml is recommended, since with smaller volumes, denatured hemoglobin and various proteinases have been recovered in various fractions [78, 79]. TMJ unloading combined with residual adhered PL and HA may become a milestone in rehabilitating the joint lubrication system. Further studies along these lines may lead to more effective intra-articular medication and improve outcomes.
Clinical Applications
Better understanding of the TMJ lubrication system and its possible collapse provides new insights into joint disorders, such as disc displacement, severe closed lock, open lock and degenerative joint disease. TMJ Disc Displacement The most common TMJ signs and symptoms, such as joint clicking, closed lock and crepitation, are associated with disc displacement. The etiologies for this condition as mentioned in the literature are various types of acute trauma, functional overloading, joint laxity, degenerative
‘Friction and Adhesive Forces’
joint disease, masticatory muscle spasm and increased friction between moving parts. To the current general consensus, TMJ disc displacement results from its inability to slide smoothly due to increased friction or degenerative changes in the joint surfaces [50, 51, 80–92]. The sequence of events, starting with increased friction in the upper joint compartment and culminating in disc displacement, has been described [93] and is best visualized schematically (fig. 1a–e). Activation of various parafunctions, such as clenching (fig. 1a), compromises the lubrication system in the upper TMJ compartment. The resulting increased friction prevents the disc from sliding together with the condyle. On jaw opening, the condyle is pulled away from the disc by the inferior head of the lateral pterygoid muscle (fig. 1b). As a result, the ligaments joining the disc to the condyle are gradually stretched (fig. 1c), and the ‘mobilized’ disc gravitates slightly downward and forward. Subsequently, on clenching, the unstable disc is propelled forward by pressure from the condyle. At this point, the force on the slightly displaced disc is shared between two vectors, one of which is directed forward [Mow, pers. commun.]. Apparently, on mouth closure, the superior belly of the lateral pterygoid muscle pulls the disc anteriorly (fig. 1d). Subsequently, during mouth opening, the condyle, which is now posterior to the loose disc, gradually pushes it down the slope of the eminence, displacing it further forward (fig. 1e). Since the lateral articular disc bears the bulk of the shearing and compressive loads, persistent loading tends to drive it in a medial direction, which is the ‘path of least resistance’ [Mohl, pers. commun.]. The arthrographic visualization of TMJs with disc displacement exhibits an intriguing feature. The anterior wall of the inferior compartment, which should be stretched by the displacement of the disc, is markedly distended and situated in front of the condyle under the disc (fig. 2). This can be explained in terms of the proposed underlying pathogenesis of disc displacement. Thus, the increased friction between the disc and the fossa leads to malformation of the lower compartment (see above). Namely, on mouth opening, the condyle is pulled forward, away from the disc, which moves more slowly, while stretching the anterior ‘wall’ of the inferior compartment (fig. 1b). In the course of time, the anterior wall of the inferior compartment gradually becomes distended, as seen in the arthrographic image in figure 2. This typical feature of disc displacement with and without reduction indirectly illustrates the role of friction in loosening and displacing the link between the disc and condyle [91].
Cells Tissues Organs 2003;174:6–16
9
Fig. 1. The process of disc displacement in the TMJ. a During joint overloading (e.g., clenching; arrowhead), the disc is pressed against the fossa. The intermittently increasing intra-articular pressure may generate reactive oxidative species. b Collapse of the lubrication system in the presence of ROS produces friction between the disc and the fossa (arrowhead). On mouth opening the condyle is pulled away from the slow-moving disc (arrow), and the condyle-disc connection slackens gradually. c The disc is loosely connected to the condyle,
10
Cells Tissues Organs 2003;174:6–16
inclining forward and downward along the slope of the eminence. d On clenching, the loosened disc is propelled forward by the pressing condyle (arrowhead). At the same time, the disc is pulled forward by the superior belly of the lateral pterygoid muscle. e On mouth opening, the condyle, which at this time is located posterior to the disc, is pulled by the inferior head of the lateral pterygoid muscle (arrow), pushing the disc into an anterior-medial-inferior direction (white arrow).
Nitzan
by disc displacement without reduction, history of clicking is obligatory and arthrocentesis has only limited effect (unpubl. data). It is assumed that joint overloading may lead to generation of ROS [97] that are responsible for the collapse of the lubrication system. Apparently, even partial collapse of the lubrication system might generate strong adhesive forces between the denuded, smooth, elastic disc and the fossa [72, 98]. In such a situation, even a small contact area between the two opposing surfaces would bring about a sudden inability of the disc to slide down the slope of the eminence or ADP (fig. 3a, b). Forced opening is not recommended as it may lead to shear and rupture of the articulating surfaces. Arthrocentesis neutralizes the adhesion forces and separates the flexible disc from the rigid surface of the eminence [75, 92–98]. The use of a jointunloading interocclusal appliance at night [52] enables the instant production of HA and PLs, thereby restoring joint lubrication. Physiotherapy, which is not indicated while the disc is ‘stuck’, can now be used intensively to rehabilitate joint movement. Recurrence is rare, probably since the meeting of two bare articular surfaces, which can be pressed together to the point of adherence, is infrequent [98]. Fig. 2. TMJ arthrography with contrast material introduced into the inferior and superior joint spaces: TMJ disc displacement in closed mouth position. There is marked distention of the inferior joint compartment in front of the condyle and under the displaced disc. This until now unexplained phenomenon can be rationalized by the pull of the head of the condyle against the anterior wall of the inferior compartment, away from the lagging disc, which rubs against the fossa as demonstrated in figure 1b.
Anchored Disc Phenomenon [92–94] ADP is characterized by sudden severely limited mouth opening (ranging between 10 and 30 mm), with deviation of the mandibular midline toward the affected side. A history of clicking is not obligatory. While there is usually no pain in the TMJ or adjacent muscles, forced mouth opening evokes pain in the affected joint. In plain open mouth radiographs and computerized tomography scans, the TMJ shows evidence of a nonsliding rotated condyle. In MRI, on the other hand, the disc seems to be ‘stuck’ to the fossa and the condyle slides under it [95]. Lavage by means of arthrocentesis, the treatment of choice, provides immediate relief with rare recurrence in long-term follow-up [75, 91–96]. As opposed to ADP in cases of limited mouth opening caused
‘Friction and Adhesive Forces’
TMJ Open Lock The sudden inability to close the mouth characterizes ‘open lock’ [99, 100]. There is usually no pain in the affected TMJ or adjacent muscles. In plain radiographs and computerized tomography scans of ‘open lock’, the condyle is located under the eminence (and not in front of it, as would be expected in condylar dislocation). MRI shows the condyle as being ‘stuck’ in front of the disc (fig. 4). Joint lavage by arthrocentesis provides relief with rare recurrence on long-term follow-up [99, 100]. As with ADP, the etiology of open lock is related to friction between the disc and the fossa. The disc, which normally moves together with the condyle, lags behind it. As a result, the condyle slides under and in front of the disc and is unable to return to its former position in the fossa, and the mouth remains open. Following lavage the lubrication system is restored, enabling the disc and the condyle to slide together, thereby preventing it from moving in front of the disc [100]. TMJ Osteoarthritis In the acute phase of osteoarthritis, patients complain of early morning stiffness in the TMJ, severe joint pain, both in the jaw at rest and in movement in all directions, limited mouth opening, and difficulty in yawning, biting
Cells Tissues Organs 2003;174:6–16
11
Fig. 3. TMJ lubrication system and its partial collapse. a Normally functioning TMJ:
SAPLs cover the fossa and disc articular surfaces, each molecule (round shape) having a hydrophobic end that is directed toward the joint space. The hydrophilic end is attached to the articular surface. The molecules are connected to each other by hydrogen bonds. The high-molecular-weight HA (cloud shape) is attached to the PL as fluid film. PLA2 is present in the joint (star shape). The HA protects the PL from being attacked by PLA2. b Following excessive overloading of the joint, free radicals may be produced, thereby degrading the HA. In its degraded form HA does not inhibit PLA2 activity, thus enabling lysis of the PL layer. On lysis of the PL, the articular surfaces are stripped of their lubricants, exposing them to unwarranted effects. Because these surfaces are smooth and elastic, and possess a high surface energy, in the presence of subboundary lubrication ‘adhesive forces’ may be generated between the ‘naked’ surfaces of the disc and fossa with subsequent disc anchorage to the fossa.
and chewing. Sometimes the symptoms are accompanied by a sensation of swelling in the TMJ area [101, 102]. On physical examination, attempts to open the mouth or to move the jaw in lateral and protrusive direction beyond the limits imposed by the disorder elicit considerable pain in the affected joint. Crepitation in the arthritic joint, with or without clicking, may occur during jaw movement [101, 102]. Imaging of an osteoarthritic joint in the advanced stages typically shows erosion of the cortical outline, osteophytes, marginal spurs, subcortical cysts, reduced joint space and a perforated disc, among other features [101, 103, 104]. Studies on human osteoarthritic TMJs revealed disc perforation, atypical cellular architecture, osteophyte formation, subchondral bone resorption, articular surface disruption, drastic changes in the extracellular collagenous matrix and proteoglycan loss [105– 108]. Osteoarthritis, a local inflammatory disease, is
12
Cells Tissues Organs 2003;174:6–16
thought to be associated with overloading beyond the joint’s adaptive capacity [11, 81–83]. It is set in motion by loss or severe impairment of the mechanism keeping degenerative processes in check [108, 110, 111]. Such processes may be subclinical or may be associated with painful and dysfunctional TMJ. The severity and duration of the clinical symptoms vary considerably from patient to patient [11, 101]. Intense pain and discomfort may alternate with asymptomatic periods. Inconsistency between the clinical symptoms and the imaging picture, the former may indicate severe disease while the radiological scan may show no evidence of joint disorder or vice versa, is characteristic of osteoarthritis. This paradoxical situation underlines the need for biomarkers and arthroscopic and biochemical [11, 50, 51, 105–111] standards to extend the diagnostic armamentarium and afford better treatment protocols [1, 106–108].
Nitzan
Fig. 4. MRI of the TMJ in open lock: The condyle is trapped in front
of the folded lagging disc (arrow) and cannot return to the fossa. Release of the disc following lavage of the upper compartment enables the reestablishment of simultaneous sliding of the disc together with the condyle.
notably metalloproteinases, which are present, in both latent and active forms, in the joint [11, 111, 112]. It is often not possible to control the damaging effects of the enzymes, i.e. various inflammatory and tissue-degrading products, by noninvasive conservative means, to the extent that supposedly protective reactions can become an almost independent virulent inflammatory core. Since the TMJ has a limited blood supply, the natural elimination of inflammatory elements greatly depends on the joint’s functional performance [52], which is often painful, and, in many instances, not recommended, especially if effusions are present. Active removal of such noxious products by lavage, in conjunction with joint unloading, is therefore essential at this stage of the disease. Following active elimination of inflamed SF, joint function is improved and the pain diminished. In the next step of the rehabilitation process, better functional performance of the joint allows for nutrition, waste removal, lubrication and absorption of medication (anti-inflammatory and analgesic drugs) in the affected area. In our experience, such treatment obviated the need for corrective surgery in 68.4% of patients. Naturally, similar symptoms caused by other pathologies, such as bone spicules, fibroankylosis or disc perforation, are not amenable to lavage [113].
Summary and Conclusion
However, as Milam [11] correctly pointed out the value of markers is questionable and the boundary between remodeling or adaptation and the disease state is, at present, ill-defined. As he also noted, some of the markers revealed by current research efforts may well be indicative of adaptation processes rather than disease states [11, 108]. Future studies are needed to establish criteria which reliably distinguish between disease and adaptation states and identify factors responsible for the disruption of the stable joint relationships [11]. Overloading is widely believed to be an important factor in disrupting stable joint relationships [11]. Tissue degeneration, one result of excessive mechanical force, in turn, is associated with free radical production [11]. Uncontrolled, it contributes to joint degeneration, starting with disruption of the lubrication system [42], leading to increased friction and adhesive force between the articular surfaces [11, 105–111]. This course of events culminates in fatigue and wear of joint components [57, 58] with underlying collagenous matrix disruption and glycosaminoglycan loss, particularly in the more elastic articular disc [58]. The matrix-degrading agents most likely to be active in the degenerative process are proteinases,
‘Friction and Adhesive Forces’
This review describes functional aspects of the TMJ. Better understanding of normal joint function would certainly improve our treatment approaches in joint disorders. This paper addresses the following main issues: the importance of the joint lubrication, the role of PLs in joint lubrication, the indirect role of HA in joint lubrication, the possible effect of joint overloading on the lubrication system, aberrations in the lubrication system and their role in joint dysfunction (disc displacement, ADP vs. disc displacement without reduction, TMJ open lock vs. dislocation, role in osteoarthritis). Better understanding of joint function and dysfunction has led to an increased use and development of simple treatment modalities, such as joint hydraulic pumps, arthrocentesis, arthroscopic lavage and lysis, and arthroscopic surgery. The use of these procedures has markedly reduced the number of TMJ surgical interventions and related complications.
Cells Tissues Organs 2003;174:6–16
13
References 1 Annandale, T. (1887) On displacement of the interarticular cartilage of the lower jaw and its treatment by operation. Lancet i 411. 2 Ireland, V.E. (1951) The problem of the jaw. Proc R Soc Med 44: 363. 3 Westesson, P.-L., M. Rohlin (1984) Internal derangement related to osteoarthrosis in temporomandibular joint autopsy specimens. Oral Surg Oral Med Oral Pathol 57: 17. 4 Dolwick, M.F. (1995) Intra-articular disc displacement. I. Its questionable role in temporomandibular joint pathology. J Oral Maxillofac Surg 53: 1069. 5 Wilkes, C.H. (1989) Internal derangement of the TMJ. Arch Otolaryngol Head Neck Surg 115: 469. 6 Nannmark, U., L. Sennerby, T. Haraldon (1990) Macroscopic, microscopic and radiologic assessment of the condylar part of the TMJ in elderly subjects: An autopsy study. Swed Dent J 14: 163. 7 Ferrar, W.B., W.J. McCarty (1990) Inferior joint space arthrography and characteristics of condylar paths on internal derangement of the TMJ. J Prosthet Dent 41: 548. 8 Isberg, A., S.E. Widmalm, R. Ivarsson (1985) Clinical, radiographic and electromyographic study of patients with internal derangement of the temporomandibular joint. Am J Orthod 88: 453. 9 Rohlin, M., P.L. Westesson, L. Eriksson (1985) The correlation of the temporomandibular joint sounds with joint morphology in fifty-five autopsy specimens. J Oral Maxillofac Surg 43: 194. 10 Gay, T., C.N. Bertolami, R.B. Donoff, et al (1987) The acoustical characteristics of the normal and abnormal temporomandibular joint. J Oral Maxillofac Surg 45: 397. 11 Milam, S.B. (1995) Articular disc replacement and degenerative temporomandibular joint disease; in Sessle, B.J., P.S. Bryant, R.A. Dionne (eds): Temporomandibular Disorders and Related Pain Conditions. Progress in Pain Research and Management. Seattle, IASP Press, vol 4, pp 89–112. 12 Westesson, P.-L., S.L. Brooks (1992) Temporomandibular joint relationship between MR evidence of effusion and the presence of pain and disc displacement. Am J Roentgenol 159: 559. 13 Widmalm, S.E., P.-L. Westesson, S.L. Brook, et al (1992) Temporomandibular joint sounds: Correlation to joint structure in fresh autopsy specimens. Am J Orthod Dentofacial Orthop 101: 60. 14 Ryan, D.E. (1996) Temporomandibular joint disc surgery: Biologic basis and treatment outcome; in Stegenga, B., L.G.M. de Bont (eds): Management of Temporomandibular Joint Diseases. Basel, Birkhäuser Verlag, pp 149– 158. 15 Dolwick, M.F., G. Dimitroulis (1994) Is there a role for temporomandibular joint surgery? Br J Oral Maxillofac Surg 32: 307.
14
16 Mercuri, L.G. (1995) Surgical management of temporomandibular joint disorders; in Sessle, B.J., P.S. Bryant, R.A. Dionne (eds): Temporomandibular Disorders and Related Pain Conditions. Progress in Pain Research and Management. Seattle, IASP Press, vol 4, pp 429– 442. 17 Dolwick, M.F. (1995) Temporomandibular joint disk displacement: Clinical perspectives; in Sessle, B.J., P.S. Bryant, R.A. Dionne (eds): Temporomandibular Disorders and Related Pain Conditions. Progress in Pain Research and Management. Seattle, IASP Press, vol 4, pp 79–87. 18 Hall, D.H. (1995) Intra-articular disc displacement. II. Its significant role in TMJ pathology. J Oral Maxillofac Surg 53: 1973. 19 Kaplan, P.A., H.K. Tu, P.R. Slender, et al (1986) Inferior joint space arthrography of normal TMJ: Reassessment of diagnostic criteria. Radiology 3: 577. 20 Kircos, L.T., D.A. Ortendahl, A.S. Mark, et al (1987) MRI of the TMJ disc in asymptomatic volunteers. J Oral Maxillofac Surg 54: 852. 21 Sanders, B. (1986) Arthroscopic surgery of the temporomandibular joint: Treatment of internal derangement with persistent closed lock. Oral Surg 62: 361. 22 Nitzan, D.W., B. Samson, H. Better (1997) Long term outcome of arthrocentesis for sudden-onset, persistent, severe closed lock in the TMJ. J Oral Maxillofac Surg 55: 151. 23 Montgomery, M.T., J.E. Van Sickels, S.E. Harms, et al (1989) Arthroscopic TMJ surgery: Effects on signs, symptoms and disc position. J Oral Maxillofac Surg 47: 1263. 24 Moses, J.J., D. Sartoris, R. Glass (1989) The effect of arthroscopic surgical lysis and lavage of the superior joint space on TMJ disc position and mobility. J Oral Maxillofac Surg 47: 674. 25 Hills, B.A. (1990) Oligolamellar nature of the articular surface. Br J Rheumatol 17: 349– 356. 26 Hills, B.A. (1996) Synovial surfactant and the hydrophobic articular surface. J Rheumatol 23: 1323. 27 Schwarz, I.M., B.A. Hills (1996) Synovial surfactant: Lamellar bodies in type B synoviocytes and phospholipids in synovial fluid and the articular lining. Br J Rheumatol 35: 821. 28 Schwarz, I.M., B.A. Hills (1998) Surface-active phospholipids as the lubricating component of lubricin. Br J Rheumatol 37: 21–26. 29 Hills, B.A. (2000) Boundary lubrication in vivo. Proc Inst Mech Eng 214/1: 83–94. 30 Higaki, H., T. Murakami, Y. Nakanishi, et al (1998) The lubricating ability of biomembrane models with dipalmitoyl phosphatidylcholine and gamma-globulin. Proc Inst Mech Eng 212: 337. 31 Williams, P.F., G.L. Powell, M. LaBerge (1993) Sliding friction analysis phosphatidylcholine as a boundary lubricant for articular cartilage. Proc Inst Mech Eng 20: 759.
Cells Tissues Organs 2003;174:6–16
32 Pickard, J.E., J. Fisher, E. Ingham, J. Egan (1998) Investigation into the effect of proteins and lipids on the functional properties of articular cartilage. Biomaterials 19: 1807–1812. 33 Bole, G.G. (1962) Synovial fluid lipids in normal individuals and patients with rheumatoid arthritis. Arthritis Rheum 5: 589–601. 34 Hills, B.A., B.D. Butler (1984) Identification of surfactants in synovial fluid and their ability to act as boundary lubricants. Ann Rheum Dis 48: 51–57. 35 Rabinowitz, J.I., J.R. Gregg, J.E. Nixon (1984) Lipid composition of the tissues of human knee joints. Clin Orthop 190: 292–298. 36 Punzi, L., S. Todesco, G. Toffano, E. Bigon, A. Bruni (1986) Phospholipids in inflammatory synovial effusions. Rheumatol Int 6/1: 7–11. 37 Hills, B.A., M.K. Monds (1999) Enzymatic identification of the load bearing boundary lubricant in the joint. Br J Rheumatol 37: 137. 38 Nitzan, D.W., U. Nitzan, P. Dan, S. Yadgar (2001) Hyaluronic acid protects surface active phospholipids from lysis by exogenous phospholipase A2. Rheumatology 40: 336–340. 39 Dan, P., D. Nitzan, A. Dagan, et al (1996) H2O2 renders cells accessible to lysis by exogenous phospholipase A2: A novel mechanism for cell damage in inflammatory processes. FEBS Lett 383: 75. 40 De Bont, L.G.M. (1985) Temporomandibular Joint Articular Cartilage Structure and Function. PhD thesis. Groningen, Van Denderen. 41 Toller, P.A. (1961) The synovial apparatus and temporomandibular joint function. Br Dent J 111: 355–362. 42 Kopp, S., B. Wenneberg, E. Clemensson (1983) Clinical microscopic and biochemical investigation of synovial fluid from temporomandibular joints. Scand J Dent Res 91: 33–41. 43 Israel, H.A. (1989) Synovial fluid analysis. Oral Maxillofac Surg Clin North Am 1/1: 85–92. 44 Marchetti, C., G. Bernasconi, M. Reguzzoni, et al (1997) The articular disc surface in different functional conditions of the human temporomandibular joint. J Oral Pathol Med 26: 278– 282. 45 Ghadially, F.N., N.K. Yong, J.-M.A. Lalonde (1982) A transmission electron microscopic comparison of the articular surface of cartilage processed attached to bone and detached from bone. J Anat 135: 685–706. 46 Forster, H., J. Fisher (1999) The influence of continuous sliding and subsequent surface wear on the friction of articular cartilage. Proc Inst Mech Eng 213: 329–345. 47 Norman, A.G., M.J. Kaab (1999) The surface contour of articular cartilage in an intact, loaded joint. J Anat 195/1: 45–56. 48 Wise, C.M., R.E. White, C.A. Agudelo (1987) Synovial fluid lipid abnormalities in various disease states: Review and classification. Semin Arthritis Rheum 16/3: 222–230.
Nitzan
49 Rachamin, E., H. Better, A. Dagan, D.W. Nitzan (2001) Electron microscope and biochemical observations of the surface active phospholipids on the articular surfaces and in the synovial fluid of the temporomandibular joint: A preliminary investigation. J Oral Maxillofac Surg 59: 1326–1332. 50 de Bont, L.G.M., B. Stegenga (1993) Pathology of the TMJ internal derangement and osteoarthrosis. Int J Maxillofac Surg 22: 71. 51 Stegenga, B., L.G.M. de Bont, G. Boering, et al (1991) Tissue responses to degenerative changes of the TMJ: A review. J Oral Maxillofac Surg 49: 1079. 52 Nitzan, D.W. (1994) Intra-articular pressure in the functioning human TMJ and its alteration by uniform elevation of the occlusal plane. J Oral Maxillofac Surg 52: 671. 53 Solberg, W.K. (1986) Temporomandibular disorders: management of internal derangement. Br Dent J 160: 379–385. 54 Forster, H., J. Fisher (1996) The influence of loading time and lubricant on the friction of articular cartilage. Proc Inst Mech Eng 210:109. 55 Boyd, R.L., C.H. Gibbs, P.E. Mahan, et al (1990) Temporomandibular joint forces measured at the condyle of Macaca arctoides. Am J Orthod Dentofacial Orthop 97: 472. 56 Chen, J., L. Xu (1994) A finite element analysis of the human TMJ. J Biomech Eng 116: 401. 57 Wright, V. (1969) Lubrication and Wear in Joints. Proc Symp Biol Eng Soc Leeds. High Wycombe, Harrison, chap 1, 4, 12, 15, 16, pp 1–14, 29–34, 110–114, 134–140, 142–149. 58 Mow, V.C., G.A. Ateshian (1997) Lubrication and wear of diarthrodial joints; in Mow, V.C., W.C. Hayes (eds): Basic Orthopaedic Biomechanics, ed 2. Philadelphia, Lippincott-Raven, pp 275–315. 59 Dowson, D. (1990) Bio-tribology of natural and replacement synovial joints; in Mow, V.C., A Radcliffe, S.L.-Y. Woo (eds): Biomechanics of Diarthrodial Joints. New York, Springer, vol 2, pp 303–345. 60 Blake, D.R., J. Unsworth, J.M. Outhwaite, et al (1989) Hypoxic reperfusion injury in the inflamed human joint. Lancet ii: 289. 61 Marry, P., R. Williams, N. Cox, et al (1991) Comparative study of intra-articular pressure dynamics in joints with acute traumatic and chronic inflammatory effusions: Potential implications for hypoxic-reperfusion injury. Ann Rheum Dis 50: 917. 62 Mapp, P.I., M.C. Grooteveld, D.R. Blake (1995) Hypoxia, oxidative stress and rheumatoid arthritis. Br Med Bull 51: 419. 63 Grootveld, M., E.B. Henderson, A. Farrell, et al (1991) Oxidative damage to hyaluronate and glucose in synovial fluid during exercise of the inflamed rheumatoid joint. Biochem J 273: 459. 64 Linn, F.C. (1968) Lubrication of animal joints: The mechanism. J Biochem 1: 193. 65 Swann, D.A., E.I. Radin, M. Nazimiec, et al (1974) Role of hyaluronic acid in joint lubrication. Ann Rheum Dis 33: 318. 66 Tabor, D. (1981) Friction – The present state of our understanding. J Lubrication Technol 103: 169.
‘Friction and Adhesive Forces’
67 Etzion, I., M. Amit (1992) The effect of small normal loads on the static friction coefficient for very smooth surfaces. J Tribol 11: 100. 68 Armstrong, C.G., V.C. Mow, C.R. Wirth (1985) Biomechanics of impact-induced microdamage to articular surface – A possible genesis for chondromalasia patella; in Finerman, G.A.M. (ed): AAOS Symposium on Sports Medicine: The Knee. St Louis, Mosby, pp 54–69. 69 Radin, E.L., R.B. Martin, D.B. Burr, et al (1984) Effects of mechanical loading on the tissue of rabbit knee. J Orthop Res 2: 221. 70 Vener, J.M., R.C. Thompson, J.L. Lewis, et al (1992) Subchondrial damage after acute transarticular loading: An in vitro model of joint injury. J Orthop Res 10: 759. 71 Bloebaum, O.Y.D., K.M. Radley (1995) Threedimensional surface analysis of young adult human articular cartilage. J Anat 187: 293. 72 Stanley, H.M., I. Etzion, D.B. Boggy (1990) Adhesion of contacting rough surfaces in the presence of subboundary lubrication. J Tribol 112: 98. 73 Hills, B.A., K. Thomas (1998) Joint stiffness and ‘articular gelling’: Inhibition of the fusion of articular surfaces by surfactant. Br J Rheumatol 37: 532. 74 Ward, D.M., R.G. Berhents, J.S. Goldberg (1990) Temporomandibular synovial fluid pressure response to altered mandibular positions. Am J Orthod Dentofacial Orthop 98: 22– 28. 75 Murakami, K.I., T. Iizuka, M. Matsuki, T. Ono (1987) Recapturing the persistent anteriorly displaced disk by mandibular manipulation after pumping and hydraulic pressure to the upper joint cavity of the temporomandibular joint. Cranio 5: 17–24. 76 Kendell, B., D.E. Frost (1996) Arthrocentesis. Atlas Oral Maxillofac Surg Clin North Am 4/2: 1–14. 77 Nitzan, D.W., F.M. Dolwick, G.A. Martinez (1991) Temporomandibular joint arthrocentesis: A simplified method for severe limited mouth opening. J Oral Maxillofac Surg 49: 1163–1167. 78 Zardeneta, G., H. Muckai, B.D. Tiner, et al (1995) Positive correlation between TMJ pain and alpha-hemoglobin recovered by TMJ lavage. J Dent Res 74: 172. 79 Zardeneta, G., J.P. Schmitz, S.B. Milam (1996) Latent proteases recovered from human TMJ lavage fluid. J Oral Maxillofac Surg 54(suppl 3): 86. 80 Tenenbaum, H.C., B.V. Freeman, D.J. Psutka, J.I. Baker (1999) Temporomandibular disorders: Disc displacement. J Orofac Pain 13/4: 285–290. 81 Laskin, D. (1994) Etiology and pathogenesis of internal derangement of the temporomandibular joint. Oral Maxillofac Surg Clin North Am 6: 217. 82 de Bont, L.G.M., G. Boering, R.S.B. Liem, et al (1986) Osteoarthritis and internal derangement of the TMJ: A light microscopic study. J Oral Maxillofac Surg 44: 634.
83 Juniper, R.P. (1984) Temporomandibular joint dysfunction: A theory based upon electromyographic studies of the lateral pterygoid muscle. Br J Oral Maxillofac Surg 22: 1. 84 Jupiter, R.P. (1981) The superior pterygoid muscle. Br J Oral Surg 19: 121. 85 Frost, H.M. (1977) Musculoskeletal pain; in Alling, C.C., P.E. Mahan (eds): Facial Pain, ed 2. Philadelphia, Lea & Febiger. 86 Ogus, H. (1987) The mandibular joint: Internal rearrangement. Br J Oral Maxillofac Surg 25: 218. 87 Mahan, P.E., C.C. Alling (1991) Facial Pain, ed 3. Philadelphia, Lea & Febiger. 88 Moses, J.J. (1989) Lateral impingement syndrome and endural surgical technique. Oral Maxillofac Surg Clin North Am 1: 165. 89 Isberg-Holm, A., R. Ivarsson (1980) The movement pattern of the mandibular condyles in individuals with and without clicking. Dent Maxillofac Radiol 9: 55. 90 Juniper, R.P. (1987) The pathogenesis and investigation of TMJ dysfunction. Br J Oral Maxillofac Surg 25: 105. 91 Nitzan, D.W. (2001) The process of lubrication impairment and its involvement in temporomandibular joint disc displacement: A theoretical concept. J Oral Maxillofac Surg 59/1: 36– 45. 92 Nitzan, D.W., Y. Marmary (1997) The ‘anchored disc phenomenon’: A proposed etiology for sudden-onset, sever and persistent closed lock of the TMJ. J Oral Maxillofac Surg 55: 797–802. 93 Benito, C., G. Csares, C. Benito (1998) TMJ static disc: Correlation between clinical findings and pseudodynamic magnetic resonance images. J Craniomand Prac 16/4: 242–251. 94 Ernandez-Sanroman, J. (2000) Morfologı´a y dina´mica de la articulacio´n temporomandibular en pacientes con sospecha clı´nica de trastorno interno: Estudio prospectivo con resonance magnética. Rev Esp Cirug Oral y Maxilofac 22: 233–240. 95 Rao, V.M., M.D. Liem, A. Farole, A.A. Razek (1993) Elusive ‘stuck’ disc in the temporomandibular joint: Diagnosis with MR imaging. Radiology 189: 823–827. 96 Nitzan, D.W., B. Samson, H. Better (1997) Long-term outcome of arthrocentesis for sudden-onset, persistent, severe closed lock of the TMJ. J Maxillofac Surg 55: 151–157. 97 Nitzan, D.W., A. Goldfarb, I. Gati, R. Kohen (2002) Changes in the reducing power of synovial fluid from temporomandibular joint with ‘anchored disc phenomenon’. J Oral Maxillofac Surg 60: 735–740. 98 Nitzan, D.W., Y. Etsion (2002) Adhesive force – The underlying cause of the ‘anchored disc phenomenon’. Int J Oral Maxillofac Surg 31/1: 94–99. 99 Kai, S., H. Kai, E. Nakayama, O. Tabata, H. Tashiro, T. Miyajima, M. Sagaguri (1992) Clinical symptoms of open-lock position of the condyle. Relation to anterior dislocation of the temporomandibular joint. Oral Surg Oral Med Oral Pathol 74: 143–148.
Cells Tissues Organs 2003;174:6–16
15
100 Nitzan, D.W. (2002) Temporomandibular joint ‘open-lock’ vs condylar dislocation: Signs and symptoms, imaging, treatment and pathogenesis. J Oral Maxillofac Surg 60: 506– 513. 101 Kopp, S. (1995) Degenerative and inflammatory temporomandibular joint disorders: Clinical perspectives; in Sessle, B.J., P.S. Bryant, R.A. Dionne (eds): Temporomandibular Disorders and Related Pain Conditions. Progress in Pain Research and Management. Seattle, IASP Press, vol 4, pp 119–131. 102 Zarb, A.Z., L.D.S. Carlsson (1999) Temporomandibular disorders: Osteoarthritis. Orofacial Pain 13/4: 295–306. 103 Dolwick, M.F., B. Sanders (1985) TMJ Internal Derangement and Arthrosis. Surgical Atlas. St Louis, Princeton, pp 49, 120, 122. 104 Gynther, G.W., G. Tronje, A.B. Holmlund (1996) Radiographic changes in the temporomandibular joint in patients with generalized osteoarthritis and rheumatoid arthritis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 81: 613.
16
105 Israel, H.A., B.E. Diamond, F. Saed-Nejad, et al (1997) Correlation between arthroscopic diagnosis of osteoarthritis and synovitis of the human temporomandibular joint and keratan sulfate levels in the synovial fluid. J Oral Maxillofac Surg 55: 210. 106 Radcliffe, A., H.A. Israel, F. Saed-Nejad, et al (1998) Proteoglycans in the synovial fluid of the temporomandibular joint as an indicator of changes in cartilage metabolism during primary and secondary osteoarthritis. J Oral Maxillofac Surg 56: 204. 107 Quinn, J.H., N.G. Bazan (1990) Identification of prostaglandin E2 and leukotriene B4 in the synovial fluid of painful, dysfunctional temporomandibular joint. J Oral Maxillofac Surg 48: 968.
Cells Tissues Organs 2003;174:6–16
108 Poole, A.R., G. Rizkalla, M. Ionescu, et al (1993) Osteoarthritis in the human knee: A dynamic process of cartilage matrix degradation, synthesis and reorganization. Agents Actions Suppl 39: 3. 109 Cameraman, D. (1989) The biology of osteoarthritis. N Engl J Med 320: 1322. 110 Kubota, E., H. Imamura, T. Kubota, et al (1997) Interleukin 1ß and stromelysin (MMP3) activity of synovial fluid as possible markers of osteoarthritis in the temporomandibular joint. J Oral Maxillofac Surg 55: 20. 111 Kubota, E., T. Kubota, J. Matsumoto, et al (1998) Synovial fluid cytokines and proteinases as markers of temporomandibular joint disease. J Oral Maxillofac Surg 56: 192. 112 Birkedal-Hansen, H., W.G.I. Moore, M.K. Bodden, et al (1993) Matrix metalloproteinases: A review. Crit Rev Oral Biol Med 4: 197. 113 Nitzan, D.W., A. Price (2001) The use of arthrocentesis for the treatment of osteoarthritic temporomandibular joint. J Oral Maxillofac Surg 59: 1154–1159.
Nitzan
Cells Tissues Organs 2003;174:17–25 DOI: 10.1159/000070571
Subchondral Bone Resorption in Temporomandibular Joint Disorders Helen E. Gruber a John Gregg b a Department b Department
of Orthopaedic Surgery, Carolinas Medical Center, Charlotte, N.C., and of Biomedical Sciences and Pathobiology, Virginia Tech University, Blacksburg, Va., USA
Key Words Bone resorption W Antiresorbing agents W Temporomandibular joint W Temporomandibular joint disease W Osteoporosis W Osteoclast
Abstract Several tissues are involved in temporomandibular joint (TMJ) health, including synovial fluid, the TMJ disc, articular cartilage, and subchondral bone. This article focuses upon bone resorption in temporomandibular joint disorders (TMD) and has the following objectives: (1) to provide a brief review of the current understanding of bone formation and bone resorption (bone remodeling); (2) to present selected case studies which illustrate the spectrum of bone resorption patterns in TMD patients of various ages; (3) to review previous reports in the literature describing loss of subchondral bone in TMD, and (4) to discuss the interaction between osteoporosis and TMD and the potential role for antiresorbing agents in TMD therapy. Copyright © 2003 S. Karger AG, Basel
Abbreviations used in this paper
M-CSF RANKL TMD TMJ
macrophage colony-stimulating factor receptor for activation of nuclear factor kappa B ligand temporomandibular joint disorders temporomandibular joint
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0017$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
Introduction to Bone Cells and Bone Remodeling
Bone remodeling consists of two phases: bone resorption by osteoclasts followed by bone formation by osteoblasts. This process occurs continually and simultaneously at different sites throughout the skeleton. Bone matrix is produced by osteoblasts at sites of previous bone resorption (Howship’s lacunae; resorption cavities). Prior to this, an initial resorption stimulus signals and recruits osteoclast precursors to the site of bone resorption. Mature osteoclasts, rich in lysosomes with high tartrate-resistant acid phosphatase content [Teitelbaum, 2000], carry out the resorption process and leave behind empty resorption cavities (Howship’s lacunae). Figure 1 illustrates bone resorption in the temporomandibular condyle of a small mammal (note numerous osteoclasts and sites of active bone resorption). In a study of human autopsy specimens of the temporomandibular joint (TMJ), Flygare et al. [2002] documented prominent histologic erosive bone changes in 31% of the condyles, and in 49% of the temporal components which were examined. Although other investigators have examined histologic bony changes in the human TMJ [Öberg et al., 1971; Bean and Omnell, 1977; Hansson and Öberg, 1977; Flygare et al., 1997], there are few studies in the literature; this appears to be an area where more research is needed correlating morphologic, imaging and clinical parameters in temporomandibular joint disorders (TMD). At the cellular and morphologic level, osteoclasts seal their perimeter against the bone surface by means of a
Helen E. Gruber, PhD Orthopaedic Research Biology Cannon Bldg, 3rd floor, Carolinas Medical Center PO Box 32861, Charlotte, NC 28232 (USA) Tel. +1 704 355 5665, Fax +1 704 355 2845, E-Mail
[email protected]
Fig. 1. Photomicrograph of the condyle and disc of the TMJ of an adult sand rat (Psammomys obesus). Arrows mark osteoclasts in sites of bone resorption. Masson’s trichrome. A !190. B !375.
podosome; inside this sealed region a ruffled border of cytoplasmic extensions forms the specialized region in which resorption occurs in the presence of low pH and secreted matrix-degrading proteinases [Teti et al., 1991; Väänäen and Horton, 1995; Roodman, 1996] (fig. 2). During remodeling, it is important to remember that although osteoclasts are not numerous, the individual cell activity of an osteoclast in removing bone matrix is much greater than the individual cell activity of an osteoblast in making new bone matrix [Gruber et al., 1986]. During resorption, signals recruit osteoblast precursors to the Howship’s lacunae. With the advent of mature osteoblasts, bone formation begins. Early osteoblasts are very active in depositing matrix, and are plump cells with abundant alkaline phosphatase demonstrable in their cell membranes. These cells secrete procollagen, collagen fibrils are formed outside the cell, and collagen bundles are then laid down in an ordered arrangement at the bone surface. As a site of bone formation nears completion, osteoblasts become less active and appear more flattened and
18
Cells Tissues Organs 2003;174:17–25
spindle-shaped; when formation has ceased, fibroblastlike cells line the quiescent bone surface. There are two important components of the remodeling cycle: (1) the depth to which osteoclasts ‘excavate’ the resorption lacunae, and (2) the efficiency with which osteoblasts are able to fill in this excavated site. If it is perfectly filled in, the amount of bone formed equals the amount of bone lost, and there is bone balance. When the amount of bone resorbed exceeds the amount formed, however, net bone loss occurs. If the imbalance between formation and resorption is low, and overall bone turnover is low, there will not be a big change in bone mass. If, however, there is an imbalance between formation and resorption which favors resorption in the presence of high turnover (i.e. the presence of a large number of active remodeling sites throughout the skeleton, also termed a high activation frequency), large bone loss will result. Following its synthesis and secretion, bone matrix must be converted into mature osteoid prior to the initiation of mineralization. Although bone matrix maturation
Gruber/Gregg
Fig. 2. Transmission electron micrograph illustrating ruffled border in the osteoclast. !5,060.
is an important but poorly understood process, two processes are reasonably well established: proteoglycan loss and formation of intermolecular cross-links. Newly secreted matrix components can be considered to be at 0% maturation, and matrix in which mineralization is being initiated can be considered to be fully matured. The histologic benchmark for the evaluation of mineralization and determination of the quality of mineralization is tetracycline incorporation at the mineralizing front. Tetracycline in vivo labelling techniques have made possible the quantitative histomorphometric determinations of bone formation and mineralization indices.
Recent Insights into Bone Remodeling
Ducy et al. [2000] have recently reviewed current information on the cell biology of the osteoblast. The osteoblast, which derives from mesenchymal stem cells,
Bone Resorption in Temporomandibular Joint Disorders
produces transcripts which distinguish it from the fibroblast cell: Cbfa1 controls the directional information for a differentiation factor for the osteoblast lineage by regulating the expression of osteocalcin, a gene expressed in terminally differentiated osteoblasts. Following embryogenesis, Cbfa1 expression is limited to osteoblasts (with some expression also occurring in hypertrophic chondrocytes). OSE1 is another regulatory element present in the osteocalcin promoter; it appears to be found in poorly differentiated osteoblasts. Osteocalcin is secreted by osteoblasts and appears to inhibit osteoblast function. During osteoblast differentiation, each of the major families of growth factors plays a role in osteoblast embryonic differentiation, and the growth factors themselves become bound into the bone matrix and may be subsequently released during later bone resorption at that site. Teitelbaum [2000] has recently summarized the current understanding of the cell biology of the osteoclast, and recent advances in osteoclastogenesis have been re-
Cells Tissues Organs 2003;174:17–25
19
viewed by Khosla [2002]. Maturation of osteoclasts from their macrophage precursors requires marrow stromal cells or their osteoblast progeny as shown by Udagawa et al. [1990]. These cells produce macrophage colony-stimulating factor (M-CSF) and the receptor for activation of nuclear factor kappa B (NF-ÎB; RANK) ligand (RANKL). M-CSF is essential for macrophage maturation, but formation of osteoclasts also requires contact between osteoclast precursors and stromal cells or osteoblasts [Udagawa et al., 1990]. As described by Hofbauer et al. [1999], the quantity of bone resorbed depends on the balance between expression of RANKL and of its inhibitor, osteoprotegerin (OPG). Hofbauer et al. suggest that the stimulation of the pool of M-CSF precursors to committed osteoclastogenesis by RANKL may be one of the central pathophysiologic pathways involved in increasing the number of osteoclasts in osteoporosis.
Selected Case Studies Illustrating the Spectrum of Bone Resorption Patterns in TMD
Because of its special form and position and its anatomical location underneath the base of the skull, the TMJ is challenging for radiographic imaging. The following case reports present panoramic tomography radiographs of the TMJ condyle. This series was selected from the patient files of one of the authors (J.G.) to illustrate the spectrum of bone resorption patterns in TMD patients of various ages. Case 1: Mild Incipient Middle-Aged Remodeling and Resorption This patient is a 49-year-old female with clinical diagnoses of temporomandibular arthralgia, masticatory myalgia and subacute osteoarthritis. She presented with a 2- to 4-month history of recent onset pain and crepitus in the right TMJ. She had suffered previous facial trauma and experiences chronic episodic sleep dysfunction, periodic limb movement and bruxism parasomnias. Clinical findings included 1+ (4) auscultated crepitus, right-deviant mandibular movement restricted to 35 mm, pain in the right lateral joint capsule upon light pressure, and myalgia in the cervical and masticatory groups. Imaging findings on the left condyle (fig. 3A) showed subtle anterior lipping; imaging of the right condyle (fig. 3B) showed mild condylar remodeling with superior and anterosuperior slope flattening. Incipient subcondylar cavitations were present.
20
Cells Tissues Organs 2003;174:17–25
Case 2: Moderate Middle-Aged Remodeling and Resorption This 49-year-old female with clinical diagnoses of temporomandibular arthralgia with acute neuropathic pain and chronic arthrosis presented with a 10-year history of TMJ pain, clicking, joint popping and locking. She experienced associated daily throbbing and migraine-like headaches, and also knee and spinal pain. There was no history of trauma. She has documented osteoporosis by lumbar QCT and DEXA scans. Clinical findings included touchevoked pain bilaterally in the preauricular regions, moderate (2+) crepitus, and restricted condylar translation. Imaging findings noted bilateral moderately severe bony condylar atrophy, anterior slope flattening and chronic (repaired) subcondylar cavitations (fig. 4). Case 3: Rapid Resorption in a Young Patient Clinical diagnoses in this 23-year-old female included bilateral temporomandibular progressive juvenile osteoarthritis, internal disc derangements and fibromyalgia. Patient history reported pain, joint locking, pulsatile midfacial headaches, and knee, back and hip pain which were exacerbated with exercise. Related medical history included familial osteoarthritis, borderline rheumatoid arthritis with ANA-positive titer, fibromyalgia with mild depression, insomnia and frequent nocturnal awakenings. There was no history of trauma. The patient exhibited multifocal tender points including intercostal, occipital and masticatory muscle groups bilaterally. Mandibular range of motion was restricted to 28 mm; reciprocal clicking was present at mandibular mid-opening. Imaging findings at age 14 documented a grossly normal right condyle but atrophy and anterior cavitation of the left condyle (fig. 5A). At age 23, there had been progression with moderately severe right condylar superior and anterior resorption (fig. 5B).
Fig. 3. Case 1. A Radiograph of the left condyle showing relatively
normal structure with mild remodeling and subtle anterior lipping. B Right condyle showing mild condylar remodeling and incipient
subcondylar cavitations. Fig. 4. Case 2 illustrating radiographic features of moderate middleaged remodeling and resorption. Fig. 5. Case 3 illustrating radiographic features of rapid resorption in a young patient. A Age 14. B Age 23. Fig. 6. Case 4 illustrating radiographic features of severe middleaged active-stage resorption. Fig. 7. Case 5 illustrating radiographic features of severe middleaged erosion of the condyle and articular fossa. Fig. 8. Case 6 illustrating radiographic features of the remission stage of a septuagenarian with previous severe erosion.
Gruber/Gregg
Bone Resorption in Temporomandibular Joint Disorders
Cells Tissues Organs 2003;174:17–25
21
Case 4: Severe Middle-Aged Active Stage Resorption This 44-year-old male with clinical diagnoses of bilateral temporomandibular severe arthrosis and osteoarthritis presented with a history of bilateral joint capsule pain and altered dental occlusal pattern over the previous 10 years. Related medical history included severe obstructive sleep apnea syndrome, hypertension, type II diabetes and progressive obesity. Upon clinical examination, severe joint crepitus (3+), mandibular retrognathia and anterior open-bite deformity were noted. Imaging findings revealed advanced arthrosis with an atrophic ‘nailhead’ left condyle, active bone resorption with vertical collapse, and a resorbed anterior condylar slope (fig. 6). Case 5: Severe Middle-Aged Erosion of the Condylar and Articular Fossa The patient was a 50-year-old male with advanced left arthrosis and maxillofacial skeletal deformities. He presented with chronic daily bitemporal headaches and observed bruxism and related medical histories of kidney stones, ulcerative colitis, hip fracture and osteoporosis. Clinical findings included severe mandibular retrognathia, hypomobility and facial asymmetry. Imaging studies revealed advanced left TMJ irregular condylar erosion, presence of a loose body, and remodeling of the anterior articular eminence and glenoid fossa (fig. 7). Case 6: Severe Erosion Pattern in a Septuagenarian; Remission Stage This 78-year-old female has a clinical diagnosis of remission stage bilateral osteoarthritis and mandibular hypomobility. Her history included bilateral temporomandibular midfacial daily pain when aged 40–55. However she notes no jaw or facial pain for the last 20 years. Related medical history included gastroesophageal reflux syndrome. Clinical findings documented severe crepitus (4+) bilaterally, absence of capsule pain on palpation, and restricted mandibular range of opening to 28 mm. Imaging findings showed ‘nail-head’ atrophy with anterior lipping of bilateral condyles (fig. 8).
Bone Resorption and TMD
In TMD, bony changes and bone resorption occur in both the condyle and temporal components of the TMJ with aging, and may range from mild decreases in cortical bone to severe destruction of condyle and temporal components [Nordahl et al., 1997]. Changes may be unilateral or bilateral. It appears that there has been no comprehen-
22
Cells Tissues Organs 2003;174:17–25
sive published analysis of the ‘normal age-related changes’ which occur in the temporal and condylar sites. An excellent description of normal and diseased TMJ radiography during aging has, however, been presented in the thesis of Boering [1994]. The normal TMJ condyle is characterized by a round condyle with a smooth convex surface. Bony trabeculae of the condyle present a homogeneous radiographic image and may have a radial structure with trabeculae directed perpendicularly onto the surface. The articular eminence is also convex with a smooth surface. Boering [1994] reports that at onset of TMD (when clicking is the main symptom) radiographic abnormalities are minimal or absent. The diseased joint may be slightly flattened. At an intermediate stage of TMD (when clicking has developed into locking), remodeling is radiographically prominent, progressive and may lead to deformation of the joint in 1–2 years. He notes that small radiolucent areas with a cyst-like appearance can develop under the cortical plate of the condyle. Cortical lamellar bone can become resorbed, the condylar surface appears rough, and the normal structure of the bony spongiosa disappears. This change can progress rapidly between 1 and 2 years and leads to flattening of the condyle and eminence, to a shortened mandibular ramus, and narrowed joint space. Boering has also described the ‘terminal stage’ in which there is marked remodeling of the mandibular head with dense sclerotic bone on the surface of the condyle that is thicker than the cortical border of the normal joint. The well-aligned pattern of trabecular bone remains absent, and the trabecular pattern is unclear and sclerotic. Osteophytes on the articular eminence and condyle may be present and may cause lipping; if the osteophyte has broken off, a loose body may be noted. In rare cases a cyst-like radiolucency may remain. Flattening and reduced size of the condyle are prominent. Idiopathic condylar resorption is a poorly understood progressive disease that affects the TMJ. Wolford and Cardenas [2002] reported MRI imaging findings which showed extreme thinness and loss of continuity of cortical bone on the condylar head in the patients with idiopathic condylar resorption. These authors also noted that this was a progressive disease and claimed that early detection and surgical treatment could minimize the amount of condylar bone resorption. Smith et al. [1999] assessed 28 joints of patients with TMD and found that 24 of them showed signs of bony degeneration such as condylar flattening, cortical erosion and joint space alterations. Subtypes of juvenile chronic arthritis with resorption of TMJ condylar bone have been reported by Pedersen et al. [2001]. These authors report that TMJ involvement
Gruber/Gregg
was most frequent when disease onset was early; more severe resorption was seen in children aged 3–6 years compared to 6–9 years. Longer duration of the arthritis was statistically correlated to severity and risk of TMJ involvement. Their study analysis also showed that patients with polyarticular arthritic involvement showed significant severe bone destruction compared to monoarticular patients.
Osteoporosis: An Overview and Current Antiresorptive Treatments
Osteoporosis is a common metabolic disease whose frequency is increasing because the elderly constitute a rapidly growing segment of our population. It is characterized by a decrease in bone volume within a normal periosteal perimeter. It is a condition of skeletal fragility in which reduced bone mass and microarchitectural deterioration of bone are associated with increased risk of fracture [Heaney, 1998]. World Health Organization classifications use the term osteoporosis to define a bone mass value greater than 2.5 standard deviations below the young adult mean [Heaney, 1998]. As reviewed by Heaney, low bone mass has a multifactorial etiology which includes genetic predisposition [Eisman, 1999], failure to achieve peak bone mass during growth and development [Chesnut, 1991], excessive thinness, disuse, gonadal hormone deficiency, inadequate calcium and vitamin D intake, other medical factors (such as alcohol abuse [Abbott et al., 1994], corticosteroid use and smoking), nutrition [Mosekilde, 1992; Bunker, 1994] and lifestyle issues. The reader is referred to several recent reviews which have discussed the hormonal alterations in osteoporotic syndromes [Avioli, 1993] and the pathophysiology of osteoporosis in greater detail than space here allows [Johnston and Slemenda, 1995; Manolagas and Jilka, 1995; Manolagas et al., 1995; Heaney, 1998]. Osteoporosis is classified into two major groups. Primary osteoporosis includes postmenopausal osteoporosis in women (type I), age-related or ‘senile’ osteoporosis [Ebeling, 1998], and idiopathic osteoporosis in juveniles and young adults. Secondary osteoporosis [Gennari et al., 1998] includes osteoporosis which is secondary to heritable or acquired abnormalities; this is a broad disease grouping which includes bone loss associated with Marfan’s syndrome, Morquio’s syndrome, homocystinuria, osteogenesis imperfecta, adult hypophosphatasia, Werner’s syndrome, lactase deficiency, male hypogonadism, gut malabsorption, renal hypercalciuria, renal tubular aci-
Bone Resorption in Temporomandibular Joint Disorders
dosis (type 2), cirrhosis, immobilization, multiple myeloma, conditions associated with low serum phosphate, selective deficiency of 1,25-dihydroxyvitamin D (adult onset), anticonvulsant drug usage, female hypogonadism, Cushing’s syndrome, thyrotoxicosis, chronic alcoholism, diabetes, chronic heparin treatment, chronic obstructive pulmonary disease, systemic mastocytosis and mild vitamin D deficiency [Gruber and Baylink, 1981]. Treatment of secondary forms of osteoporosis is achieved by treating the primary cause or, as with glucocorticoid treatment [Manelli and Giustina, 2000], treatment with antiresorbers. As reviewed by Teitelbaum [2000], the most common form of osteoporosis is that due to decreased estrogen with menopause. Increased osteoclast numbers and increased bone resorption are associated with estrogen loss. Increased osteoclasts result from increases in a variety of cytokines which regulate this cell: RANKL, tumor necrosis factor alpha, interleukin 1, 6 and 11, M-CSF, prostaglandin E and osteoprotegerin [Jilka et al., 1992; Teitelbaum, 2000]. This brief discussion of osteoporosis has shown that there are many conditions which may make the individual at risk for significant bone loss. It is important to realize that osteoporosis-induced bone loss can be superimposed upon TMD bone loss. The following section reviews previous clinical reports which focus upon bone loss in TMD. Since cartilage is avascular, healthy subcondylar bone is important not only as a support for the cartilage but also since the vascularization of bone serves, along with synovial fluid, as a source of nutrition for chondrocytes [Malinin and Ovellette, 2000]. In addition to coexisting conditions listed above which may place the TMD patient at increased risk for osteoporosis, several additional physiologic factors are also important, such as age, family history of osteoporosis, nutritional status (which may be compromised for TMD patients with reduced calcium and vitamin D intake), lifestyle factors (such as smoking, exercise, and alcohol intake), concurrent medications which impact skeletal health (such as steroids or antiepileptic therapies) and for female patients, years since menopause and hysterectomy status. Arnett et al. [1996] have also suggested that attention be paid to pregnancy and lactation in respect to bone remodeling in the TMJ. Current antiresorptive treatments have recently been reviewed by Marcus et al. [2002]. Although this review focuses upon postmenopausal osteoporosis and clinical trials which used fractures as an endpoint, the reader is referred to this publication because it provides an excel-
Cells Tissues Organs 2003;174:17–25
23
lent current summary of pharmacologic therapies for osteoporosis, including estrogen and hormone replacement therapies, selective estrogen receptor modulators, bisphosphonates, salmon calcitonin nasal spray, and antiresorptive combination therapies.
Challenges
In the case reports presented above, two subjects presented with documented osteoporosis (cases 2 and 5). In Kirk’s report [2002] on risk factors for TMJ, osteoporosis was found to carry the highest risk of potential failure in initial surgical outcomes. Fourteen out of 28 patients identified with low bone density in this study experienced surgical failure. The authors suggested that preoperative diagnosis and evaluation of this risk factor, coupled with selection of the type of surgical procedure, would minimize the impact of osteoporosis on surgical outcome. These findings suggest that early detection and possible antiresorptive therapy for bone loss/osteoporosis merit future consideration in TMD therapy. This review of bone resorption in TMD also points to the need for further study of bone remodeling in the TMJ and investigations of degradative pathways in the TMJ. Puzas et al. [2001] have recently investigated cartilage degradative mechanisms in the TMJ of a transgenic mouse model with overexpression of tumor necrosis factor alpha), but there have been few basic science investigations into bone resorption in TMD. Yasuoka et al. [2000] investigated the effect of estrogen replacement on TMJ remodeling in ovariectomized rats and noted that in animal studies it is important to relate the region of study in the mandibular condyle to site-related differential growth due to feeding habits.
Tanguay et al. [1993] evaluated the effect of human TMJ synovial fluid on bone resorption in cultured mouse calvaria and found that in 5 out of 18 samples, bone resorption increased significantly over controls. Not all synovial fluid specimens which increased bone resorption were from patients with arthroscopically observable osteoarthritis, reflecting the diversity of TMJ disorders. It should be noted, however, that much more information is needed regarding the effect of the osteoarthritic disease process on TMD. Dijkgraaf et al. [1997, 1999] have contributed to this area by their studies of arthroscopically observed changes and histopathologic changes in the synovium of TMD patients. Their work identified an early proliferative phase and a later fibrous phase in the involved synovial membrane in osteoarthritis of the TMJ. Such findings are important to a better understanding of the epidemiology and natural progression of articular TMD [de Bont et al., 1997]. Studies such as those by Milam et al. [1998], which suggested that mechanical stresses may lead to the accumulation of damaging free radicals in the TMJ, point to the need for greater understanding of the extracellular matrix of the normal and diseased TMJ [Milam et al., 1991], and of the role of matrix metalloproteinase activity and the potential inhibition of MMPs by low-dose tetracyclines [Moses et al., 2001; Ryan et al., 2001]. In summary, this review has illustrated and reviewed the patterns of bone loss in TMD and highlighted the potential impact of osteoporosis in patients with both TMD and osteoporosis. Since bone resorption is important in TMD, and since osteoporosis poses a threat for approximately 55% of the US population aged 50 years and older, clinical consideration of the potential role of antiresorbing agents in TMJ bone health is timely and important.
References Abbott, L., J. Nadler, R.K. Rude (1994) Magnesium deficiency in alcoholism: Possible contribution to osteoporosis and cardiovascular disease in alcoholics. Alcoholism Clin Exp Res 18: 1076–1082. Arnett, G.W., S.B. Milam, L. Gottesman (1996) Progressive mandibular retrusion – Idiopathic condylar resorption. Part I. Am J Orthod Dentofacial Orthop 110: 8–15. Avioli, L.V. (1993) Hormonal alterations and osteoporotic syndromes. J Bone Miner Res 8(suppl 2): S511–S514. Bean, L.R., K.-Ä.Ö.T. Omnell (1977) Comparison between radiologic observations and macroscopic tissue changes in temporomandibular joints. Dentomaxillofac Radiol 6: 90–106.
24
Boering, G. (1994) Temporomandibular Joint Osteoarthrosis. A Clinical and Radiographic Investigation; thesis University of Groningen. Bunker, V.W. (1994) The role of nutrition in osteoporosis. Br J Biomed Sci 51: 228–240. Chesnut, C.H., 3rd (1991) Theoretical overview: Bone development, peak bone mass, bone loss, and fracture risk. Am J Med 91: 2S–4S. de Bont, L.G., L.C. Dijkgraaf, B. Stegena (1997) Epidemiology and natural progression of articular temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 83: 72–76.
Cells Tissues Organs 2003;174:17–25
Dijkgraaf, L.C., R.S. Liem, L.G. de Bont (1997) Synovial membrane involvement in osteoarthritic temporomandibular joints: A light microscopic study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 83: 373–386. Dijkgraaf, L.C., R.S. Liem, L.T. van der Weele, L.B. de Bont (1999) Correlation between arthroscopically observed changes in synovial light microscopic findings in osteoarthritic temporomandibular joints. Int J Oral Maxillofac Surg 28: 83–89. Ducy, P., T. Schinke, G. Karsenty (2000) The osteoblast: A sophisticated fibroblast under central surveillance. Science 289: 1501–1504.
Gruber/Gregg
Ebeling, P.R. (1998) Osteoporosis in men. New insights into aetiology, pathogenesis, prevention, and management. Drugs Aging 13: 421–434. Eisman, J.A. (1999) Genetics of osteoporosis. Endocr Rev 20: 788–804. Flygare, L., H. Hosoki, A. Petersson, M. Rohlin, S. Akerman (1997) Bone volume in human temporomandibular autopsy joints with and without erosive changes. Acta Odontol Scand 55: 167–172. Flygare, L., M. Rohlin, S. Åkerman (2002) Microscopy and tomography of erosive changes in the temporomandibular joint. An autopsy study. Acta Odontol Scand 53: 297–303. Gennari, C., G. Martini, R. Nuti (1998) Secondary osteoporosis. Aging Clin Exp Res 10: 214–224. Gruber, H.E., D.J. Baylink (1981) The diagnosis of osteoporosis. J Am Geriatr Soc 29: 490–497. Gruber, H.E., J.L. Ivey, E.R. Thompson, C.H.I. Chesnut, D.J. Baylink (1986) Osteoblast and osteoclast cell number and cell activity in postmenopausal osteoporosis. Miner Electrolyte Metab 12: 246–254. Hansson, T., T. Öberg (1977) Arthrosis and deviation in form in the temporomandibular joint. Acta Odontol Scand 35: 167–174. Heaney, R.P. (1998) Pathophysiology of osteoporosis. Endocrinol Metab Clin North Am 27: 255– 265. Hofbauer, L.C., F. Gori, B.L. Riggs, D.L. Lacey, C.R. Dunstan, T.C. Spelsberg, S. Khosla (1999) Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: Potential paracrine mechanisms of glucocorticoid-induced osteoporosis. Endocrinology 140: 4382–4389. Jilka, R.L., G. Hangoc, G. Firasole, G. Passeri, D.C. Williams, J.S. Abrams, B. Boyce, H. Broxmeyer, S.C. Manolagas (1992) Increased osteoclast development after estrogen loss: Mediation by interleukin-6. Science 257: 88–91. Johnston, C.C., Jr., C.W. Slemenda (1995) Pathogenesis of osteoporosis. Bone 17(suppl): 19S– 22S. Khosla, S. (2002) Minireview: The OPG/RANKL/ RANK system. Endocrinology 142: 5050– 5055.
Bone Resorption in Temporomandibular Joint Disorders
Kirk, W.S., Jr. (2002) Risk factors and initial surgical failures of TMJ arthrotomy and arthroplasty: A four to nine year evaluation of 303 surgical procedures. J Craniomandibular Pract 16: 154–161. Malinin, T., E.A. Ouellette (2000) Articular cartilage nutrition is mediated by subchondral bone: A long-term autograft study in baboons. Osteoarthritis Cartilage 8: 483–491. Manelli, F., A. Giustina (2000) Glucocorticoidinduced osteoporosis. TEM 11: 79–85. Manolagas, S.C., T. Bellido, R.L. Jilka (1995) New insights into the cellular, biochemical, and molecular basis of postmenopausal and senile osteoporosis: Roles of IL-6 and gp130. Int J Immunopharmacol 17: 109–116. Manolagas, S.C., R.L. Jilka (1995) Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med 332: 305–311. Marcus, R., M. Wong, H. Heath 3rd, J.L. Stock (2002) Antiresorptive treatment of postmenopausal osteoporosis: Comparison of study designs and outcomes in large clinical trials with fracture as an endpoint. Endocr Rev 23: 16– 37. Milam, S.B., R.J. Klebe, R.G. Triplett, D. Herbert (1991) Characterization of the extracellular matrix of the primate temporomandibular joint. J Oral Maxillofac Surg 49: 381–391. Milam, S.B., G. Zardeneta, J.P. Schmitz (1998) Oxidative stress and degenerative temporomandibular joint disease: A proposed hypothesis. J Oral Maxillofac Surg 56: 214–223. Mosekilde, L. (1992) Osteoporosis and calcium. J Intern Med 231: 145–149. Moses, O., C.E. Nemcovsky, H. Tal, R. Zohar (2001) Tetracycline modulates collagen membrane degradation in vitro. J Periodontol 72: 1588–1593. Nordahl, S., P. Alstergren, A. Appelgran, B. Appelgren, S. Eliasson, S. Kopp (1997) Pain, tenderness, mandibular mobility, and anterior open bite in relation to radiographic erosions in temporomandibular joint disease. Acta Odontol Scand 55: 18–22. Öberg, T., C.E. Carlsson, C.M. Fajers (1971) The temporomandibular joint. A morphologic study on a human autopsy material. Acta Odontol Scand 29: 349–384. Pedersen, T.K., J.J. Jensen, B. Melsen, T. Herlin (2001) Resorption of the temporomandibular condylar bone according to subtypes of juvenile chronic arthritis. J Rheumatol 28: 2109– 2115.
Puzas, J.E., J.-M. Landeau, R. Tallents, J. Albright, E.M. Schwartz, R. Landesberg (2001) Degradative pathways in tissues of the temporomandibular joint. Cells Tissues Organs 169: 248– 256. Roodman, G.D. (1996) Advances in bone biology: The osteoclast. Endocr Rev 17: 308–332. Ryan, M.E., A. Usman, N.S. Ramamurthy, L.M. Golub, R.A. Greenwald (2001) Excessive matrix metalloproteinase activity in diabetes: Inhibition by tetracycline analogues with zinc reactivity. Curr Med Chem 8: 305–316. Smith, J.A., N.A. Sandler, W.H. Ozaki, T.W. Braun (1999) Subjective and objective assessment of the temporalis myofascial flap in previously operated temporomandibular joints. J Oral Maxillofac Surg 57: 1058–1065. Tanguay, D., R.D. Schwarz, C.S. Greene, D.P. Forbes, H.T. Perry, P. Lakatos, P.H. Stern (1993) Temporomandibular joint synovial fluid effect on resorption of mouse calvarial bone in vitro. Northwestern Dent Res 4: 11– 16. Teitelbaum, S.L. (2000) Bone resorption by osteoclasts. Science 289: 1504–1508. Teti, A., P.C. Marchisio, A.Z. Zallone (1991) Clear zone in osteoclast function: Role of podosomes in regulation of bone-resorbing activity. Am J Physiol 261: C1–C7. Udagawa, N., N. Takashashi, T. Akatsu, H. Tanaka, T. Sasaki, T. Nishihara, T. Koga, T.J. Martin, T. Suda (1990) Origin of osteoclasts: Mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc Natl Acad Sci USA 87: 7260–7264. Väänäen, H.K., M. Horton (1995) The osteoclast clear zone is a specialized cell-extracellular matrix adhesion structure. J Cell Sci 108: 2729– 2732. Wolford, L.M., L. Cardenas (2002) Idiopathic condylar resorption: Diagnosis, treatment protocol, and outcomes. Am J Orthod Dentofacial Orthop 116: 667–677. Yasuoka, T., M. Nakashima, T. Okuda, N. Tatematsu (2000) Effect of estrogen replacement on temporomandibular joint remodeling in ovariectomized rats. J Oral Maxillofac Surg 58: 189– 196.
Cells Tissues Organs 2003;174:17–25
25
Cells Tissues Organs 2003;174:26–33 DOI: 10.1159/000070572
Nitric Oxide in Experimental Joint Inflammation Benefit or Detriment?
S.M. Wahl a N. McCartney-Francis a J. Chan a R. Dionne b L. Ta b J.M. Orenstein c a Oral Infection and Immunity Branch and b Pain and Neurosensory Mechanisms Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, Md., and c Department of Pathology, George Washington University, Washington, D.C., USA
Key Words Nitric oxide W Inflammation W Arthritis W Tumor necrosis factor W Temporomandibular joint disorder
Abstract The host response to infection or injury initiates a cascade of events involving recruitment of leukocytes and the release of multiple inflammatory mediators. One of these mediators, nitric oxide (NO), not only represents an important microbicidal agent in host defense, but also functions as a biological signaling and effector molecule in inflammation and immunity. However, overproduction of NO can be autotoxic and contribute to tissue damage and has been implicated in pathogenesis of tumors, and infectious, autoimmune and chronic degenerative diseases. NO is generated via constitutive and inducible nitric oxide synthases (iNOS) which catalyze the oxidation of a guanidino nitrogen associated with L-arginine. Whereas endothelial NOS (eNOS) and neuronal NOS (nNOS) are constitutively expressed, iNOS is transcriptionally induced by bacterial constituents and inflammatory mediators, including TNF· and IL-1. In an experimental model of bacterial component-induced joint inflammation and tissue degradation, functionally distinct roles of the constitutive NOS and iNOS were demon-
ABC Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2003 S. Karger AG, Basel
Accessible online at: www.karger.com/cto
strated. Following systemic delivery of an arthritogenic dose of streptococcal cell walls (SCW), these bacterial peptidoglycan-polysaccharide complexes disseminate and target the peripheral joints, liver and spleen of the treated animals. Following deposition of the SCW in the peripheral joints, an initial innate inflammatory response
Abbreviations used in this paper
AG cNOS COX eNOS IL-1ß iNOS L-NIL NMMA nNOS NO NSAIDs rAAV SCW Th1 TMD TNF· TNFR:Fc
aminoguanidine constitutive nitric oxide synthase cyclooxygenase endothelial nitric oxide synthase interleukin-1ß inducible nitric oxide synthase N-iminoethyl-L-lysine NG-monomethyl-L-arginine neuronal nitric oxide synthase nitric oxide nonsteroidal anti-inflammatory drugs recombinant adenoassociated viral vector streptococcal cell wall T helper type 1 temporomandibular joint disorders tumor necrosis factor-· tumor necrosis factor receptor-Fc fusion construct
Sharon M. Wahl Building 30, Room 320 30 Convent Dr., MSC 4352, NIDCR, NIH Bethesda, MD 20892-4352 (USA) Tel. +1 301 496 4178, Fax +1 301 402 1064, E-Mail
[email protected]
to the bacterial components progresses into an adaptive immune response with the recruitment and activation of mononuclear phagocytes and T lymphocytes. With the release of cytokines and inflammatory mediators, there is an upregulation of gene expression for iNOS, but not the constitutive nNOS or eNOS. Nonetheless, the constitutive NOS isoforms, regulated by calcium fluxes and interaction with calmodulin, may also enhance NO production. Increased release of NO was detected not only in the synovium, but also in the circulation, and plasma levels of nitrate plus nitrite, the stable products of NO reactions, correlated with disease progression. Following inhibition of NO production with nonspecific NOS inhibitors, such as NG-monomethyl-L-arginine, which target all three isoforms, there is a striking therapeutic benefit with reduced signs and symptoms of erosive arthritis. In contrast, selective targeting of iNOS with Niminoethyl-L-lysine resulted in exacerbation of the synovial inflammation and degradation of joint structures. Based on these data, it appears that the constitutive isoforms of NOS contribute to the pathophysiology of the arthropathy, and that induced NOS and NO may function, in part, in a protective pathway. Moreover, the suppression of NO following treatment with TNF· antagonists results in reduced inflammation and the associated synovial pathology. Collectively, these data implicate discrete roles for the NOS isoforms in the emergence of local tissue pathology and underscore the need to define the specific pathways that are being targeted for interventional strategies.
pain and levels of TNF· in synovial fluid of TMD [Shafer et al., 1994], and TNF· levels were higher in TMD with bony changes [Fu et al., 1995]. Since TNF· is one of multiple inducers of nitric oxide [Bogdan, 2001], it was anticipated and demonstrated that NO was elevated in affected synovial tissues of TMD, as in other arthritides [Homma et al., 2001], and might correlate with joint pain [Suenaga et al., 2001] and/or degenerative changes [Takahashi et al., 1996, 1999]. Disruption of the inflammatory cascade with nonsteroidal anti-inflammatory agents (NSAIDs), cyclooxygenase 2 (COX2)-selective inhibitors, and/or surgical intervention, which may include repair, removal or replacement of the intra-articular disc have had limited success [Ta et al., 2002]. Cytokine-specific antagonists such as TNF· inhibitors, effective in rheumatoid arthritis [Graninger and Smolen, 2002], may offer a relevant approach to disengage the inflammatory pathogenesis in TMD.
NO and Nitric Oxide Synthase Isoforms in Immunopathology
Temporomandibular joint disorders (TMD) may afflict up to 12% of the US population [Carlsson and LaResche, 1995] either independently or as part of a constellation of joint pathologies associated with rheumatoid arthritis [Holmlund et al., 1992]. Immunopathology of TMD often parallels that observed in other synovial joints with infiltration of inflammatory cells, production of cytokines [i.e. tumor necrosis factor-· (TNF·) and interleukin-1ß (IL-1ß)] and inflammatory mediators such as nitric oxide (NO) and matrix metalloproteases with subsequent bone resorption [reviewed in Kacena et al., 2001]. The generation of these cytokines and inflammatory mediators likely represent key events in the cascade leading to connective tissue destruction. In this regard, a significant correlation has been reported between severity of
For more than a decade, the prevailing paradigm for NO in immunobiology has revolved around its induction in phagocytic cells coupled with its antimicrobial and tumoricidal functions. NO is a short-lived free radical gas, which is synthesized through the five-electron oxidation of one of the guanidino nitrogens associated with L-arginine in the presence of requisite cofactors [MacMicking et al., 1997]. NO production by activated macrophages is orchestrated by an inducible nitric oxide synthase (iNOS, NOS2) in response to cytokines and microbes or their products [Nathan, 1997; Bogdan, 2001]. Although the half-life of NO is measured in seconds, the sustained generation of high levels of NO by iNOS can lead to production of reactive nitrogen species which upon reaction with oxygen can result in nitration of cellular proteins. In particular, the S-nitrosation of free thiols in plasma or tissue significantly increases the half-life of NO, effectively creating a storage form of NO. Whereas low levels of NO appeared critical in host defense, excess NO is toxic and blamed for chronic immunopathology. As continued exploration of the origins, regulation and functions of this small, highly diffusible radical gas revealed new insights, there has been, as is often the case, a paradigm shift. Through the availability of NOS inhibitors, NO donors and NOS-deficient mice, it has become evident that in the immune system, NO plays a much more diverse role than initially conceived. NO governs a multiplicity of cellular
NO in Experimental Joint Inflammation
Cells Tissues Organs 2003;174:26–33
Copyright © 2003 S. Karger AG, Basel
Introduction
27
processes including proliferation, differentiation, production of cytokines, expression of adhesion and costimulatory molecules, generation of matrix and apoptosis. Whereas iNOS may be detrimental in some immune responses, it paradoxically may also exhibit protective functions in autoimmunity [Kubes, 2000]. Besides iNOS, there are two other isoforms of the enzyme which catalyze the oxidation of L-arginine to NO and L-citrulline: constitutively expressed neuronal NOS (nNOS, NOS1) and endothelial NOS (eNOS, NOS3). Both nNOS and eNOS exist as preformed proteins within the cells, are calcium/calmodulin-dependent with eNOS bound to the caveolin scaffolding in plasmalemma vesicles (caveolae) and nNOS interacting with N-methyl-Daspartate receptors. The constitutive isoforms, once considered restricted to neurons and endothelial cells, are now known to be the products of multiple cell types, to be regulatable, and to contribute to immune response pathways, including antimicrobial activity, apoptosis, cell adhesion and/or autoimmune pathogenesis [Salerno et al., 2002]. Collectively, iNOS and constitutive NOS (cNOS) drive many diverse responses underlying both normal and aberrant host defense [see review, Abramson et al., 2001; Bogdan, 2001].
Experimental Model of Joint Pathology
In order to explore these potentially dichotomous roles for NOS and NO in the development of acute and chronic inflammation and tissue destruction and the merit of nonspecifically or specifically antagonizing them to alleviate synovial pathology, we utilized an experimental model of peripheral arthritis triggered by microbial products. While not a model of TMD, this arthritis model is characterized by acute and chronic synovial inflammation and tissue damage and provides insight into cellular and molecular pathways which underlie joint pathology. Intraperitoneal delivery of peptidoglycan-polysaccharide components of the cell walls of group A streptococci results in dissemination of the bacterial products to three primary target tissues, the spleen, liver and peripheral joints [Hines et al., 1993; McCartney-Francis et al., 1993]. Deposition of these proinflammatory, immunogenic and degradation-resistant bacterial constituents in the peripheral joints results in a rapid inflammatory response in the synovium, with adhesion, recruitment and infiltration of leukocytes, local release of inflammatory mediators, and tissue swelling. After several days, a remission phase is evident clinically and microscopically, which is followed
28
Cells Tissues Organs 2003;174:26–33
by recruitment of lymphocytes and macrophages that are instrumental in the subsequent synovial hyperplasia and erosive destruction of the subchondral and periarticular bone and cartilage. A predominant T helper type 1 (Th1) lymphocyte response dominated by IL-2, IFNÁ and TNF· drives the immunopathogenesis, and the paucity of antiinflammatory Th2-type cytokines, such as IL-4, IL-10 and IL-5, is underscored by the ability of exogenous IL-4 to reverse the chronic autoimmune-like inflammation [Allen et al., 1993].
TNF· Antagonist Inhibits NO and Joint Pathology
Multiple proinflammatory cytokines and inflammatory mediators have been implicated in the evolution of this progressive, chronic destructive lesion in the synovium and adjacent bone and cartilage structures. One of the central cytokines associated with emergent pathology in the SCW-induced arthritis model is TNF· and a dramatic reversal of the persistent destructive phase of the disease occurs when the animals are treated with a TNF· antagonist [Chan et al., 2002]. The p75 TNF receptor (TNFR) linked to the Fc region of immunoglobulin G, which stabilizes the molecule, binds to TNF· and neutralizes its activity by preventing it from binding to the cell surface TNF· receptor and has been successfully used in human rheumatoid disease [reviewed in Graninger and Smolen, 2002]. In our studies, this dimeric fusion protein, TNFR:Fc, was delivered by gene transfer, either in a plasmid vector or in a recombinant adenoassociated viral vector (rAAV). Delivery of the rAAV-TNFR:Fc either intramuscularly for systemic delivery or directly into the afflicted joint (intra-articular) resulted in increased circulating TNFR, reduced circulating TNF· and consequently, the streptococcal cell wall (SCW)-induced arthritis was suppressed clinically and at the cellular and molecular levels [Chan et al., 2002] (fig. 1). Intra-articular delivery resulted in significantly less systemic TNFR:Fc which may minimize potential side effects of this TNF· antagonist. Nonetheless, whether the generation of TNF· or its actions are disrupted by TNFR:Fc or with systemic delivery of human chorionic gonadotropin, secretory leukocyte inhibitor (SLPI) or TGF-ß [Song et al., 1998, 1999, 2000], the propagation of the synovial pathology can be interrupted. Since TNF· is a pivotal upstream trigger of a cascade of proinflammatory molecules, inhibition of its activity also blocked downstream mediators, such as NO (fig. 2).
Wahl/McCartney-Francis/Chan/Dionne/ Ta/Orenstein
Fig. 1. Gene transfer of TNFR:Fc suppresses SCW-induced arthritis.
Female Lewis rats receiving an arthritogenic dose of group A SCW were untreated or treated with TNFR-immunoglobulin G Fc fusion construct (TNFR:Fc) in an AAV vector (rAAV-TNFR:Fc) [Chan et al., 2002]. A Histopathology of a representative joint from an untreated SCW arthritic animal demonstrating synovial hyperplasia, inflammatory cells, pannus formation, joint space narrowing and
destruction of bone and cartilage. B Joint from an arthritic animal receiving rAAV-TNFR:Fc appears nearly normal. C Articular indices during acute (day 5 post-SCW) and chronic (day 33 post-SCW) arthritis in animals receiving rAAV-TNFR:Fc on day 0. Although TNFR:Fc has little effect during acute disease, it dramatically suppresses chronic erosive arthritis.
In our earlier studies, we documented a role for iNOS and NO in the arthropathology of the SCW model of erosive arthritis [McCartney-Francis et al., 1993, 1999,
2001]. The bacterial cell wall products directly induce iNOS and NO synthesis in macrophages as part of an innate immune response, evident in vitro [McCartneyFrancis et al., 2001] and in vivo, since iNOS and NO were also detected at the molecular and protein levels within
NO in Experimental Joint Inflammation
Cells Tissues Organs 2003;174:26–33
Effect of NO Inhibitors on Arthritis
29
Fig. 2. TNF antagonist (TNFR:Fc) inhibits NO. Serum from control,
SCW arthritic, and arthritic animals treated with rAAV-TNFR:Fc was assayed for nitrite and nitrate levels with an NO analyzer. Serum levels of the NO stable end products, nitrite and nitrate, in the treated animals were significantly reduced compared to untreated arthritic animals (* p ! 0.0001).
the acutely and chronically inflamed synovium. Moreover, elevated levels of the metabolites of NO, nitrite and nitrate, were detected in the serum of arthritic animals, but not in control animals (fig. 2), not only providing evidence of excess levels of these molecules, but also perhaps a prognostic marker for tissue disease. NO has multiple cytotoxic effects, including DNA damage, induction of apoptosis, vascular permeability, aberrant reactive oxygen intermediate metabolism and proinflammatory signal cascades which could favor an autotoxic response. Thus, it seemed a likely target for therapeutic intervention and to this end, animals were treated with a nonspecific inhibitor of NOS, NG-monomethyl-L-arginine (NMMA), which is an arginine analogue that competitively suppresses both cNOS and iNOS isoforms. A dramatic protective role was evident in those animals receiving NMMA, with reduced inflammatory cell infiltrates, suppressed cytokine production, and inhibition of the distinctive cartilage and bone erosions characteristic of this disease model. Importantly, even if the intravenous NMMA treatment was initiated up to 12 days after the initiation of arthritis in this model, a significant benefit was evident [McCartney-Francis et al., 1993]. These data documented that NO is a major contributing factor to the immunologically mediated joint destruction induced by the bacterial cell walls.
30
Cells Tissues Organs 2003;174:26–33
Emboldened by these positive effects of nonspecific inhibition of NOS, we explored alternative agents that might suppress the high levels of NO in the joints. One such inhibitor, hemoglobin, serves as a scavenger of NO by binding to its heme-containing domain, thereby blocking its activity. If hemoglobin could deplete excess NO, it might reduce inflammation and injury in synovial tissue from joints with inflammatory arthritis. Hemoglobin, delivered systemically (intraperitoneal injection) to animals at the onset of disease or during the development of the chronic phase of SCW-induced synovitis, also effectively reduced NO levels as detected in the circulation and in the locally inflamed synovium to ameliorate the disease process, and provided further evidence of an autotoxic role for NO in chronic inflammation and autoimmune-like lesions [McCartney-Francis et al., 1999]. Because NMMA and hemoglobin are nonspecific inhibitors of NOS, targeting the constitutive isoforms as well as the iNOS, it was reasonable to assume that a more specific targeting of iNOS might leave the presumed physiological and necessary actions of eNOS and nNOS intact while blocking the high levels of induced NO associated with activation of macrophages. Whereas interference with eNOS and nNOS can be beneficial by reducing blood flow to the joint as well as pain [Grisham et al., 2002], inhibition of these constitutively expressed isoforms could conceivably alter cell-cell communication, vascular tone and neurotransmission. To this end, the NOS inhibitor, aminoguanidine (AG), while not completely specific for iNOS, has a higher affinity for iNOS than cNOS [Salerno et al., 2002], and N-iminoethyl-L-lysine (L-NIL), which even more selectively targets iNOS [Moore et al., 1994], were compared for their efficacy in ameliorating arthropathogenesis. L-NIL, like AG and NMMA, inhibited nitrite production by SCW-stimulated macrophages in culture, and whether delivered orally or by the intraperitoneal route reduced circulating levels of nitrite in blood. Unanticipated, however, was the ineffective reduction in arthritis by AG and the striking exacerbation of arthritis in animals treated with L-NIL [McCartney-Francis et al., 2001]. Whereas NMMA and hemoglobin were associated with reduced inflammatory cell infiltrates, a decrease in cytokine production, reduced synovial hyperplasia and ablation of cartilage and bone matrix erosion, animals receiving L-NIL had a paradoxical enhancement of inflammation and massive tissue destruction.
Wahl/McCartney-Francis/Chan/Dionne/ Ta/Orenstein
Identification of iNOS, eNOS and nNOS in Inflamed Synovium
Because a specific inhibitor of iNOS was therapeutically less effective than the nonspecific antagonists which inhibit both iNOS and cNOS, we reasoned that either cNOS and eNOS were contributing to the autotoxic response in the synovium and/or that iNOS had a beneficial functional component beyond its expected detrimental role in joint disease as would have been predicted by the prevailing paradigm of cNOS and NO favoring beneficial physiologic processes, whereas iNOS and NO, in excess, promote pathology. In fact, both of these hypotheses may come into play within the synovial lesions. By immunohistochemical analysis, iNOS was readily detected in arthritic lesions, predominantly in macrophages, but some chondrocytes and synovial lining cells were also positive. After treatment with L-NIL, the number of positive cells was reduced, although some staining persisted in the proximal bone. Surprisingly, parallel staining for eNOS and nNOS revealed that they were also expressed in the synovium of arthritic, but not control animals. In addition to eNOS staining in the endothelial cells, multinucleated osteoclasts were eNOS-positive, but did not stain for iNOS or nNOS. By comparison, nNOS was in the bone and also in infiltrating leukocytes, clearly not restricted to neuronal cells [McCartney-Francis et al., 2001]. After L-NIL treatment, no reduction in either nNOS nor eNOS was evident, even though iNOS was clearly reduced. Interestingly, eNOS mRNA was actually enhanced in the joint tissues of arthritic rats following LNIL treatment. Moreover, the inflammatory cells themselves were contributing to the pool of NO not only via iNOS, but also through the constitutive isoforms of NOS. These observations implicate the ‘constitutive’ isoforms in the sequelae of chronic synovial inflammation and tissue damage, emphasizing that the roles of cNOS and iNOS are clearly context-dependent [Singh et al., 2000].
favoring Th1 has been reported [Fiorucci, 2001]. This enhanced Th1 response has been linked to the absence of NO which otherwise inhibits caspase 1. By posttranslational nitrosation and inactivation of cysteine proteases, such as the IL-1ß-converting enzyme (ICE/caspase 1), NO can interrupt pro-cytokine processing and the downstream consequences of active IL-1ß release [Fiorucci, 2001]. Without inhibition, caspase 1 cleaves the inactive pro-form of IL-1ß and other pro-cytokines (IL-18) into their biologically active form to initiate and perpetuate the inflammation [Salvesen and Dixit, 1997]. In vitro, NO downregulates IL-1ß, IL-2, IL-12, IL-18 and INFÁ [Kolb and Kolb-Bachofen, 1998]. Reflecting these observations, the iNOS null mice are more susceptible, rather than resistant to, both adjuvant and septic arthritis [Gilkeson et al., 1997; McInnes et al., 1998]. Conversely, endogenous NO or NO-releasing NSAIDs (NO-NSAIDs) suppress caspase 1-dependent Th1 cytokine cascades, as well as inhibit platelet aggregation [Fiorucci, 2001]. NO-NSAIDs, which are a new class of compounds obtained by adding a nitroxybutyl moiety to a conventional NSAID, are more effective than their parent compounds [Fiorucci, 2001] because they not only inhibit COX, but also have COX-independent actions involving NO. Thus, ablation of NO may well promote a Th1 response and although counterintuitive, may enhance rather than inhibit synovial pathology. These and other recent provocative studies have uncovered an otherwise unappreciated role for the constitutive and inducible NOS isoforms in the evolution of acute and chronic inflammatory pathology, which may be important in the design of therapeutic agents.
Summary and Conclusion
A new body of literature is also emerging which does, in fact, point to evidence that NO can have beneficial properties in the regulation of an inflammatory response. Of particular relevance is the recent documentation that NO is a potent inhibitor of Th1 lymphocytes [Fiorucci et al., 2000], the likely perpetrators of SCW arthritis and many other chronic destructive diseases. Also, in the absence of iNOS (iNOS null mice), a shift to an imbalance
One such chronic inflammatory syndrome includes TMD in which NOS and NO have been identified both in the synovial fluid [Takahashi et al., 1996, 1999; Suenaga et al., 2001] and in the synovial tissues of individuals experiencing pain and tissue pathology [Homma et al., 2001]. In this regard, the histopathology associated with TMD, particularly in individuals who have had surgical intervention and disc replacement, includes mononuclear cell infiltrates and the presence of multinucleated giant cells (fig. 3) [Trumpy et al., 1996], typical of chronic inflammation. Not only is there evidence of NOS in these diseased tissues, but it is also possible to detect enhanced circulating levels of nitrite/nitrate in the blood of TMD patients (fig. 4). In a limited number of individuals, repre-
NO in Experimental Joint Inflammation
Cells Tissues Organs 2003;174:26–33
A Beneficial Side to NO
31
3
Fig. 3. Inflammatory cells in synovial tissue from
TMD patient. Histopathology of synovial tissue obtained at the time of surgical excision of implant material reveals mononuclear and multinucleated giant cells (arrows) observed by light microscopy (A, B) and electron microscopy (C). Fig. 4. Nitrite and iNOS in TMD. A Plasma samples obtained from TMD patients who have had disc implants (n = 6) were evaluated by NO analyzer for nitrite and nitrate levels and compared to plasma samples from control subjects (n = 3). p ! 0.05. B Immunohistochemical staining for iNOS in a multinucleated giant cell in TMD synovial tissue.
32
Cells Tissues Organs 2003;174:26–33
4
Wahl/McCartney-Francis/Chan/Dionne/ Ta/Orenstein
senting a subset of a larger investigation [Ta et al., 2002], levels of nitrite/nitrate measured in the serum were significantly higher than in control subjects (fig. 4), reflective of tissue inflammation. Although iNOS has been identified in the temporomandibular joint tissues [Homma et al., 2001], limited studies have been done to date to characterize which, if any, of the other isoforms of NOS are
represented locally in the TMD tissues. This essential information will provide important insights into the potential for targeting NO/NOS in therapeutic strategies for patients suffering from TMD. Clearly, the emergent complexities of NO regulation require diligence in assessment and manipulation of NO via the NOS isoforms in the pathogenesis of chronic joint diseases.
References Abramson, S., A. Amin, R. Clancy, M. Attur (2001) The role of nitric oxide in tissue destruction. Best Pract Res Clin Rheumatol 15: 831–845. Allen, J., H. Wong, G. Costa, M. Bienkowski, S. Wahl (1993) Suppression of monocyte function and differential regulation of IL-1 and IL-1ra by IL-4 contribute to resolution of experimental arthritis. J Immunol 151: 4344–4351. Bogdan, C. (2001) Nitric oxide and the immune response. Nat Immunol 2: 907–916. Carlsson, G., L. LeResche (1995) Epidemiology of temporomandibular disorders; in Sessle, B., P. Bryant, R. Dionne (eds): Temporomandibular Disorders and Related Pain Conditions. Seattle, IASP Press, pp 211–226. Chan, J., G. Villarreal, W. Jin, T. Stepan, H. Burstein, S. Wahl (2002) Intraarticular gene transfer of TNFR:Fc suppresses experimental arthritis with reduced systemic distribution of the gene product. Mol Ther 6: 727–736. Fiorucci, S. (2001) NO-releasing NSAIDS are caspase inhibitors. Trends Immunol 22: 232–235. Fiorucci, S., L. Santucci, E. Antonelli, E. Distrutti, G. del Sero, O. Morelli, L. Romani, B. Federici, P. del Soldato, A. Morelli (2000) NO-aspirin protects from T cell-mediated liver injury by inhibiting caspase-dependent processing of Th1-like cytokines. Gastroenterology 118: 404–421. Fu, K., X. Ma, Z. Zhang, W. Chen (1995) Tumor necrosis factor in synovial fluid of patients with temporomandibular disorders. J Oral Maxillofac Surg 53: 424–426. Gilkeson, G., J. Mudgett, M. Seldin, P. Ruiz, A. Alexander, M. Misukonis, D. Pisetsky, J. Weinberg (1997) Clinical and serologic manifestations of autoimmune disease in MRL-1pr/ 1pr mice lacking nitric oxide synthase type 2. J Exp Med 186: 365–373. Graninger, W., J. Smolen (2002) Treatment of rheumatoid arthritis by TNF-blocking agents. Int Arch Allergy Immunol 127: 10–14. Grisham, M.B., K.P. Pavlick, F.S. Laroux, J. Hoffman, S. Bharwani, R.E. Wolf (2002) Nitric oxide and chronic gut inflammation: Controversies in inflammatory bowel disease. J Invest Med 50: 272–283. Hines, K., M. Christ, S. Wahl (1993) Cytokine regulation of the immune response: An in vivo model. Immunomethods 3: 13–22.
NO in Experimental Joint Inflammation
Holmlund, A., G. Gynther, F. Reinholt (1992) Rheumatoid arthritis and disk derangement of the temporomandibular joint. A comparative arthroscopic study. Oral Surg Oral Med Oral Pathol 73: 273–277. Homma, H., T. Takahashi, H. Seki, M. Ohtani, T. Kondoh, M. Fukuda (2001) Immunohistochemical localization of inducible nitric oxide synthase in synovial tissue of human temporomandibular joints with internal derangement. Arch Oral Biol 46: 93–97. Kacena, M., G. Merrel, S. Konda, K. Wilson, Y. Xi, M. Horowitz (2001) Inflammation and bony changes at the temporomandibular joint. Cells Tissues Organs 169: 257–264. Kolb, H., V. Kolb-Bachofen (1998) Nitric oxide in autoimmune disease: Cytotoxic or regulatory mediator? Immunol Today 19: 556–561. Kubes, P. (2000) Inducible nitric oxide synthase: A little bit of good in all of us. Gut 47: 6–9. MacMicking, J., Q. Xie, C. Nathan (1997) Nitric oxide and macrophage function. Ann Rev Immunol 15: 323–250. McCartney-Francis, N., J. Allen, D. Mizel, J. Albina, Q. Xie, C. Nathan, S. Wahl (1993) Suppression of arthritis by an inhibitor of nitric oxide synthase. J Exp Med 178: 749–754. McCartney-Francis, N., X. Song, D. Mizel, C. Wahl, S. Wahl (1999) Hemoglobin protects from streptococcal cell wall-induced arthritis. Arthritis Rheum 42: 1119–1127. McCartney-Francis, N., X. Song, D. Mizel, S. Wahl (2001) Selective inhibition of inducible nitric oxide synthase exacerbates erosive joint disease. J Immunol 166: 2734–2740. McInnes, I., B. Leung, X. Wei, C. Gemmell, F. Liew (1998) Septic arthritis following Staphylococcus aureus infection in mice lacking inducible nitric oxide synthase. J Immunol 160: 308– 315. Moore, W., R. Webber, G. Jerome, F. Tjoeng, T. Misko, M. Currie (1994) L-N6-(1-iminoethyl)lysine: A selective inhibitor of inducible nitric oxide synthase. J Med Chem 37: 3886– 3888. Nathan, C. (1997) Perspective series: Nitric oxide and nitric oxide synthases. J Clin Invest 100: 2417–2423. Salerno, L., V. Sorrenti, C. Di Giacomo, G. Romeo, M. Siracusa (2002) Progress in the development of selective nitric oxide synthase (NOS) inhibitors. Curr Pharm Des 8: 177–200. Salvesen, G., V. Dixit (1997) Caspases: Intracellular signaling by proteolysis. Cell 91: 443–446.
Shafer, D., L. Assael, L. White, E. Rossomando (1994) Tumor necrosis factor-alpha as a biochemical marker of pain and outcome in temporomandibular joints with internal derangements. J Oral Maxillofac Surg 52: 786–791. Singh, V., S. Mehrotra, P. Narayan, C. Pandey, S. Agarwal (2000) Modulation of autoimmune diseases by nitric oxide. Immunol Res 22: 1– 19. Song, X., G. MiLi, W. Jin, D. Klinman, S. Wahl (1998) Plasmid DNA encoding transforming growth factor-ß1 suppresses chronic disease in a streptococcal cell wall-induced arthritis model. J Clin Invest 101: 2615–2621. Song, X., L. Zeng, W. Jin, C. Pilo, M. Frank, S. Wahl (2000) Suppression of streptococcal cell wall-induced arthritis by human chorionic gonadotropin. Arthritis Rheum 43: 2064–2072. Song, X., L. Zeng, W. Jin, J. Thompson, D. Mizel, K. Lei, R. Billinghurst, A. Poole, S. Wahl (1999) Secretory leukocyte protease inhibitor suppresses the inflammation and joint damage of bacterial cell wall-induced arthritis. J Exp Med 190: 535–542. Suenaga, S., K. Abeyama, A. Hamasaki, T. Mimura, T. Noikura (2001) Temporomandibular disorders: Relationship between joint pain and effusion and nitric oxide concentration in the joint fluid. Dentomaxillofac Radiol 30: 214– 218. Ta, L., J. Phero, S. Pillemer, H. Hale-Donze, N. McCartney-Francis, A. Kingman, M. Max, S. Gordon, S. Wahl, R. Dionne (2002) Clinical evaluation of patients with TMJ implants. J Oral Maxillofac Surg 60: 1389–1399. Takahashi, T., T. Kondoh, K. Kamei, H. Seki, M. Fukuda, H. Nagai, H. Takano, Y. Yamazaki (1996) Elevated levels of nitric oxide in synovial fluid from patients with temporomandibular disorders. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 82: 505–509. Takahashi, T., T. Kondoh, M. Ohtani, H. Homma, M. Fukuda (1999) Association between arthroscopic diagnosis of temporomandibular joint osteoarthritis and synovial fluid nitric oxide levels. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 88: 129–136. Trumpy, I., B. Roald, T. Lyberg (1996) Morphologic and immunohistochemical observation of explanted Proplast-Teflon temporomandibular joint interpositional implants. J Oral Maxillofac Surg 54: 63–68.
Cells Tissues Organs 2003;174:26–33
33
Cells Tissues Organs 2003;174:34–48 DOI: 10.1159/000070573
Pathophysiological Mechanisms in Osteoarthritis Lead to Novel Therapeutic Strategies Charles J. Malemud a–c Najmul Islam d Tariq M. Haqqi a Departments of a Medicine, b Anatomy and d Orthopaedics, Case Western Reserve University School of Medicine and c Research Institute of University Hospitals of Cleveland, Cleveland, Ohio, USA
Key Words Chondrocytes, human W Apoptosis W Osteoarthritis W Cytokine W Signaling pathways
Abstract Osteoarthritis (OA) is a debilitating, progressive disease of diarthrodial joints associated with aging. At the molecular level, OA is characterized by an imbalance between anabolic (i.e. extracellular matrix biosynthesis) and catabolic (i.e. extracellular matrix degradation) pathways in
which articular cartilage is the principal site of tissue injury responses. The pathophysiology of OA also involves the synovium in that ‘nonclassical’ inflammatory synovial processes contribute to OA progression. Chondrocytes are critical to the OA process in that the progression of OA can be judged by the vitality of chondrocytes and their ability to resist apoptosis. Growth factors exemplified by insulin-like growth factor-1, its binding proteins and transforming growth factor-ß contribute to anabolic pathways including compensatory biosynthesis of extracellular matrix proteins. Catabolic pathways are
Abbreviations used in this paper
AP-1 BH DD DED EGCG ERK FADD FLICE IAP IGF IGFBP IL JNK MACH MAPK MMP
activating protein-1 Bcl-2 homologous domain death domain death effector domain epigallocatechin-3-gallate extracellular matrix-responsive kinase Fas-activated death domain protein Fas-like interleukin-1-converting enzyme inhibitor of apoptosis protein insulin-like growth factor insulin-like growth factor-binding protein interleukin c-Jun-amino-terminal kinase MORT-1-associated CED-3 homolog mitogen-activated protein kinase matrix metalloproteinase
ABC Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2003 S. Karger AG, Basel
Accessible online at: www.karger.com/cto
MORT-1 MT-MMP NF-ÎB NO OA PA RIP TACE TGF-ß TIMP TMJ TNF-· TNFR TRADD TRAF-6
mediator of receptor-induced toxicity-1 membrane-type MMP nuclear factor-kappa B nitric oxide osteoarthritis plasminogen activator receptor-interacting protein TNF-·-converting enzyme transforming growth factor-ß tissue inhibitor of metalloproteinases temporomandibular joint tumor necrosis factor-· tumor necrosis factor receptor tumor necrosis factor receptor-activated death domain protein tumor necrosis factor receptor-associated protein-6
Charles J. Malemud, PhD Professor of Medicine and Anatomy, University Hospitals of Cleveland Foley Building, 2061 Cornell Rd. Cleveland, OH 44106-5076 (USA) Tel. +1 216 844 7846, Fax +1 216 844 2288, E-Mail
[email protected]
altered by cytokine genes such as interleukin-1 (IL-1) and tumor necrosis factor-· (TNF-·) which are upregulated in OA. In addition, IL-1 and TNF-· downregulate extracellular matrix protein biosynthesis while concomitantly upregulating matrix metalloproteinase (MMP) gene expression. When MMPs are activated, cartilage extracellular matrix degradation ensues apparently because levels of endogenous cartilage MMP inhibitors cannot regulate MMP activity. Therapeutic strategies designed to modulate the imbalance between anabolic and catabolic pathways in OA may include neutralizing cytokine activity or MMP gene expression or inhibiting signaling pathways which result in apoptosis dependent on mature caspase activity or mitogen-activated protein kinase (MAPK) activity. MAPK activity appears critical for regulating chondrocyte and synoviocyte apoptosis and MMP genes.
type II, type IX, and type XI collagens, accessory proteins such as cartilage oligomeric matrix protein and link protein [Sandell, 1995a; Thonar et al., 1999] and fibronectin [Homandberg, 1999]. Cartilage is also a rich source of hyaluronan which forms a noncovalent association with the hyalectans [Thonar et al., 1999]. Chondrocytes also synthesize a ‘free’ form of hyaluronan [Malemud et al., 1987] which may subserve hyaluronan complexed with hyalectans. A specialized pericellular matrix surrounds the chondrocyte. This matrix may consist of proteoglycans, small amounts of type VI collagen and collagen receptor proteins, such as anchorin [Sandell et al., 1995b; Thonar et al., 1999]. The pericellular matrix is often degraded in OA cartilage well before macroscopic changes are detected at the articular cartilage surface [Sachs et al., 1982].
Copyright © 2003 S. Karger AG, Basel
Cartilage Extracellular Matrix Turnover Introduction
The extracellular matrix of articular cartilage consists of proteoglycans, each belonging to several distinct proteoglycan families. These include perlecan and the hyalectans exemplified by aggrecan, and versican [Iozzo and Murdoch, 1996], the small leucine-rich family of proteoglycans, including decorin, biglycan, and fibromodulin. The extracellular matrix also consists of hybrid collagen fibers comprised of three collagen isotypes, namely
Cartilage extracellular matrix turnover is regulated by matrix metalloproteinases (MMPs) which are synthesized as latent proenzymes that require activation in order to degrade cartilage extracellular matrix proteins. Three enzymes are believed to regulate the turnover of extracellular matrix proteins, namely collagenase, responsible for degradation of native collagen fibers, stromelysins which degrade proteoglycan and type IX collagen and gelatinase(s) which degrade denatured collagen [Martel-Pelletier et al., 1999a; Smith, 1999]. Of particular importance to an understanding of the pathology of OA is the notion that while many potential MMPs have the capacity to degrade cartilage matrix proteins [Puzas et al., 2001], not all appear to be particularly important in OA. Thus, it is critical to catalog and understand the regulation of those MMP genes that appear to be intimately involved in OA as distinct from MMPs that participate in extracellular matrix protein turnover, in general. Stromelysin-1 (MMP-3) appears to be involved in OA [Sirum and Brinckerhoff, 1989; Okada et al., 1992; Hembry et al., 1995], although not even human chondrocytes derived from femoral head OA cartilage expressed MMP3 constitutively in vitro [Ganu et al., 1994]. Among the other MMPs, MMP-2 (i.e. 72-kDa gelatinase A) can degrade proteoglycans, fibronectin, and type XI collagen. MMP-9 (i.e. 92-kDa gelatinase B) can degrade proteoglycan. Both MMP-2 and MMP-9 have been found in human articular joints, but only MMP-9 has been found in increased amounts in human OA cartilage [Mohtai et al., 1993; Smith, 1999].
Pathophysiology of Osteoarthritis
Cells Tissues Organs 2003;174:34–48
Osteoarthritis (OA) is a debilitating and progressive disease of diarthrodial joints [Malemud, 1999]. At the molecular level, OA is characterized by an imbalance between the synthesis and integration of extracellular matrix proteins such as proteoglycans, collagen and link protein [Poole, 1999] on one hand, and the degradation of both newly synthesized extracellular matrix proteins [Martel-Pelletier et al., 1988] and extracellular matrix proteins integrated into the already existing matrix [Frenkel and Di Cesare, 1999], on the other. The chondrocyte plays a central role in maintaining cartilage homeostasis [Haqqi et al., 2000]. The vitality of articular cartilage can be judged on the basis of the capacity of chondrocytes to resist apoptosis (programmed cell death) and to regulate the biosynthesis and degradation of extracellular matrix proteins.
Molecular Composition of Cartilage Extracellular Matrix
35
With respect to the potential role of MMPs in temporomandibular joint (TMJ) disease, gelatin zymography detected MMP-2 at all ages in the ICR mouse mandibular condyle whereas MMP-9 levels were low at all ages [Gebstein et al., 2002]. TMJ disc cells were found to synthesize both 92- and 72-kDa gelatinase(s) in vitro [Puzas et al., 2001]. These results suggested that several MMPs relevant to peripheral joint OA might have pathophysiological significance in TMJ arthritis as well. The enzyme group that appears most relevant to cartilage degradation in OA includes a subgroup of metalloproteinases called ADAMTS because they possess disintegrin, metalloproteinase characteristics with a thrombospondin motif in their structure [Kuno et al., 2000]. This enzyme subgroup is viewed as responsible for the proteolytic degradation of the hyalectins of extracellular matrix and the molecular structure of the ADAMTS enzyme subgroup has been recently reviewed by Gao et al. [2002]. With particular reference to cartilage proteoglycan degradation, ADAMTS1, 4 and 5, exhibited activity towards aggrecan [Abbaszade et al., 1999; Tortorella et al., 1999; Kuno et al., 2000; Gao et al., 2002] whereas ADAMTS1 and 4 (‘aggrecanase-1’) also demonstrated versican-degrading activity [Sandy et al., 2001]. The activity of the MMP which ultimately proved to be ADAMTS4 was found to be elevated in OA joints [Sandy et al., 1992]. Recently, Malfait et al. [2002] showed that ADAMTS4 and 5 were present in OA human cartilage. These enzymes appear to be responsible for aggrecan degradation without MMP participation. Thus, inhibition of these enzymes may have utility in OA therapy. However, differential regulation of mRNA levels of MMP-1, MMP-3, and MMP-13 as well as ‘aggrecanase-1’ and ‘aggrecanase2’, (i.e. ADAMTS11) has also been reported in experimental rabbit OA induced by anterior cruciate ligament transection [Bluteau et al., 2001] suggesting that temporal expression of MMP genes may play a role in the sequence of cartilage alterations typical of pathophysiological changes in OA. The activity of MMPs is regulated by endogenous MMP inhibitors, called tissue inhibitor of metalloproteinases (TIMPs) [for review, see Smith, 1999]. Among the TIMPs, human joint tissue contains TIMP-1, TIMP-2 and TIMP-3 [Apte et al., 1994; Martel-Pelletier et al., 1994]. TIMP-4 is functionally redundant to TIMP-2 [Bigg et al., 1997]. TIMP-1, TIMP-2 and TIMP-3 expression is differentially downregulated by tumor necrosis factor-· (TNF-·) [Hui et al., 2001]. Recently, TIMP-4 mRNA was found in 2 of 9 nonarthritic and 12 of 14 OA femoral head cartilages [Huang et al., 2002]. A strong correlation be-
36
Cells Tissues Organs 2003;174:34–48
tween TIMP-4 mRNA and protein levels was found. Further, an increase in TIMP-4 levels in human OA cartilage [Huang et al., 2002] suggested that TIMP-4 may regulate the activity of MMPs in OA cartilage and is therefore the relevant TIMP responsible for regulating MMP activity and extracellular matrix turnover. This is important new information, as previous studies [Pelletier et al., 1990] had suggested that a relative deficit of TIMPs exists in OA cartilage when compared to the MMP levels which was thought to account for the imbalance between anabolic and catabolic pathways. Which of the MMPs relevant to OA are regulated by TIMP-4 is presently unknown. However, significant inhibition of MMP-2 (72-kDa gelatinase) by TIMP-4 has been reported [Bigg et al., 1997]. Regulating the activity of MMPs is critical in maintaining normal cartilage structure. The activation of MMP proenzymes is a key regulatory step in OA pathogenesis. A family of membrane-bound MMPs (i.e. MTMMP1 to MT-MMP-4) were shown to activate collagenase-3, but the role of MT-MMPs in the activation of proenzymes in OA cartilage has not been determined [Martel-Pelletier et al., 1999a]. Gao et al. [2002] have proposed a sequence of steps leading to ADAMTS activation which may be relevant to OA cartilage in vivo. Thus, activation of ADAMTS4 not only required truncation at the prodomain by a furin-like activity, but MMP-like removal of a segment of the C-terminal spacer domain as well [Gao et al., 2002]. Several other enzyme systems have been studied as potential activators of pro-MMPs in OA; included among these are serine- and cysteine-dependent proteases, such as plasminogen activator (PA)/plasmin and cathepsin B [Martel-Pelletier et al., 1990, 1991a, b]. Cathepsin B was found in OA cartilage [Buttle et al., 1993]. Synovial cathepsin B was shown to play a critical role in cartilage degradation in an experimental rabbit model of OA induced by partial medial meniscectomy [Mehraban et al., 1997]. There are additional examples where MMPs activate pro-MMPs [Smith, 1999] and several working models for accelerated cartilage extracellular matrix turnover in OA have been proposed [Malemud, 1999; MartelPelletier et al., 1999a; Smith, 1999; Pelletier et al., 2001].
Growth Factors and Cartilage Repair
Cartilage extracellular matrix synthesis and repair are strongly regulated by insulin-like growth factors 1 and 2 (IGF-1, IGF-2) [Messai et al., 2000] and IGF-1-binding proteins (IGFBPs) [Middleton et al., 1996; Martin et al.,
Malemud/Islam/Haqqi
1997]. Transforming growth factor-ß (TGF-ß) may also play a role in OA cartilage repair as recombinant TGFß1-stimulated chondrocyte proteoglycan core protein synthesis in vitro [Malemud et al., 1991]. TGF-ß production is also elevated in response to load [van den Berg et al., 1999] and TGF-ß augments TIMP synthesis [Wright et al., 1991; Gunther et al., 1994; Hui et al., 2001]. Growth hormone promotes longitudinal long bone growth by improving responsiveness of differentiated chondrocytes to IGF-1 [Isaksson et al., 1987]. Relatively low levels of IGF-1 are found in normal human articular cartilage [Middleton and Tyler, 1992]. However, elevated levels of IGF-1 mRNA and IGF-1 protein in human OA cartilage have been reported [Middleton and Tyler, 1992]. OA chondrocytes appear to be unresponsive to IGF-1 which may be a result of elevated levels of IGFBPs produced by OA chondrocytes [Doré et al., 1994]. The growth hormone/IGF-1 paracrine axis was found to be altered in OA. Symptomatic OA patients had elevated serum and intraerythrocyte growth hormone levels and serum IGF-1 levels when compared to normal volunteers [Denko et al., 1996; Denko and Malemud, 1999; Denko et al., 2003]. That growth hormone potentially plays a role in the clinical symptoms of OA was supported by the finding that clinically asymptomatic OA patients had serum growth hormone levels that were in the normal range [Denko and Boja, 1993]. These results suggested that aberrant growth hormone/IGF-1/IGFBP metabolism may play a role in altering IGF-1 which may limit cartilage repair pathways in OA.
Cytokines, such as interleukin-1 (IL-1) and TNF-·, play key roles in altering cartilage extracellular matrix turnover [Martel-Pelletier et al., 1999a, b]. In support of this view, human chondrocytes derived from OA cartilage failed to express MMP-3 (stromelysin-1) mRNA unless incubated with recombinant IL-1· or IL-1ß [Ganu et al., 1994]. When MMP-3 was expressed after incubation with rhIL-1· or IL-1ß, the molecular characteristics of chondrocyte MMP-3 were identical to fibroblast MMP-3 [Ganu et al., 1994]. IL-1ß suppressed TIMP synthesis while concomitantly upregulating MMP production, increasing PA synthesis and reducing plasminogen activator inhibitor-1 levels [Martel-Pelletier et al., 1991b]. Active IL-1 was found in OA synovial membrane [Pelletier et al., 1995] suggesting that activated synoviocytes and not chondrocytes were responsible for producing the IL-1
that upregulated MMP gene expression in OA cartilage. Additional cytokines produced during an ‘inflammatory’ episode in OA may serve as cofactors for regulatory molecules (i.e. IL-6, IL-8) while other molecules may inhibit the ‘inflammation’ (i.e. IL-4, IL-10, IL-13, interferon-Á) [Goldring, 1999]. Taken together, these results support two recent proposals [Pelletier et al., 2001; Attur et al., 2002a] that soluble mediators characteristic of ‘nonclassical’ inflammation pathways result in chondrocyte and synoviocyte activation that drive the progression of human OA and that therapies could be targeted towards affecting these pathways. TNF-· is synthesized as a membrane-bound precursor which is proteolytically cleaved by MMPs resulting in activated TNF-· which then forms TNF-· trimers [Gearing et al., 1994]. The enzyme responsible for the proteolytic cleavage is the TNF-·-converting enzyme (TACE). TACE gene expression was found in normal human cartilage, but TACE expression was upregulated in OA cartilage [Attur et al., 2002a]. The proinflammatory pathways initiated by IL-1 and TNF-· require receptor-mediated binding. Each cytokine binds with high affinity to specific IL-1 and TNF receptors [for review, see Martel-Pelletier et al., 1999a, b]. The binding of IL-1· to the IL-1 type I receptor served as the basis for therapeutic treatment of rheumatoid arthritis with an IL-1 receptor antagonist [Jiang et al., 2000; Bresnihan, 2001] or a type II IL-ß decoy receptor for treatment of OA [Attur et al., 2002b]. There is also increasing evidence that TNF-· is involved in the pathogenesis of OA [Goldring, 1999; Westacott et al., 2000]. This evidence serves as the impetus for elucidating the mechanisms and signaling pathways induced in chondrocytes and synoviocytes by TNF-· and its receptors. Human chondrocytes expressed both the p55 and p75 TNF-· receptors [Westacott et al., 1994]. Synoviocytes and chondrocytes derived from OA joints showed increased TNF receptors compared to their normal counterparts [Alaaeddine et al., 1997; Webb et al., 1997]. In addition, culture supernates from OA synovium or synovial fluid from OA joints apparently resulted in increased TNF receptors on normal chondrocytes when these synovial fluids were added to chondrocytes in vitro [Webb et al., 1998]. Taken together, these results suggested that TNF-· receptors were increased in OA and provide the coupling mechanism for proinflammatory TNF-·-mediated effects in OA. A strategy for neutralizing the activity of TNF-· has proven efficacious in rheumatoid arthritis patients receiving methotrexate [Lipsky et al., 2000]. Two TNF-·-blocking agents, infliximab and etanercept
Pathophysiology of Osteoarthritis
Cells Tissues Organs 2003;174:34–48
Cytokines and Other Soluble Mediators
37
(employed in the therapy of rheumatoid arthritis), were shown to suppress TNF-·-induced nitric oxide (NO) synthesis in OA cartilage explants [Vuolteenaho et al., 2002]. TGF-ß1 or IGF-1 prevented TNF-·-induced MMP synthesis [Hui et al., 2001]. IL-6 [Middleton et al., 1996], leukemia inhibitory factor and IL-17 [Martel-Pelletier et al., 1999a] are additional cytokines that target extracellular matrix synthesis, MMP gene expression and TIMP production, which suggests their potential role in OA. IL-17 is particularly interesting as it enhances the production of NO in chondrocyte cultures [Attur et al., 1997]. NO is found in increased amounts in arthritic serum, synovial fluid, and synovial membrane [Farrell et al., 1992; McInnes et al., 1996]. NO was found in OA synovium by immunolocalization [Grabowski et al., 1997]. Increased NO levels were found in medial meniscus and tibial articular cartilage compared to controls in experimental OA following partial meniscectomy [Kobayashi et al., 2001]. NO can also activate articular cartilage MMPs [Murrell et al., 1995]. IL-1ß and IL-17 augmented NO production by human OA chondrocytes [Martel-Pelletier et al., 1999a; Singh et al., 2002] and IL-1ß upregulated IL-6 synthesis by cultured human OA chondrocytes [Singh et al., 2001]. Further, IL-6 mRNA was elevated in fibrillated OA cartilage [Middleton et al., 1996]. Taken together, these results suggest that cytokine networks involving IL-1 and TNF-· and soluble proinflammatory mediators, exemplified by NO, working individually or synergistically, amplify cartilage tissue injury responses in OA.
al., 2002], compression [Bee et al., 2000], hydrostatic pressure [Malemud et al., 2001; Islam et al., 2002], mechanical injury [D’Lima et al., 2001], impact loading [Chen et al., 2001a] and inhibition of type I IGF receptor [Loesser and Shanker, 2000].
Chondrocyte Apoptosis Is Linked to Bcl-2 Gene Expression
The chondrocyte apoptosis pathway is linked to structural changes in mitochondrial membranes and alterations in the expression of a family of proteins known as Bcl-2 [Feng et al., 1998]. Mitochondrial changes and alterations in Bcl-2 are also characteristic of other cell types undergoing apoptosis [Adams and Cory, 1998; Green and Reed, 1998]. Studies of the structure and function of the Bcl-2 protein family indicate that all family members contain four conserved motifs (BH1, BH2, BH3 and BH4) which are homologous to Bcl-2. Most of the Bcl-2 family members that are affected by cytotoxic molecules contain the BH1 and BH2 motif, while those that are most similar to Bcl-2 contain all four domains [Adams and Cory, 1998]. Bcl-2 forms homodimers with itself and heterodimers with Bax, another member of the Bcl-2 family. The interaction of Bcl-2 with Bax appears essential for the antiapoptotic activity of Bcl-2 [Yin et al., 1994]. Cleavage of Bcl-2 by recombinant caspase-3 in vitro resulted in ablation of Bcl-2 interactions with itself or Bax [Lin et al., 2000]. Cleavage of Bcl-2 by caspases also promoted apoptosis [Tomicic and Kaina, 2000].
Induction of Chondrocyte Apoptosis
Chondrocyte viability and resistance to programmed cell death (apoptosis) is critical for compensatory extracellular matrix biosynthesis and repair. Chondrocyte apoptosis is not associated with articular cartilage development [Kavanagh et al., 2002]. However, the frequency of apoptotic chondrocytes increased with age in animal articular cartilages [Adams and Horton, 1998]. Induction of apoptosis by proinflammatory cytokines or other disturbances in the synovial joint milieu would be expected to compromise extracellular matrix protein compensatory and cartilage repair pathways. Chondrocyte apoptosis can be initiated in vitro by IL1ß [Singh et al., 2002, 2003], TNF-· [Fischer et al., 2000; Aizawa et al., 2001; Petterson et al., 2002], NO [Blanco et al., 1995; Kim et al., 2002b], Fas/anti-Fas [Hashimoto et al., 1997; Kühn et al., 2000; Kühn and Lotz, 2001; Sun et
38
Cells Tissues Organs 2003;174:34–48
Bcl-2 and Caspase Activation
In general two pathways, intrinsic and extrinsic, result in apoptosis [Reed, 2000]. The extrinsic pathway can be initiated via coupling of TNF-· to TNFR1 or Fas, while the intrinsic pathway is activated when cytochrome c is released from mitochondria. Release of cytochrome c is inhibited by Bcl-2 [Reed, 1997; Kluck et al., 1997; Yang et al., 1997]. Thus, direct or indirect suppression or loss of Bcl-2 function enhances cytochrome c release [O’Connor and Strasser, 1999]. The presence of cytochrome c in the cytosol results in its interaction with Apaf-1 (a flavoprotein with homology to ascorbate reductases and bacterial NADH oxidases) and procaspase-9 [Li et al., 1997; Susin et al., 1997]. When procaspase-9 activation occurs, processing and activation of other caspases occur which carry
Malemud/Islam/Haqqi
out the final ‘executioner’ signals. In this regard, mature caspase-9 (intrinsic) and mature caspase-8 (extrinsic) can both activate procaspase-3 [Reed, 2000]. To prevent apoptosis at inappropriate times a family of caspase inhibitors, inhibitors of apoptosis proteins (IAPs), inhibit mature caspases [Nicholson, 2001; Gupta, 2002]. Thus, the prototypical IAP, namely XIAP, inhibited caspase-9 and caspase-3 via specific binding regions (BIR) that interact with mature caspases [Nicholson, 2001]. The structural relationship between XIAP and its antagonist, namely Smac/DIABLO, an apoptosis-promoting mitochondrial protein appears to regulate apoptosis dynamics [Nicholson, 2001]. In another regulatory pathway, TNF-· binding to tumor necrosis factor receptor (TNFR) activates nuclear factor kappa B (NF-ÎB) which induces IAPs and mitochondrial stabilizing proteins, such as Bcl-xL and Bfl-1 [Dahl et al., 2000; Yang and Li, 2000]. Thus, NF-ÎB is a repressor of apoptosis [Gupta, 2002]. Recent evidence suggests, however, that Apaf-1 does not physically interact with Bcl-2, Bcl-xL or Bax [Conus et al., 2000]. Bcl-2 activity may also be regulated by phosphorylation. The activity of the proapoptosis Bcl-2 family members, namely Bax, Bak, Bad, Bin and Bim, often functions according to whether they are phosphorylated and where they are sequestered in the cell [Green and Reed, 1998]. Bax is a cytosolic protein which inserts into the mitochondrial membrane upon induction of apoptosis. Alterations in mitochondrial energy dynamics which can be induced by inhibitors of the electron transport system can initiate Bax association with mitochondria [Smaili et al., 2001]. Bad, another cytosolic protein, was retained in the cytosol only to be released from its cytosolic docking site when Bad was dephosphorylated [Green and Reed, 1998]. At another site for regulation of apoptosis, caspases may convert Bcl-2 to Bax-like effector molecules [Cheng et al., 1997] potentiating apoptosis.
apoptosis is associated with zones of degenerating cartilage [Hashimoto et al., 1998; Kim et al., 2000] and the frequency of apoptosis is increased in OA cartilage and dissociated chondrocytes when compared to nonarthritic cartilage and chondrocytes [Héraud et al., 2000], although countervailing data indicated that the frequency of apoptotic cells in OA cartilage could be as little or less than 1% [Aigner and Kim, 2002]. Bcl-2 is expressed in healthy adult and OA human cartilage [Erlacher et al., 1995; Malemud et al., 2001; Islam et al., 2002], but Bcl-2 expression was significantly higher in normal cartilage than in OA cartilage [Kim et al., 2000].
Induction of Chondrocyte Apoptosis by Hydrostatic Pressure
Recent evidence indicated that Bcl-2 expression also played a role in aggrecan biosynthesis [Feng et al., 1999]. Thus, underexpression of Bcl-2 in the clonal rat chondrocyte line IRC resulted in aggrecan suppression and overexpression of Bcl-2 blocked the loss of aggrecan expression induced by withdrawal of serum from the cultures [Feng et al., 1999]. Suppression of Bcl-2 may also result in additional changes in extracellular matrix proteins as well. Thus,
Mechanical loading of joint cartilage may provide the primary environmental stimulus to initiate cartilage apoptosis during aging and in early OA. To explore the possibility that mechanical loading could initiate apoptosis, we subjected human chondrocytes obtained by enzymatic dissociation of OA cartilage to continuous cyclic hydrostatic pressure levels which have been measured during the normal walking cycle (i.e. 0.8–6.3 MPa). These studies showed that human OA chondrocytes subjected to hydrostatic pressure in vitro (5 MPa; 1 Hz continuous cyclic sinusoidal waveform, 30 min to 4 h) in a servopneumatic pressure-loading device [Angele et al., 2003] were induced to undergo apoptosis [Malemud et al., 2001; Islam et al., 2002]. Continuous cyclic hydrostatic pressure (5 MPa) resulted in decreased chondrocyte viability which was independent of initial cell density [Islam et al., 2002]. That decreased chondrocyte viability induced by hydrostatic pressure resulted from apoptosis was shown by the effect of hydrostatic pressure on chondrocyte internucleosomal DNA fragments and their migration on agarose gels stained with ethidium bromide [Islam et al., 2002]. DNA fragmentation was undetectable in nonloaded chondrocytes even after 4 h or after 30 min of pressure. The pattern of DNA fragment migration after 30 min of pressure was identical to nonloaded chondrocytes. However, DNA fragments were seen after pressureloading for 2 h which was sustained at the 4-hour time point. The migration pattern of DNA fragments resembled an authentic apoptosis DNA laddering standard [Islam et al., 2002] as well as the pattern of human chondrocyte DNA fragment migration induced by incubation of human chondrocytes with TNF-· or sodium nitroprusside (unpubl. data).
Pathophysiology of Osteoarthritis
Cells Tissues Organs 2003;174:34–48
Bcl-2 and Chondrocyte Gene Expression
39
Other markers of apoptosis were also found in chondrocytes subjected to hydrostatic pressure. Thus, we first observed activation of caspase-3 after 2 h of pressure loading and cleavage of poly-ADP-ribose-polymerase (a substrate for caspase-3) after 4 h of pressure loading [Malemud et al., 2001; Islam et al., 2002]. Apoptosis is also linked to transcriptional upregulation of the tumor suppressor gene, p53 [Levine, 1997; Polyak et al., 1997; Burns and El-Deiry, 1999; Saeed Sheikh and Fornace, 2000] and the protooncogene, c-myc [Evan and Littlewood, 1998]. In this regard, hydrostatic pressure also induced Bax-· mRNA and p53 and c-myc protein but suppressed Bcl-2 mRNA and inhibited Bcl-2 protein synthesis after 4 h [Malemud et al., 2001; Islam et al., 2002]. Whether the expression of c-myc resulted in p53 gene expression and activation remains to be determined, but p53 has been shown to mediate the downregulation of Bcl-2 [Reed, 1994]. It is also noteworthy that when wildtype p53 is expressed at high levels, other pathways involved in chondrocyte apoptosis may be altered. In this regard, p53 targets the Bax-· [Wolter et al., 1997], IGF-1 receptor [Prisco et al., 1997] and IGFBP3 genes [Buckbinder et al., 1995]. Thus, upregulation of Bax-·, p53 and c-myc chondrocyte gene expression by hydrostatic pressure may not only induce chondrocyte apoptosis, but further exacerbate alterations in chondrocyte homeostasis such as those that occur in OA cartilage [Islam et al., 2002]. This could be accomplished by targeting proapoptotic genes affecting IGF-1 and IGF-1 receptor genes that likely contribute to cartilage proliferation pathways [Doré et al., 1994; Morales, 1995] or MMP gene expression and MMP activation [Sun et al., 2000a] which contribute to cartilage degradation.
Death Domain Proteins and Apoptosis
The apoptosis cascade ultimately involves the synthesis and activation of caspases, cysteine proteases with aspartate specificity [Salvesen and Dixit, 1997; Thornberry and Lazebnik, 1998; Elkon, 1999; Reed, 2000]. Latent caspases are activated by several mechanisms including the binding of adaptor proteins in the cytosol to latent caspases resulting in cleavage and procaspase activation [Verhagen and Vaux, 1999]. Prototypical among the mammalian adaptor proteins required from procaspase activation are MORT-1/FADD [mediator of receptor-induced toxicity-1/Fas-associated death domain protein) [Chinnaiyan et al., 1995; Chao et al., 2002]. In the extrinsic pathway, apoptosis can be
40
Cells Tissues Organs 2003;174:34–48
induced when Fas ligand or anti-Fas antibody binds to the Fas receptor on cell membranes. Induction of Fasinduced apoptosis requires the presence of a conserved cytoplasmic protein structure known as the death domain (DD) [Cleveland and Ihle, 1995] present in the C-terminus of Fas and FADD. The Fas and FADD DD interact linking Fas/FADD to other downstream signaling activation steps which can also be initiated by TNF-·/TNFR interactions. In keeping with this view, it was shown that cells from FADD knockout mice resisted TNF-induced apoptosis [for review, see Ashkenazi and Dixit, 1998], indicating the key role FADD plays in the extrinsic apoptosis pathway. The DD motif in FADD is not sufficient, however, to trigger apoptosis. FADD also possesses a death effector domain (DED) which is present in the prodomain of latent caspase-8, also called Fas-associated ICE-like protease (FLICE) or MACH-1 (MORT-associated CED-3 homolog) [Boldin et al., 1996; Muzio et al., 1996]. CED-3 is an apoptosis protein in Caenorhabditis elegans. MACH binds to MORT-1/FADD. Recognition by FADD of the procaspase-8 DED recruits procaspase-8 to the cell surface [Chinnaiyan et al., 1995] resulting in autoprocessing, release of activated caspase-8 and initiation of downstream apoptosis events [Juo et al., 1999]. Taken together with data showing that FADD overexpression leads to cell death [reviewed in Duckett, 2002], these results indicated that FADD itself is a death-inducing protein. In this regard, intracellular FADD levels may be important in regulating the efficiency of apoptosis [Chao et al., 2002].
Effect of Hydrostatic Pressure on Chondrocyte FADD
To test the hypothesis that human OA chondrocyte apoptosis induced by hydrostatic pressure was accompanied by changes in FADD, chondrocytes were subjected to hydrostatic pressure (5 MPa for 1–4 h; 1-Hz continuous cyclic sinusoidal waveform). Chondrocyte protein lysate was immunoprecipitated, subjected to SDS-PAGE and Western blotted with murine anti-human FADD, clone A66-2, IgG1 (Pharmingen International). The results shown in figure 1 indicated that that chondrocyte FADD was produced constitutively (fig. 1, lane 1). Pressure loading (5 MPa for 1–4 h) caused an apparent increase in FADD after 1 h, which was sustained over the 4-hour time course. No significant change in ß-actin protein measured by Western blotting was seen over the identical 4-hour incubation period [Islam et al., 2002].
Malemud/Islam/Haqqi
1
2
3
4
48 kD
26 kD
Fig. 1. Hydrostatic pressure increases FADD synthesis in human
OA chondrocytes. Human chondrocytes from OA femoral heads were grown in high-density monolayer culture, transferred to heatsealable bags and subjected to hydrostatic pressure (5 MPa for 1–4 h) [Islam et al., 2002; Angele et al., 2003]. Chondrocyte protein lysate was immunoprecipitated with purified mouse anti-human FADD (Pharmingen International). The detection of antigen/antibody complexes employed the ECL + Plus Western blotting detection system (Amersham, Arlington Heights, Ill., USA). After SDS-PAGE, the proteins were transferred to a blotting membrane and probed with mouse anti-human FADD. FADD has a calculated molecular weight of 24 kDa and migrates in SDS-PAGE as a F27-kDa protein (Pharmingen International technical data sheet). The arrows show the position of the expected 26-kDa FADD-reactive product and a 48-kDa anti-FADD-reactive product whose identity is unknown. Lane 1 = Nonloaded chondrocytes, 4 h; lane 2 = pressure-loaded (5 MPa, 1 Hz frequency), 1 h; lane 3 = pressure-loaded, 2 h; lane 4 = pressure-loaded, 4 h.
However, no correlation was found between FADD and apoptosis in nonneoplastic adjacent liver [Sun et al., 2000b]. Western blot evidence of constitutive FADD synthesis was also found in B cell lymphomas resistant to Fasmediated apoptosis which was not a result of defective FADD [Xerri et al., 1999]. These results indicated that upregulation or overexpression of FADD contributed to human OA chondrocyte apoptosis induced by hydrostatic pressure. These results also suggested that a mechanotransduction signaling pathway [French, 1992; Ko and McCulloch, 2001] activated by hydrostatic pressure could engage the TNFR and/or Fas receptor which share a common signaling pathway involving the adaptor protein, TNF-receptor-activated death domain protein (TRADD) and FADD [Chinnaiyan et al., 1995; Muzio et al., 1996]. In keeping with this view, TRADD binds to the TNF-p55 receptor and receptorinteracting protein (RIP) to initiate TNF/TNFR-mediated apoptosis signaling downstream events. The implication that FADD plays a critical role in apoptosis has resulted in its exploitation for developing experimental therapeutic strategies for treating arthritis. Thus, synovial hyperplasia caused by inhibition of apoptosis plays a role in arthritis [Urayama et al., 2001]. In this regard, local injection of FADD adenovirus induced apoptosis in vivo in proliferating human rheumatoid synovium engrafted into the severe combined immunodeficient mouse [Kobayashi et al., 2000]. In this study, chondrocytes were unaffected by the FADD adenovirus construct. This result suggested that novel strategies employed to upregulate FADD might be useful to treat synovial hyperplasia resulting from synoviocyte activation which may impact on the progression of articular cartilage changes in inflammatory arthritis or OA.
FADD has a calculated molecular weight of 24 kDa. The 26-kDa band reactive with anti-human FADD is about the predicted size of FADD which is F27–28 kDa in SDS-PAGE [Boldin et al., 1996]. The 48-kDa band shown in figure 1 also reacted with anti-human FADD, but its identity is presently unknown. It may represent a complex of FADD and procaspase-8. FADD/procaspase-8 complex formation may be recognized by antiFADD since other adaptor proteins containing DD sequences form intramolecular complexes [Cleveland and Ihle, 1995]. Constitutive expression of FADD by nonloaded human OA chondrocytes (fig. 1) was not correlated with apoptosis as nonloaded chondrocytes did not exhibit DNA fragmentation [Islam et al., 2002]. Constitutive FADD expression has been reported in primary hepatocarcinoma even when low levels of apoptosis were found, but strong evidence for hepatic apoptosis was characterized by increased levels of FADD [Sun et al., 2000b].
Many of the downstream events resulting in death signals involve signal transduction pathways. These signaling pathways link IL-1ß, TNF-· and NO to MMP transcription, mitogen-activated protein kinases (MAPKs), NF-ÎB and FADD [Chen et al., 2001b; Notoya et al., 2000; Vincenti and Brinckerhoff, 2001; Duckett, 2002]. NF-ÎB is a transcription factor that has been identified as playing an important role in apoptosis. In quiescent cells, NF-ÎB is sequestered in the cytoplasm in an inactive form through its interaction with several inhibitory components, namely IÎB-·, IÎB-ß, IÎB-Â, p105 and p100 [for
Pathophysiology of Osteoarthritis
Cells Tissues Organs 2003;174:34–48
Apoptosis and Signal Transduction Pathways: Potential Intervention by Novel Therapies
41
review, see Chen et al., 2001b]. Activation of NF-ÎB results in degradation of IÎB and p105 and p100 precursors resulting in NF-ÎB translocation to the nucleus where it induces transcription. Rapid activation of NF-ÎB by TNF-·, c-myc and Fas ligand [Hsu et al., 1998; Matsui et al., 1998] suggested that NF-ÎB was a critical proapoptotic factor. More recent evidence, however, suggested that NF-ÎB may behave in an antiapoptotic role as well since in murine gene ‘knockouts’, where genes encoding members of the NF-ÎB family or kinase activity regulated by NF-ÎB were deleted, resulted in embryonic lethals characterized by liver cell apoptosis and other cellular abnormalities [Tanaka et al., 1999]. The importance of NF-ÎB activity in apoptosis has led to experimental studies designed to regulate NF-ÎB activity in an attempt to exploit this pathway in developing novel strategies for treating autoimmunity and arthritis. Recent studies employing naturally occurring molecules, such as hyaluronic acid, chondroitin sulfate and glucosamine, have indicated that a site of biological activity of these molecules is NF-ÎB [Gouze et al., 2002; Yang et al., 2002]. Thus, hyaluronic acid and chondroitin sulfate augmented NF-ÎB activity in immature dendritic cells which promoted their differentiation [Yang et al., 2002]. Dendritic cells have been identified as a potent antigen-presenting cell in that they effectively bind antigen and migrate to lymphoid organs where they induce primary T cell responses [Banchereau and Steinman, 1998]. In another study, glucosamine added to rat chondrocytes in culture primed with IL-1ß decreased the activation of NFÎB and the expression of the IL-ß receptor [Gouze et al., 2002]. This was significant as IL-1 binds to its receptor, IL-1R1 which activates the TNF receptor molecule-associated protein-6 (TRAF-6) leading to activation of the NF-ÎB-inducing kinase [Ninomiya-Tsugi et al., 1999]. In turn, NF-ÎB activates nuclear target genes, important in apoptosis and MMP transcription. Thus, IL-1-activated TRAF-6 can also activate MAPKs. MAPKs promote phosphorylation of c-Jun-NH2-terminal kinase (JNK) and members of the Fos gene family, both of which are associated with activating protein-1 (AP-1) activation [Barchowsky et al., 2000]. NF-ÎB activity is critical for MMP transcription [Mengshol et al., 2000; Vincenti and Brinckerhoff, 2001] and indeed MEKK3 has been implicated in TNF-induced NF-ÎB activation [Yang et al., 2000]. MEK-kinases appear to play a critical role as p38 and JNK activators [Deacon and Blank, 1999]. Recent studies have also identified JNK-2 as a potential site for therapeutic intervention. Han et al. [2001a] showed that following the induction of type II collagen-induced
42
Cells Tissues Organs 2003;174:34–48
arthritis in JNK knockout mice, the cartilage from the JNK-2 knockouts was more safraninophilic suggesting that less proteoglycan was degraded in these mice compared to their wild-type counterparts. The cartilage in the JNK-2 knockouts was also less focally damaged than the cartilage in the wild-type mice expressing JNK. However, the JNK2 ‘knockout’ mice and wild-type mice exhibited similar levels of AP-1 binding and MMP-13 (collagenase-3) activity suggesting that both JNK-1 and JNK-2 expression are required for maximal AP-1 binding and MMP-13 expression. As JNK activity is critical in the MAPK pathway for IL-1-induced collagenase expression in synoviocytes of arthritic joints [Han et al., 2001b], inhibition of JNK with the novel experimental JNK inhibitor SP600125 {anthra[1,9-cd]pyrazol-6(2H)-one} completely blocked IL-1induced accumulaton of phosphorylated Jun and induction of c-Jun transcription in synoviocytes [Han et al., 2001b]. This was accompanied by suppression of AP-1 binding and collagenase mRNA levels. Taken together, these results suggested that inhibition of JNK activity may be expected to regulate MMP expression in arthritis. The interest in mechanisms underlying the potential use of a biologically active polyphenolic fraction of green tea (Camellia sinensis) epigallocatechin-3-gallate (EGCG) as a therapeutic strategy for treating arthritis was stimulated by the novel finding by Haqqi et al. [1999] that orally administered EGCG ameliorated inflammatory arthritis in murine collagen-induced arthritis. The Haqqi group also showed that EGCG induced apoptosis in an immortalized line of chondrosarcoma chondrocytes (HTB-94] [Islam et al., 2000], suggesting that ablation of inflammatory synovitis by EGCG in arthritis could occur by induction of apoptosis. Indeed, HTB-94 transcribes a mutant p53 isoform [Islam et al., 2000] which may be responsible for the resistance to apoptosis in this cell line. The results of Islam et al. [2000] also indicated that EGCG-induced apoptosis in HTB-94 was mediated via direct activation of caspase-3 by EGCG. To further explore the role of MAPKs in human chondrocyte apoptosis and MMP gene expression, MAPK gene expression and phosphorylation in human OA chondrocytes incubated in vitro with IL-1ß with and without EGCG was examined. IL-1ß-activated human OA chondrocytes treated with EGCG showed complete inhibition of the two phospho-JNK isoforms, accumulation of phospho-c-Jun and DNA-binding activity of AP-1 [Singh et al., 2003]. EGCG did not inhibit synthesis of JNK nor did EGCG affect p38-MAPK or p44/042 extracellular signalregulated kinase (ERK) [Singh et al., 2003]. EGCG also inhibited IL-1ß-induced NO production via inhibi-
Malemud/Islam/Haqqi
tion of iNOS. EGCG inhibition of iNOS was regulated by controlling NF-ÎB activation and translocation to the nucleus by suppressing IÎB-· degradation in the cytoplasm [Singh et al., 2002]. In addition, EGCG inhibited IL-6 synthesis [Singh et al., 2001]. The suppression of NO by EGCG supports the findings of previous study which showed that EGCG inhibited NO-induced apoptosis cell cycle arrest and apoptosis in the U937 human promonocyte cell line [Kelly et al., 2001]. Taken together, these results indicated that EGCG has the capacity to alter signaling pathways intimately involved in apoptosis and MMP gene transcription. In support of this, EGCG was recently shown to inhibit bovine and human cartilage proteoglycan and type II collagen degradation in cartilage explants without toxicity [Adcocks et al., 2002]. This suggests that EGCG inhibition of JNK could be coupled to preservation of cartilage extracellular matrix in OA. Another site for potential inhibition of signaling pathways regulating apoptosis and MMPs are the MAPKs and ERKs [Geng et al., 1996; Islam et al., 2001; Shakibaei et al., 2001; Liacini et al., 2002]. NO induces apoptosis in human chondrocytes [Blanco et al., 1995]. Human OA chondrocytes express ERKs and ERK1/2 and p38 kinase has been implicated in NO induction of cyclooxygenase-2 and PGE2 in human chondrocyte cultures [Notoya et al., 2000]. Blocking MAPK activity with the ERK1/2 inhibitor, PD98059 or the p38 inhibitor, SB202190 inhibited NO-induced apoptosis and PGE2 production [Notoya et al., 2000]. PGE2 alone did not induce apoptosis. In another study, the stimulation of apoptosis by NO in rabbit articular chondrocyte cultures coupled with ERK inhibition was accompanied by increased p53 accumulation and caspase-3 activity [Kim et al., 2002], while inhibition of p38 kinase activity resulted in reduced p53 and caspase-3, suggesting that ERK and p38 counteract each other in regulating chondrocyte apoptosis. However, ERK and p38 are both required for differentiation of human mesenchymal stem cells into chondrocytes in vitro while JNK is not [Friess et al., 2002] and the expression of ERK and p38 are likely critical for maintenance of the chondrogenic phenotype [Kim et al., 2002]. The potential for p38 MAPK inhibition of apoptosis in OA cartilage was recently suggested by Sun et al. [2002] in that repression of p38 by the p38 inhibitor, SB203580, inhibited apoptosis by suppressing the apoptosis-associated factor-activating transcription factor-2 and caspase-2 activity. Anti-Fas activation of apoptosis was accompanied by elevated p38 activity [Sun et al., 2002].
Pathophysiology of Osteoarthritis
Caspase inhibition as a potential target for interfering with apoptosis has been previously proposed by Thornberry and Lazebnik [1998], and differential inhibition of caspases has been implicated in CD95 (Fas/Apo-1)induced apoptosis in cultured human chondrocytes [Kühn and Lotz, 1991]. It is therefore instructive that a novel experimental disease-modifying antirheumatic drug, 2-acetylthiomethyl-4-(4-methylphenyl)-4-oxobutanoic acid (KE-298), increased caspase-3 activation and augmented activation-induced T cell death [Urayama et al., 2001]. Nicholson [2001] has proposed another potential strategy to induce apoptosis whereby a protein mimetic designed to interact with IAPs blocks IAP-mediated caspase inhibition.
Conclusions
An increasing understanding of the pathways at the molecular level that initiate and result in progression of human OA has provided a platform for developing novel therapeutic strategies that could regulate cartilage homeostasis. The proinflammatory cytokines, IL-1 and TNF-·, have been identified as playing critical roles in initiating apoptosis and upregulating MMP gene expression in peripheral joint OA. These mediators have also been identified as playing key roles in TMJ disease [Puzas et al., 2001; Kacena et al., 2001]. An understanding of the signaling pathways that regulate the expression of tumor suppressor genes, such as p53, must also be examined in that p53 not only alters IGF-1 [Prisco et al., 1997] and IGFBP gene expression [Buckbinder et al., 1995], but MMP-13 expression as well [Sun et al., 2000a]. The ability to modify the activity of cytokines, the signaling pathways they activate, the synthesis and activation of MMPs and caspases all appear to be important targets for novel drug discovery programs [Nuttall, 2001] designed to provide new agents for the medical management of OA and TMJ disease.
Acknowledgement The experimental studies were supported, in part, by grants from NIH and a development award from the Department of Orthopaedics, Case Western Reserve University School of Medicine. We wish to acknowledge the participation of Karl J. Jepsen, PhD and Jean F. Welter, PhD who designed the servopneumatic pressure-loading device, Matthew F. Kraay, MD and Victor M. Goldberg, MD for their continuous support of the project entitled ‘In vitro Behavior of Human Cartilage’, and the Cooperative Human Tissue Network of University Hospitals of Cleveland, Cleveland, Ohio.
Cells Tissues Organs 2003;174:34–48
43
References Abbaszade, I., R.Q. Liu, F. Yang, S.A. Rosenfeld, O.H. Ross, J.R. Link, D.M. Ellis, M.D. Tortorella, M.A. Pratta, J.M. Hollis, R. Wynn, J.L. Duke, H. George, M.C. Hillman, Jr., K. Murphy, B.H. Wiswall, R.A. Copeland, C.P. Decicco, R. Bruckner, H. Nagase, Y. Itoh, R.C. Newton, R.L. Magolda, J.M. Trzaskos, G.F. Hollis, E.C. Arner, T.C. Burn (1999) Cloning and characterization of ADAMTS11, an aggrecanase from the ADAMTS family. J Biol Chem 274: 23443–23450. Adams, C.S., W.E. Horton, Jr. (1998) Chondrocyte apoptosis increases with age in the articular cartilage of adult animals. Anat Rec 250: 418– 425. Adams, J.M., S. Cory (1998) The Bcl-2 protein family: Arbiters of cell survival. Science 281: 1322–1325. Adcocks, C., P. Collin, D.J. Buttle (2002) Catechins from green tea (Camellia sinensis) inhibit bovine and human cartilage proteoglycan and type II collagen degradation in vitro. J Nutr 132: 341–346. Aigner, T., H.A. Kim (2002) Apoptosis and cellular viability. Issues in osteoarthritic cartilage degradation. Arthritis Rheum 46: 1986–1996. Aizawa, T., T. Kon, T.A. Einhorn, L.C. Gerstenfeld (2001) Induction of apoptosis in chondrocytes by tumor necrosis factor-alpha. J Orthop Res 19: 785–796. Alaaeddine, N., J.A. Di Battista, J.-P. Pelletier, J.M. Cloutier, K. Kiansa, M. Dupuis, J. MartelPelletier (1997) Osteoarthritic synovial fibroblasts possess an increased level of tumor necrosis factor-receptor 55 (TNF-R55) that mediates biological activation of TNF-alpha. J Rheumatol 24: 1985–1994. Angele, P., J.U. Yoo, C. Smith, J. Mansour, K.J. Jepsen, M. Nerlich B. Johnstone (2003) Cyclic hydrostatic pressure enhances chondrogenic phenotype of human mesenchymal cell differentiated in vitro. J Orthop Res, in press. Apte, S.S., M. Mattei, B.R. Olsen (1994) Cloning of the cDNA encoding human tissue inhibitor of metalloproteinases-3 (TIMP-3) and mapping of the TIMP-3 gene to chromosome 22. Genomics 19: 86–90. Ashkenazi, A., V. Dixit (1998) Death receptors: Signaling and modulation. Science 281: 1305– 1308. Attur, M.G., M. Dave, M. Akamatsu, M. Ktaoh, A.R. Amin (2002a) Osteoarthritis or osteoarthrosis: The definition of inflammation becomes a semantic issue in the era of molecular medicine. Osteoarthritis Cartilage 10: 1–4. Attur, M.G., M.N. Dave, M.Y. Leung, C. Cipolletta, M. Meseck, S.L.C. Woo, A.R. Amin (2002b) Functional genomic analysis of type II IL-1ß decoy receptor: Potential for gene therapy in human arthritis and inflammation. J Immunol 168: 2001–2010. Attur, M.G., R.N. Patel, S.B. Abramson, A.R. Amin (1997) Interleukin-17 up-regulation of nitric oxide production in human osteoarthritis cartilage. Arthritis Rheum 140: 1050–1053.
44
Banchereau, J., R.M. Steinman (1998) Dendritic cells and the control of immunity. Nature 392: 245–252. Barchowsky, A., D. Frieta, M.P. Vincenti (2000) Integration of the NF-kappaB and mitogenactivated protein kinase/AP-1 pathways at the collagenase-1 promoter: Divergence of IL-1 and TNF-dependent signal transduction of rabbit primary synovial fibroblasts. Cytokine 12: 1469–1479. Bee, Z., G.V. Crossingham, K.M. Clements, M.A. Adams, M. Sharif (2000) Chondrocyte death induced by cyclic compressive loading of articular cartilage: Evidence for involvement of apoptosis (abstract). Osteoarthritis Cartilage 8(suppl B): S53. Bigg, H.F., Y.E. Shi, Y.E. Lui, B. Steffenson, C.M. Overall (1997) Specific, high affinity binding of tissue inhibitor of metalloproteinases-4 (TIMP-4) to the COOH-terminal hemopexinlike domain of human gelatinase A. TIMP-4 binds progelatinase A and the COOH-terminal domain in a similar manner to TIMP-2. J Biol Chem 272: 15496–15500. Blanco, F.J., R.L. Ochs, H. Schwarz, M. Lotz (1995) Chondrocyte apoptosis induced by nitric oxide. Am J Pathol 146: 75–85. Bluteau, G., T. Conrozier, P. Mathieu, E. Vignon, D. Herbage F. Mallein-Gerin (2001) Matrix metalloproteinase-1, -3, -13 and aggrecanase-1 and -2 are differentially expressed in experimental osteoarthritis. Biochim Biophys Acta 1526: 147–158. Boldin, M.P., T.M. Goncharov, Y.V. Goltsev, D. Wallach (1996) Involvement of MACH, a novel MORT1/FADD-interacting protease, in Fas/ Apo-1 and TNF receptor-induced cell death. Cell 85: 803–815. Bresnihan, B. (2001) The safety and efficacy of interleukin-1 receptor antagonist in the treatment of rheumatoid arthritis. Semin Arthritis Rheum 30(suppl 2): 17–20. Buckbinder, L., R. Talbott, S. Valesco-Miguel, I. Takenaka, B. Faha, B.R. Seizinger, N. Kley (1995) Induction of the growth inhibitor IGFbinding protein 3 by p53. Nature 377: 646– 649. Burns, T.F., W.S. El-Deiry (1999) The p53 pathway and apoptosis. J Cell Physiol 181: 231–239. Buttle, D., C.J. Handley, M.Z. Ilic, J. Saklatvala, J. Murata, A.J. Barrett (1993) Inhibition of proteoglycan release by a specific inhibitor of cathepsin B and an inhibitor of metalloproteinases. Evidence for two converging pathways of chondrocyte-mediated degradation. Arthritis Rheum 36: 1709–1717. Chao, W., Y. Shen, L. Li, A. Rosenzweig (2002) Importance of FADD signaling in serum deprivation and hypoxia-induced cardiomyocyte apoptosis. J Biol Chem 277: 31639–31645. Chen, C.-T., N. Burton-Wurster, C. Borden, K. Hueffer, S.E. Bloom, G. Lust (2001a) Chondrocyte necrosis and apoptosis in impact damaged articular cartilage. J Orthop Res 19: 703–711.
Cells Tissues Organs 2003;174:34–48
Chen, F., V. Castranova, X. Shi (2001b) New insights into the role of nuclear factor-ÎB in cell growth regulation. Am J Pathol 159: 387–397. Cheng, E.H.-Y., D.G. Kirsch, R.J. Clem, R. Ravi, M.B. Kastan, A. Bedl, K. Ueno, J.M. Hardwick (1997) Conversion of Bcl-2 to Bax-like death effector by caspases. Science 278: 1966–1968. Chinnaiyan, A.M., K. O’Rourke, M. Terwari, V.M. Dixit (1995) FADD, a novel death-domaincontaining protein, interacts with the death domain of Fas and initiates apoptosis. Cell 81: 505–512. Cleveland, J.L., J.N. Ihle (1995) Contenders in FasL/TNF death signaling. Cell 81: 479–482. Conus, S., T. Rosse, C. Borner (2000) Failure of Bcl-2 family members to interact with Apaf-1 in normal and apoptotic cells. Cell Death Differ 7: 947–954. Dahl, A.M., C. Klein, P.G. Andres, C.A. London, M.P. Lodge, R.C. Mulligan, A.K. Abbas (2000) Expression of bcl-X(L) restores cell survival, but not proliferation of effector differentiation, in CD28-deficient T lymphocytes. J Exp Med 191: 2031–2038. Deacon, K., J.L. Blank (1999) MEK-kinases directly activates MMK6 and MKK7 specific activators of the p38 and c-Jun-NH2-terminal kinases. J Biol Chem 274: 16604–16610. Denko, C.W., B. Boja (1993) Growth factors in asymptomatic osteoarthritis – Insulin, insulinlike growth factor-1, growth hormone. Inflammopharmacology 2: 71–76. Denko, C.W., B. Boja, C.J. Malemud (2003) Intraerythrocyte deposition of growth hormone in rheumatic diseases. Rheumatol Int 23: 11–14. Denko, C.W., B. Boja, R.W. Moskowitz (1996) Growth factors, insulin-like growth factor-1 and growth hormone, in synovial fluid and serum of patients with rheumatic diseases. Osteoarthritis Cartilage 4: 245–249. Denko, C.W., C.J. Malemud (1999) Metabolic disturbances and synovial joint responses in osteoarthritis. Front Biosci 4: d686–d693. D’Lima, D.D., S. Hashimoto, P.C. Chen, C.W. Colwell, Jr., M.K. Lotz (2001) Human chondrocyte apoptosis in response to mechanical injury. Osteoarthritis Cartilage 9: 712–719. Doré, S., J.-P. Pelletier, J.A. Di Battista, G. Tardif, P. Brazeau, J. Martel-Pelletier (1994) Human osteoarthritic chondrocytes possess an increased number of insulin-like growth factor-1 binding sites but are unresponsive to its stimulation. Arthritis Rheum 37: 253–263. Duckett, C.S. (2002) Apoptosis and NF-ÎB: The FADD connection. J Clin Invest 109: 579– 580. Elkon, K.B. (1999) Caspases: Multifunctional proteases. J Exp Med 190: 1725–1727. Erlacher, L., R. Maier, R. Ullrich, H. Kiener, M. Aringer, M. Menschik, W. Graninger (1995) Differential expression of the protooncogene bcl-2 in normal and osteoarthritic human articular cartilage. J Rheumatol 22: 926–931.
Malemud/Islam/Haqqi
Evan, G., T. Littlewood (1998) A matter of life and death. Science 281: 1317–1321. Farrell, A.J., D.R. Blake, R.M. Palmer, S. Moncada (1992) Increased concentrations of nitrite in synovial fluid and serum samples suggest increased nitric oxide synthesis in rheumatic diseases. Ann Rheum Dis 51: 1219–1222. Feng, L., R. Balakir, P. Precht, W.E. Horton, Jr. (1999) Bcl-2 regulates chondrocyte morphology and aggrecan gene expression independent of caspase activation and full apoptosis. J Cell Biochem 74: 576–586. Feng, L., P. Precht, R.R. Balakir, W.E. Horton, Jr (1998) Evidence for a direct role of Bcl-2 in the regulation of articular chondrocyte apoptosis under conditions of serum withdrawal and retinoic acid treatment. J Cell Biochem 71: 303– 309. Fisher, B.A., S. Mundle, A.A. Cole (2000) Tumor necrosis factor-alpha induced DNA cleavage in human articular chondrocytes may involve multiple endonucleolytic activities during apoptosis. Microsc Res Tech 50: 236–242. Freiss, D., J. Yoo, B. Johnstone (2002) MAP kinase signaling during in vitro chondrogenesis (abstract 0313). Trans Orthop Res Soc 27. French, A.S. (1992) Mechanotransduction. Annu Rev Physiol 54: 135–152. Frenkel, S.R., P.E. Di Cesare (1999) Degradation and repair of articular cartilage. Front Biosci 4: d671–d685. Ganu, V.S., S.-I. Hu, R. Melton, C. Winter, V.M. Goldberg, T.M. Haqqi, C.J. Malemud (1994) Biochemical and molecular characterization of stromelysin synthesized by human osteoarthritic chondrocytes in culture stimulated with recombinant interleukin-1. Clin Exp Rheumatol 12: 489–496. Gao, G., J. Westling, V.P. Thompson, T.D. Howell, P.E. Gottschall, J.D. Sandy (2002) Activation of the proteolytic activity of ADAMTS4 (aggrecanase-1) by C-terminal trunction. J Biol Chem 277: 11034–11041. Gearing, A.J., P. Beckett, M. Christidoulous, M. Churchill, J. Clements, A.H. Davidson, A.H. Drummond, W.A. Galloway, R. Gilbet, J.L. Gordon, T.M. Leber, M. Manigan, K. Miller, P. Nayee, K. Owen, S. Patel, W. Thomas, G. Wells, L.M. Wood, K. Woolley (1994) Processing of tumor necrosis factor-alpha precursor by metalloproteinases. Nature 370: 555–557. Gebstein, A., S. Shapiro, N. Lahat, E. Livne (2002) Matrix metalloproteinase activity and mRNA expression in cartilage of temporomandibular and knee joints in growth development and aging (abstract). Bone 30(suppl): 27S. Geng, Y., J. Valbracht, M. Lotz (1996) Selective activation of the mitogen-activated protein kinase subgroups c-Jun NH2 terminal kinase and p38 by IL-1 and TNF in human articular chondrocytes. J Clin Invest 98: 2292–2299. Goldring, M.B. (1999) The role of cytokines as inflammatory mediators in osteoarthritis: Lessons from animal models. Connect Tissue Res 40: 1–11.
Pathophysiology of Osteoarthritis
Gouze, J.N., A. Bianchi, P. Bécuwe, M, Dauça, P. Netter, J. Magdalou, B. Terlain, K. Bordji (2002) Glucosamine modulates IL-1 induced activation of rat chondrocytes at a receptor level, and by inhibiting the NF-ÎB pathway. FEBS Lett 510: 166–170. Grabowski, P.S., P.K. Wright, R.J. Van’t Hof, M.F. Helfrich, H. Ohshima, S.H. Ralston (1997) Immunolocalization of inducible nitric oxide synthase in synovium and cartilage in rheumatoid and osteoarthritis. Br J Rheumatol 36: 651– 655. Green, D.R., J.C. Reed (1998) Mitochondria and apoptosis. Science 281: 1309–1312. Gunther, M., H.D. Haubeck, E. van der Leur, J. Blaser, S. Bender, I. Gutgemann, D.C. Fischer, H. Tschesche, H. Greiling, P.C. Heinrich, L. Graeve (1994) TGF-ß1 regulates tissue inhibitor of metalloproteases-1 expression in differentiated human articular chondrocytes. Arthritis Rheum 37: 395–405. Gupta, S. (2002) A decision between life and death during TNF-· signaling. J Clin Immunol 22: 185–194. Han, Z., D.L. Boyle, L. Chang, B. Bennett, M. Karin, L. Yang, A.M. Manning, G.S. Firestein (2001b) c-Jun-N-terminal kinase is required for metalloproteinase expression and joint destruction in inflammatory arthritis. J Clin Invest 108: 73–81. Han, Z., L. Chang, Y. Yamanishi, M. Karin, G.S. Firestein (2001a) Joint damage and inflammation in c-Jun-N-terminal kinase 2 knockout mice with passive murine collagen-induced arthritis. Arthritis Rheum 46: 818–823. Haqqi, T.M., D.D. Anthony, S. Gupta, N. Ahmed, M.-S. Lee, G.K. Kumar, H. Mukhtar (1999) Prevention of collagen-induced arthritis in mice by a polyphenolic fraction from green tea. Proc Natl Acad Sci USA 96: 4524–4529. Haqqi, T.M., D.D. Anthony, C.J. Malemud (2000) Chondrocytes; in Tsokos, G.C. (ed): Current Molecular Medicine: Principles of Molecular Rheumatology. Totowa, Humana Press, pp 267–277. Hashimoto, S., R.L. Ochs, S. Komiya, M. Lotz (1998) Linkage of chondrocyte apoptosis and cartilage degradation in human osteoarthritis. Arthritis Rheum 41: 1632–1638. Hashimoto, S., K. Takahashi, D. Amiel, R.D. Coutts, M. Lotz (1997) Fas/Fas ligand expression and induction of apoptosis in chondrocytes. Arthritis Rheum 40: 1749–1755. Hembry, R.M., M.R. Bagga, J.J. Reynolds, D.L. Hamblen (1995) Immunolocalisation studies on six matrix metalloproteinases and their inhibitors, TIMP-1 and TIMP-2, in synovia from patients with osteo- and rheumatoid arthritis. Ann Rheum Dis 54: 25–32. Héraud, F., A. Héraud, M.-F. Harmand (2000) Apoptosis in normal and osteoarthritic human articular cartilage. Ann Rheum Dis 59: 959– 965. Homandberg, G.A. (1999) Potential regulation of cartilage metabolism in osteoarthritis by fibronectin fragments. Front Biosci 4: d713– d730.
Hsu, S.C., M.A. Gavrilin, H.H. Kee, C.C. Wu, S.H. Han, M.Z. Lai (1998) NF-ÎB-dependent Fas ligand expression. Eur J Immunol 29: 2948– 2956. Huang, W., W.Q. Li, F. Dehnade, M. Zafarullah (2002) Tissue inhibitor of metalloproteinase-4 (TIMP-4) gene expression is increased in human osteoarthritic femoral head cartilage. J Cell Biochem 85: 295–303. Hui, W., A.D. Rowan, T. Cawston (2001) Modulation of the expression of matrix metalloproteinase and tissue inhibitors of metalloproteinases by TGF-ß1 and IGF-1 in primary human articular and bovine nasal chondrocytes stimulated with TNF-·. Cytokine 16: 31–35. Iozzo, R.V., A.D. Murdoch (1996) Proteoglycans of the extracellular environment: Clues from the gene and protein side offer novel perspectives in molecular diversity and function. FASEB J 10: 598–614. Isaksson, O.G.P., A. Lindahl, A. Nilsson, J. Isgaard (1987) Mechanism of the stimulatory effect of growth hormone on longitudinal bone growth. Endocr Rev 8: 426–438. Islam, N., T.M. Haqqi, K.J. Jepsen, M. Kraay, J.F. Welter, V.M. Goldberg, C.J. Malemud (2002) Hydrostatic pressure induces apoptosis in human chondrocytes from osteoarthritic cartilage through up-regulation of tumor necrosis factor·, inducible nitric oxide synthase, p53, c-myc and bax-· and suppression of bcl-2. J Cell Biochem 87: 266–178. Islam, S., N. Islam, T. Kermode, B. Johnstone, H. Mukhar, R.W. Moskowitz, V.M. Goldberg, C.J. Malemud, T.M. Haqqi (2000) Involvement of caspase-3 in epigallocatechin-3-gallatemediated apoptosis of human chondrosarcoma cells. Biochem Biophys Res Commun 270: 793–797. Islam, S., T. Kermode, D. Sultana, R.W. Moskowitz, H. Mukhtar, C.J. Malemud, V.M. Goldberg, T.M. Haqqi (2001) Expression profile of protein kinase genes in human osteoarthritis chondrocytes. Osteoarthritis Cartilage 9: 684– 693. Jiang, Y., H.K. Genant, I. Watt, M. Cobby, B. Bresnihan, R. Aitchison, D. McCabe (2000) A multicenter, double-blind dose-ranging, randomized, placebo-controlled study of recombinant human interleukin-1 receptor antagonist in patients with rheumatoid arthritis. Arthritis Rheum 43: 1001–1009. Juo, P., M.S. Woo, C.J. Kuo, P. Signorelli, H.P. Biemann, Y.A. Hannun, J. Blenis (1999) FADD is required for multiple signaling events downstream of the receptor Fas. Cell Growth Differ 10: 797–804. Kacena, M., G.A. Merrel, S. Konda, K.M. Wilson, Y. Xi, M.C. Horowitz (2001) Inflammation and bony changes at the temporomandibular joint. Cells Tissues Organs 169: 257–264. Kavanagh, E., M. Abiri, Y.S. Bland, D.E. Ashhurst (2002) Division and death of cells in developing synovial joints and long bones. Cell Biol Int 26: 679–688.
Cells Tissues Organs 2003;174:34–48
45
Kelly, M., C.M. Geigerman, G. Loo (2001) Epigallocatechin gallate protects U937 cell against nitric-oxide induced cell cycle arrest and apoptosis. J Cell Biochem 81: 647–658. Kim, S.-J., J.-W. Ju, C.-D. Oh, Y.-M. Yoon, W.K. Song, J.-H. Kim, Y.J. Yoo, O.-S. Bang, S.-S. Kang, J.-S. Chun (2002) ERK-1/2 and p38 kinase oppositely regulate nitric oxide-induced apoptosis of chondrocytes in association with p53, caspase-3 and differentiation status. J Biol Chem 277: 1332–1339. Kim, H.A., Y.J. Lee, S.C. Seong, K.W. Choe, Y.W. Song (2002) Apoptotic chondrocyte death in human osteoarthritis. J Rheumatol 27: 455– 462. Kluck, R.M., E. Bossy Wetzel, D.R. Green, D.D. Newmeyer (1997) The release of cytochrome c from mitochondria: A primary site for Bcl-2 regulation of apoptosis. Science 275: 1132– 1136. Ko, K.S., C.A.G. McCulloch (2001) Intracellular mechanotransduction: Cellular circuits that coordinate tissue responses to mechanical loading. Biochem Biophys Res Commun 285: 1077–1083. Kobayashi, K., H. Mishima, S. Hashimoto, R.S. Goomer, F.L. Harwood, M. Lotz, H. Moriya, D. Amiel (2001) Chondrocyte apoptosis and regional expression of nitric oxide in the medical meniscus following partial meniscectomy. J Orthop Res 19: 802–808. Kobayashi, T., K. Okamoto, T. Kobata, T. Hasunuma, T. Kato, H. Hamada, K. Nishioka (2000) Novel gene therapy for rheumatoid arthritis by FADD gene transfer: Induction of apoptosis of rheumatoid synoviocytes but not chondrocytes. Gene Ther 7: 527–533. Kühn, K., S. Hashimoto, M. Lotz (2000) IL-1ß protects human chondrocytes from CD-95-induced apoptosis. J Immunol 164: 2233–2239. Kühn, K., M. Lotz (2001) Regulation of CD95 (Fas/APO-1)-induced apoptosis in human chondrocytes. Arthritis Rheum 44: 1644– 1653. Kuno, K., Y. Okada, H. Kawashima, H. Nakamura, M. Miyasaka, H. Ohno, K. Matsushima (2000) ADAMTS-1 cleaves a cartilage proteoglycan, aggrecan. FEBS Lett 478: 241–245. Levine, A. (1997) p53, the cellular gatekeeper for growth and division. Cell 88: 323–331. Li, P., D. Nijhawan, I. Budihardjo, S.M. Srinivasula, M. Ahmad, E.S. Alnemri, X. Wang (1997) Cytochrome c and dATP-dependent formation of Apaf-1/caspase 9 complex initiates an apoptotic protease cascade. Cell 91: 479–489. Liacini, A., J. Sylvester, W.Q. Li, M. Zafarullah (2002) Inhibition of interleukin-1-stimulated MAP kinases, activating protein-1 (AP-1) and nuclear factor kappa B (NF-ÎB) transcription factors down-regulates matrix metalloproteinase gene expression in articular chondrocytes. Matrix Biol 21: 251–262. Lin, H., H.-M. Zhang, C. Chen, B.D. Chen (2000) Apoptosis of Mo7e leukemia cells is associated with the cleavage of Bcl-2 into a shortened fragment that is not functional for heterodimerization with Bcl-2 and Bax. Exp Cell Res 261: 180–186.
46
Lipsky, P.E., D.M. van der Heijde, E.W. St. Clair, D.E. Furst, F.C. Breedveld, J.R. Kalden, J.S. Smolen, M. Weisman, P. Emery, M. Feldmann, G.R. Harriman, R.N. Maini (2000) Infliximab and methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. N Engl J Med 343: 1594–1602. Loesser, R.F., G. Shanker (2000) Autocrine stimulation by insulin-like growth factor 1 and insulin-like growth factor 2 mediates chondrocyte survival in vitro. Arthritis Rheum 43: 1552– 1559. Malemud, C.J. (1999) Fundamental pathways in osteoarthritis: An overview. Front Biosci 4: d659–d661. Malemud, C.J., V.M. Goldberg, L.M. Miller (1987) Synthesis of a ‘free’ form of hyaluronic acid by articular chondrocytes in monolayer culture. Biochem Int 14: 987–995. Malemud, C.J., T.M. Haqqi, K.J. Jepsen, N. Islam, M. Kraay, J.Welter, V.M. Goldberg (2001) Induction of apoptosis in human chondrocytes by hydrostatic pressure (abstract). Osteoarthritis Cartilage 9(suppl B): S8. Malemud, C.J., W. Killeen, T.M. Hering, A.F. Purchio (1991) Enhanced sulfated-proteoglycan core protein synthesis by incubation of rabbit chondrocytes with recombinant TGF-ß1. J Cell Physiol 149: 152–159. Malfait, A.-M., R.-Q. Liu, K, Ijiri, S. Komiya, M.D. Tortorella (2002) Inhibition of ADAM-TS4 and ADAM-TS5 prevents aggrecan degradation in osteoarthritic cartilage. J Biol Chem 277: 22201–22208. Martel-Pelletier, J., N. Alaaeddine, J.-P. Pelletier (1999b) Cytokines and their role in the pathophysiology of osteoarthritis. Front Biosci 4: d694–d703. Martel-Pelletier, J., J.-M. Cloutier, J.-P. Pelletier (1990) Cathepsin B and cysteine protease inhibitors in human OA: Effect of intra-articular steroid injections. J Orthop Res 8: 336–344. Martel-Pelletier, J., J. Di Battista, D. Lajeunesse (1999a) Biochemical factors in joint articular tissue degradation in osteoarthritis; in Reginster, J.-Y., J.-P. Pelletier, J. Martel-Pelletier, Y. Henrontin (eds): Osteoarthritis – Clinical and Experimental Aspects. Berlin, Springer, pp 156–187. Martel-Pelletier, J., M.P. Faure, R. McCollum, F. Mineau, J.-M. Cloutier, J.-P. Pelletier (1991a) Plasmin, plasminogen activators and inhibitor in human osteoarthritic cartilage. J Rheumatol 18: 1863–1871. Martel-Pelletier, J., R. McCollum, N. Fujimoto, K. Obata, J.-M. Cloutier, J.-P. Pelletier (1994) Excess of metalloproteinases over tissue inhibitor of metalloproteinase may contribute to cartilage degradation in osteoarthritis and rheumatoid arthritis. Lab Invest 70: 807–815. Martel-Pelletier, J., J.-P. Pelletier, C.J. Malemud (1988) Activation of neutral metalloproteases in human osteoarthritic knee cartilage: Evidence for degradation in the core protein of sulphated proteoglycan. Ann Rheum Dis 47: 801– 808.
Cells Tissues Organs 2003;174:34–48
Martel-Pelletier, J., M. Zafarullah, S. Kodama, J.-P. Pelletier (1991b) In vitro effects of interleukin 1 on the synthesis of metalloproteases, TIMP, plasminogen activators and inhibitors in human articular cartilage. J Rheumatol 18(suppl 27): 80–84. Martin, J.A., S.M. Ellerbrook, J. Buckwalter (1997) Age-related decline in chondrocyte response to insulin-like growth factor-I: The role of growth factor binding proteins. J Orthop Res 15: 491– 498. Matsui, K., A. Fine, B. Zhu, A. Marshak-Rothstein, S.T. Ju (1998) Identification of two NF-ÎB sites in CD95 ligand (Fas ligand) promoter: Functional analysis in T cell hybridoma. J Immunol 161: 3469–3473. McInnes, I.B, B.P. Leung, M. Field, X.Q. Wei, F.P. Huang, R.D. Sturrock, A. Kinninmonth, J. Weidner, R. Mumford, F.Y. Liew (1996) Production of nitric oxide in the synovial membrane of rheumatoid and osteoarthritis patients. J Exp Med 184: 1519–1524. Mehraban, F., M.H. Tindal, M.M. Proffitt, R.W. Moskowitz (1997) Temporal pattern of cysteine endopeptidase (cathepsin B) expression in cartilage and synovium from experimental osteoarthritis rabbit knees: Gene expression in chondrocytes in response to interleukin-1 and matrix depletion. Ann Rheum Dis 56: 108– 115. Mengshol, J.A., M.P. Vincenti, C.I. Coon, A. Barchowsky, C.E. Brinckerhoff (2000) Interleukin1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-Jun N-terminal kinase and nuclear factor kappaB: Differential regulation of collagenase 1 and collagenase 3. Arthritis Rheum 43: 801–811. Messai, H., Y. Duchossoy, A.-M. Khatib, A. Panasyuk, D.R. Mitrovic (2000) Articular chondrocytes from aging rats respond poorly to insulinlike growth factor-1: An altered signaling pathway. Mech Ageing Dev 115: 21–37. Middleton, J., A. Manthey, J. Tyler (1996) Insulinlike growth factor (IGF) receptor, IGF-I, interleukin-1ß (IL-1ß), and IL-6 mRNA expression in osteoarthritic and normal human cartilage. J Histochem Cytochem 44: 133–141. Middleton, J.F.S., J.A. Tyler (1992) Upregulation of insulin-like growth factor I gene expression in lesions of osteoarthritic human articular cartilage. Ann Rheum Dis 51: 440–447. Mohtai, M., R.L. Smith, D.J. Schurman, Y. Tsuji, F.M. Torti, N.I. Hutchinson, G. Stetler-Stevenson, G.I. Goldberg (1993) Expression of a 92kDa type IV collagenase/gelatinase (gelatinase B) in osteoarthritic cartilage and its induction in normal human articular cartilage by interleukin-1. J Clin Invest 92: 179–185. Morales, T.I. (1995) The role of signaling factors in articular cartilage homeostasis and osteoarthritis; in Kuettner, K.E., V.M. Goldberg (eds): Osteoarthritic Disorders. Rosemont, American Academy of Orthopaedic Surgeons, pp 261– 270.
Malemud/Islam/Haqqi
Murrell, G.A.C., D. Jang, R.J. Williams (1995) Nitric oxide activated metalloprotease enzymes in articular cartilage. Biochem Biophys Res Commun 206: 15–21. Muzio, M., A.M. Chinnaiyan, F.C. Kischkel, K. O’Rourke, A. Shevchenko, J. Ni, C. Scaffidi, J.D. Bretz, M. Zhang, R. Gentz, M. Mann, P.H. Krammer, M.E. Peter, V.M. Dixit (1996) FLICE, a novel FADD-homologous ICE/ced-3like protease, is recruited to the CD95 (Fas/ apo-1) death-inducing signaling complex. Cell 85: 817–825. Nicholson, D.W. (2001) Apoptosis. Baiting death inhibitors. Nature 410: 33–34. Ninomiya-Tsugi, J., K. Kashimoto, A. Hiyama, J.I. Inoue, Z. Cao, K. Matsumoto (1999) The kinase TAK1 can activate NIK-IÎB as well as MAP kinase cascade in the IL-1 signalling pathway. Nature 398: 252–256. Notoya, K., D.G. Jovanovic, P. Raboul, J. MartelPelletier, F. Mineau, J.-P. Pelletier (2000) The induction of cell death in human osteoarthritis chondrocytes by nitric oxide is related to the production of prostaglandin E2 via the induction of cyclooxygenase-2. J Immunol 165: 3402–3410. Nuttall, M.E. (2001) Drug discovery and target validation. Cells Tissues Organs 169: 265–271. O’Connor, L., A. Strasser (1999) The Bcl-2 protein family; in Kumar, S. (ed): Apoptosis: Biology and Mechanisms. Heidelberg, Springer, pp 173–207. Okada, Y., M. Shinmei, O. Tanaka, K. Naka, A. Kimura, I. Nakanishi, M. Bayliss, K. Iwata, H. Nagase (1992) Localization of matrix metalloproteinase 3 (stromelysin) in osteoarthritic cartilage and synovium. Lab Invest 66: 680–690. Pelletier, J.-P., J. Martel-Pelletier, S.B. Abramson (2001) Osteoarthritis, an inflammatory disease. Potential implication for the selection of new therapeutic targets. Arthritis Rheum 44: 1237–1247. Pelletier, J.-P., R. McCollum, J.-M. Cloutier, J. Martel-Pelletier (1995) Synthesis of metalloproteases and interleukin-6 (IL-6) in human osteoarthritic synovial membrane is an IL-1 mediated process. J Rheumatol 22: 109–114. Pelletier, J.-P., F. Mineau, M.P. Faure, J. MartelPelletier (1990) Imbalance between the mechanisms of activation and inhibition of metalloproteases in the early lesions of experimental osteoarthritis. Arthritis Rheum 33: 1466– 1476. Petterson, I., Y. Figenshau, E. Olsen, W. Bakkelund, B. Smedsröd, B. Sveinbjörnsson (2002) Tumor necrosis factor-related apoptosis-inducing ligand induces apoptosis in human articular chondrocytes in vitro. Biochem Biophys Res Commun 296: 671–676. Polyak, K., Y. Xia, J.L. Zweier, K.W. Kinzler, B. Vogelstein (1997) A model for p53-induced apoptosis. Nature 389: 300–305. Poole, A.R. (1999) An introduction to the pathophysiology of osteoarthritis. Front Biosci 4: d662–d670.
Pathophysiology of Osteoarthritis
Prisco, M., A. Hongo, M.G. Rizzo, A. Sacchi, R. Baserga (1997) The insulin-like growth factor I receptor is a physiologically relevant target of p53 in apoptosis caused by interleukin-3 withdrawal. Mol Cell Biol 17: 1084–1092. Puzas, J.E., J.-M. Landreau, R. Tallents, J. Albright, E.D. Schwarz, R. Landesberg (2001) Degradative pathways in tissues of the temporomandibular joint. Cells Tissues Organs 169: 248–256. Reed, J.C. (1994) Bcl-2 and the regulation of programmed cell death. J Cell Biol 124: 1–6. Reed, J.C. (1997) Cytochrome c: Can’t live with it; can’t live without it. Cell 91: 559–562. Reed, J.C. (2000) Mechanisms of apoptosis. Am J Pathol 157: 1415–1430. Sachs, B.L., V.M. Goldberg, L. Getzy, R.W. Moskowitz, C.J. Malemud (1982) A histopathologic differentiation of tissue types in human osteoarthritic cartilage. J Rheumatol 9: 210–216. Saeed Sheikh, M., A.J. Fornace, Jr. (2000) Role of p53 family members in apoptosis. J Cell Physiol 182: 171–181. Salvesen, G.S., V.M. Dixit (1997) Caspase activation: The induced-proximity model. Proc Natl Acad Sci USA 96: 10964–10967. Sandell, L.J. (1995a) Molecular biology of collagens in normal and osteoarthritic cartilage; in Kuettner, K.E., V.M. Goldberg (eds): Osteoarthritic Disorders. Rosemont, American Academy of Orthopaedic Surgeons, pp 131–146. Sandell, L.J., H. Chansky, O. Zamparo, T.M. Hering (1995b) Molecular biology of cartilage proteoglycans and link protein; in Kuettner, K.E., V.M. Goldberg (eds): Osteoarthritic Disorders. Rosemont, American Academy of Orthopaedic Surgeons, pp 117–130. Sandy, J.D., C.R. Flannery, P.J. Neame, L.S. Lohmander (1992) The structure of aggrecan fragments in human synovial fluid: Evidence for involvement in osteoarthritis of a novel proteinase which cleaves the Glu 373-Ala 374 bond of the interglobular domain. J Clin Invest 89: 1512–1516. Sandy, J.D., J. Westling, R.D. Kenagy, M.L. IruelaArispe, C. Verscharen, J.C. Rodriguez-Mazaneque, D. Zimmermann, J.M. Lemire, J.W. Fischer, T.N. Wight, A.W. Clowes (2001) Versican V1 proteolysis in human aorta in vivo occurs at Glu441-Ala442 bond, a site that is cleaved by recombinant ADAMTS-1 and ADAMTS-4. J Biol Chem 276: 13372–13378. Shakibaei, M., G. Schulze-Tanzil, P. de Souza, T. John, M. Rahmanzadeh, R. Rahmanzadeh, H.J. Merker (2001) Inhibition of mitogen-activated protein kinase induces apoptosis of human chondrocytes. J Biol Chem 276: 13289– 13294. Singh, R., S. Ahmed, N. Islam, V.M. Goldberg, T.M. Haqqi (2002) Epigallocatechin-3-gallate inhibits interleukin-1ß-induced expression of nitric oxide synthase and production of nitric oxide in human chondrocytes. Suppression of nuclear factor ÎB activation by degradation of the inhibitor of nuclear factor ÎB. Arthritis Rheum 46: 2079–2086.
Singh, R., S. Ahmed, T. Kermode, C.J. Malemud, R.W. Moskowitz, T.M. Haqqi (2001) Epigallocatechin-3-gallate (EGCG) inhibits interleukin-1ß-induced activation of mitogen-activated protein kinases (p-38MAPK, JNK/ SAPK), nuclear factor-ÎB (NF-ÎB), expression of nitric oxide synthase, production of nitric oxide and IL-6 in human osteoarthritis chondrocytes (abstract). Osteoarthritis Cartilage 9(suppl): S40. Singh, R., S. Ahmed, C.J. Malemud, V.M. Goldberg, T.M. Haqqi (2003) Epigallocatechin-3gallate selectively inhibits interleukin-1ß-induced expression of mitogen activated protein kianse subgroup c-Jun-N-terminal kinase in human osteoarthritis chondrocytes. J Orthop Res 21: 102–107. Sirum, K.L., C.E. Brinckerhoff (1989) Cloning of the genes for human stromelysin and stromelysin 2: Differential expression in rheumatoid synovial fibroblasts. Biochemistry 28: 8691– 8698. Smaili, S.S., Y.T. Hsu, K.M. Sanders, J.T. Russell, R.L. Roule (2001) Bax translocation to mitochondria subsequent to rapid loss of mitochondrial membrane potential. Cell Death Differ 8: 909–920. Smith, R.L. (1999) Degradative enzymes in osteoarthritis. Front Biosci 4: d704–d712. Sun, Y., J.M. Cheung, J. Martel-Pelletier, J.P. Pelletier, L. Wenger, R.D. Altman, D.S. Howell, H.S. Cheung (2000a) Wild type and mutant p53 differentially regulate the gene expression of human collagenase-3 (hMMP-13). J Biol Chem 275: 11327–11332. Sun, X., L. Wie, C.M. Davis, Q. Chen (2002) Antifas induction of apoptosis in human osteoarthritic chondrocytes through activating transcription factor 2 and caspase-3: Dependence on p38 mitogen-activated protein kinase (abstract 0195). Trans Orthop Res Soc 27. Sun, B.H., X.P. Zhao, B.J. Wang, D.L. Yang, L.J. Hao (2000b) FADD and TRADD expression and apoptosis in primary hepatocellular carcinoma World J Gastroenterol 6:223–227. Susin, S.A., N. Zamzami, M. Castedo, E. Daugas, H.G. Wang, S. Geley, F. Fassy, J.C. Reed, G. Kroemer (1997) The central executioner of apoptosis: Multiple connections between protease activation and mitochondria in Fas, APO-1/CD95- and ceramide-induced apoptosis. J Exp Med 186: 25–37. Tanaka, M., M.E. Fuentes, K. Yamaguchi, M.H. Durnin, S.A. Dalrymple, K.L. Hardy, D.V. Goeddel (1999) Embryonic lethality, liver degeneration and impaired NF-ÎB activation in IKK-ß-deficient mice. Immunity 10: 421–429. Thonar, E.J.-M.A., K. Matsuda, D.H. Manicourt, K.E. Kuettner (1999) Structure and function of normal human articular cartilage; in Reginster, J.-Y., J.-P. Pelletier, J. Martel-Pelletier, Y. Henrontin (eds): Osteoarthritis – Clinical and Experimental Aspects. Berlin, Springer, pp 1– 19. Thornberry, N.A., Y. Lazebnik (1998) Caspases: Enemies within. Science 281: 1312–1316.
Cells Tissues Organs 2003;174:34–48
47
Tomicic, M.T., B. Kaina (2000) Hamster bcl-2 protein is cleaved in vitro and in cells by caspase-9 and caspase-3. Biochem Biophys Res Commun 281: 404–408. Tortorella, M.D., T.C. Burn, M.A. Pratta, I. Abbaszade, J.M. Hollis, R. Liu, S.A. Rosenfeld, R.A. Copeland, C.P. Decicco, R. Wynn, A. Rockwell, F. Yang, J.L. Duke, K. Solomon, H. George, R. Bruckner, H. Nagase, Y. Itoh, D.M. Ellis, H. Ross, B.H. Wiswall, K. Murphy, M.C. Hillman, Jr., G.F. Hollis, R.C. Newton, R.L. Magolda, J.M. Trzaskos, E.C. Arner (1999) Purification and cloning of aggrecanase-1: A member of the ADAMTS family of proteins. Science 284: 1664–1666. Urayama, S., A. Kawakami, T. Nakashima, S. Yamasaki, A. Hida, H. Ida, M. Kamamichi, H. Nakamura, T. Origuchi, K. Migita, Y. Kawabe, K. Eguchi (2001) New disease-modifying antirheumatic drug 2-acetylthiomethyl-4-(4-methylphenyl)-4-oxobutanoic acid (KE-298) selectively augments activation-induced T cell death. J Lab Clin Med 138: 11–17. Van den Berg, W.B., P.M. van der Kraan, H.M. van Beuningen (1999) Role of growth factors and cartilage repair; in Reginster, J.-Y., J.-P. Pelletier, J. Martel-Pelletier, Y. Henrontin (eds): Osteoarthritis – Clinical and Experimental Aspects. Heidelberg, Springer, pp 188–209. Verhagen, A.M., D.L. Vaux (1999) Molecular mechanisms in apoptosis; an overview; in Kumar, S. (ed): Apoptosis: Biology and Mechanisms. Berlin, Springer, pp 11–24. Vincenti, M.P., C.E. Brinckerhoff (2001) The potential of signal transduction inhibitors for the treatment of arthritis: Is it all just JNK? J Clin Invest 108: 181–183.
48
Vuolteenaho, K., T. Moilanen, M. Hämäläinen, E. Moilanen (2002) Effects of TNF-· antagonists on nitric oxide production in human cartilage. Osteoarthritis Cartilage 10: 327–332. Webb, G.R., C.I. Westacott, C.J. Elson (1997) Cartilage tumor necrosis factor alpha receptors and focal loss of cartilage in osteoarthritis. Osteoarthritis Cartilage 5: 427–437. Webb, G.R., C.I. Westacott, C.J. Elson (1998) Osteoarthritic synovial fluid and synovium supernatants up-regulate tumor necrosis factor receptors on human articular chondrocytes. Osteoarthritis Cartilage 6: 167–176. Westacott, C.I., R.M. Atkins, P.A. Dieppe, C.J. Elson (1994) Tumor necrosis factor-alpha receptor expression on chondrocytes isolated from human articular cartilage. J Rheumatol 21: 1710–1715. Westacott, C.I., A.F. Baraket, L. Wood, M.J. Perry, P. Nelson, I. Bisbinas, L. Armstrong, A.B. Miller, C.J. Elson (2000) Tumor necrosis factor alpha can contribute to focal loss of cartilage in osteoarthritis. Osteoarthritis Cartilage 8: 213– 221. Wolter, K.G., Y.-T. Hsu, C.L. Smith, A. Nechushtan, X.-G. Xi, R.J. Youle (1997) Movement of Bax from cytosol to mitochondria during apoptosis. J Cell Biol 139: 1281–1292. Wright, J.K., T.E. Cawston, B.I. Hazleman (1991) TGF-ß stimulates the production of tissue inhibitor of metalloproteases (TIMP) by human synovium and skin fibroblasts. Biochim Biophys Acta 1094: 207–210.
Cells Tissues Organs 2003;174:34–48
Xerri, L., E. Devilard, R. Bouabdallah, A.-M. Stoppa, J. Hassoun (1999) FADD expression and caspase activation in B-cell lymphomas resistant to Fas-mediated apoptosis. Br J Haematol 106: 652–661. Yang, Y.L., X.M. Li (2000) The IAP family: Endogenous caspase inhibitors with multiple biological activities. Cell Res 10: 169–177. Yang, J., Y. Lin, Z. Guo, J. Cheng, J. Huang, L. Deng, W. Liao, Z. Chen, Z.-G. Liu, B. Su (2000) The essential role of MEKK3 in TNFinduced NF-ÎB activation. Nat Immunol 2: 620–624. Yang, J., X. Liu, K. Bhalla, C.N. Kim, A.M. Ibrado, J. Cai, T.-I. Peng, D.P. Jones, X. Wang (1997) Prevention of apoptosis by Bcl-2: Release of cytochrome c from mitochondria blocked. Science 275: 1129–1132. Yang, R., Z. Yan, F. Chen, G.K. Hansson, R. Kiessling (2002) Hyaluronic acid and chondroitin sulfate A rapidly promote differentiation of immature DC with upregulation of costimulatory and antigen-presenting molecules and enhancement of NF-ÎB and protein kinase activity. Scand J Immunol 55: 2–13. Yin, X.-M., Z.N. Oltval, S.J. Korsmeyer (1994) BH1 and BH2 domains are required for inhibition of apoptosis and heterodimerization with Bax. Nature 369: 321–323. Zafarullah, M., J. Su, J. Martel-Pelletier, J.A. Di Battista, B.G. Costello, W.G. Stetler-Stevenson, J.-P. Pelletier (1996) Tissue inhibitor of metalloproteinase-2 (TIMP-2) mRNA is constitutively expressed in bovine, human normal and osteoarthritic articular chondrocytes. J Cell Biochem 60: 211–217.
Malemud/Islam/Haqqi
Cells Tissues Organs 2003;174:49–62 DOI: 10.1159/000070574
Analgesic Effect of Elastoviscous Hyaluronan Solutions and the Treatment of Arthritic Pain Endre A. Balazs Matrix Biology Institute, Ridgefield, N.J., USA
Key Words Arthritic pain W Hyaluronan W Hylan W Temporomandibular joint W Viscosupplementation
Abstract Elastoviscous hyaluronan solutions have an analgesic effect when injected intra-articularly in animal and human joints. This was first discovered using animal behavioral models and later confirmed in neurophysiological studies in cat and rat joints. These studies on both normal and experimentally produced arthritis in joints confirmed that only elastoviscous solutions of hyaluronan or certain of its derivatives (hylans) have a desensitizing effect on nociceptive sensory receptors. Recently, this desensitizing effect of elastoviscous hyaluronan solutions was also demonstrated on intact or on isolated patches of oocyte cell membranes. Viscosupplementa-
Abbreviations used in this paper
DDN DDR IL-1ß MW NIF-NaHA TMJ
displaced disc without reduction displaced disc with reduction interleukin-1ß molecular weight noninflammatory fraction of Na-hyaluronan temporomandibular joint
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0049$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
tion, the exchange of pathological synovial fluid in arthritic joints with pure elastoviscous solutions of hyaluronan or hylans, is a widely accepted therapeutic modality used to provide long-lasting analgesia in human knee joints. The clinical studies performed on human and animal temporomandibular joints since the mid1970s are reviewed. These trials used three distinctly different preparations made from hyaluronan of different average molecular weight, polydispersity and, consequently, different elastoviscous properties. These differences are demonstrated and the consequences on the potential efficacy of the preparations are discussed. Copyright © 2003 S. Karger AG, Basel
Introduction
This article reviews the discovery of the analgesic effect of elastoviscous hyaluronan solutions and the neurophysiological research that supports the hypothesis that this effect is the basis for the therapeutic application of this biopolymer in the treatment of osteoarthritis and traumatic arthritis in human and veterinary medicine. It also reviews the clinical studies performed since the late 1970s on the temporomandibular joint (TMJ) using very different hyaluronan preparations to evaluate their use in the treatment of the diseases of this joint. The considerable clinical work on viscosupplementation of the TMJ
Endre A. Balazs Matrix Biology Institute 65 Railroad Avenue Ridgefield, NJ 07657 (USA) Tel. +1 201 945 5885, Fax +1 201 945 1430, E-Mail
[email protected]
during the past three decades has not resulted in acceptance of this relatively new therapeutic modality for treating arthritis in this joint. This review questions the reason for this and whether or not viscosupplementation can be regarded as a universal treatment of chronic pain in all joints.
The Analgesic Effect of Elastoviscous Hyaluronan Solutions
In the 1960s, Balazs and coworkers discovered that hyaluronan solutions made of highly purified Na-hyaluronan of relatively high average molecular weight (12 million) in concentrated (1%) solutions (dissolved in phosphate-buffered physiological NaCl) had an analgesic effect on joint pain in animals and humans [for review of this early work, see Balazs, 1971]. It was important that the hyluronan used was a highly purified fraction of the molecule prepared from human umbilical cord or rooster comb, and that it was sterile, pyrogen-free and caused no inflammation when injected into various tissues and tissue spaces of animals (vitreous, joints, peritoneal cavity) [Balazs, 1971; Denlinger and Balazs, 1980]. This highly purified fraction with only trace amounts of protein was also nonimmunogenic and was called the noninflammatory fraction of Na-hyaluronan (NIF-NaHA). The first preparation made for medical use was called by the trade name Healon® (Biotrics, Arlington, Mass., USA, 1972). Later the same preparation was named Hylartil® (Pharmacia, Uppsala, Sweden, 1981) for use in veterinary medicine. The analgesic effect was first shown in dogs with experimentally induced, traumatic arthritis and in racehorses with traumatic arthritis and posttraumatic osteoarthritis [Rydell et al., 1970]. It was also discovered that the pain relief (of lameness in dogs and in horses) lasted much longer than the residence time of the injected hyaluronan in the joint. The analgesic effect was the consequence of the elastoviscous properties of the hyaluronan, because injecting the same amount and concentration of the molecule in a nonelastoviscous form (low average molecular weight) failed to produce long-lasting analgesia [Rydell and Balazs, 1971]. During the early 1970s, clinical studies on osteoarthritic patients demonstrated the analgesic effect of Healon. The elastoviscous solutions of Healon prepared from human umbilical cord or rooster comb, when injected once or twice into the knee joint after the effusion was removed, produced an analgesic effect that lasted for sev-
50
Cells Tissues Organs 2003;174:49–62
eral months [Balazs, 1971]. This preparation of hyaluronan was used extensively in clinical trials during the 1970s [for review, see Adams, 1998; Denlinger, 1998]. The same preparations (Healon and Hylartil) were used for the first time by Kopp and Wenneberg [1979] in the TMJ to test their analgesic effect. More than a decade later, the animal behavioral studies on dogs and horses were extended to rats when pain caused by the intraperitoneal injection of bradykinin was alleviated by simultaneous injection of hyaluronan of the same concentration used in the joint studies (1%) but of various average molecular weights (MW). The analgesic effect was directly proportional to the molecular weight and, therefore, to the elastoviscosity of the solutions used [Miyazaki et al., 1984; Gotoh et al., 1988]. In one other important behavioral study, Japanese researchers measured the bradykinin concentration in the rat knee joint after injection of monosodium urate crystals. One hour before the crystals were injected, hyaluronans (1%) of various average MW were injected into the same joint. When the average MW was 2 million, significantly less bradykinin was found in the joint after the injection of the crystals. Hyaluronan solutions (1%) made with average MW of 0.47 and 0.95 million showed no effect [Aihara et al., 1992]. By the late 1980s, it was well established by animal behavioral and human clinical studies that elastoviscous solutions of NIF-NaHA could have an analgesic effect when injected into arthritic joints if applied appropriately to patients with acute arthritic pain. It was necessary to confirm these analgesic effects with neurophysiological methods. The question to answer was whether or not elastoviscous hyaluronan solutions, when applied to the sensory nerve endings in animals joints, decreased the frequency and intensity of nerve impulses and, therefore, acted as analgesics on nociceptor terminals. In the mid-1990s, electrophysiological studies were initiated by Belmonte, Balazs and their coworkers using cat knee joints to demonstrate the analgesic effect of elastoviscous hylan G-F 20 solutions. Hylan G-F 20 (trade name Synvisc®) is a chemically modified hyaluronan with a very high average molecular weight (approximately 6 million) developed and manufactured by Biomatrix (now Genzyme Biosurgery; Ridgefield, N.J., USA) [Balazs and Leshchiner, 1989; Balazs et al., 1999]. The derivation of the molecule does not affect the carboxylic acid and the N-acetyl groups of the molecule. When this highly elastoviscous solution of hylan G-F 20 was injected into the healthy cat knee, the rate of neural discharges of the nociceptive afferent fibers of the nerve innervating the syno-
Balazs
vial tissue produced by noxious movements (beyond the normal range of motion) of the joint were significantly attenuated. Nonelastoviscous solutions of the same or higher concentration made of low average molecular weight (F30,000) hylan and physiological buffer solution had no effect (fig. 1). The analgesic effect of this elastoviscous hylan solution was also tested on cat knees with acute inflammation produced by intra-articular injection of kaolin and carrageenan. Nerve discharges developed and reached a constant rate 2–3 h after injection. Ongoing discharges were present in the resting joint (pain at rest), and the rate of discharge increased substantially when the joint was moved within the normal range. Both ongoing and movement-evoked discharges were attenuated after intra-articular injection of the elastoviscous hylan G-F 20 solution. The injection of physiological saline or nonelastoviscous hylan G-F 20 had no effect [Pozo et al., 1997; Belmonte et al., 1998]. Nociceptors from the afferent fibers (A-delta and C fibers) of the medial articular nerve of rat knee were also studied before and after acute inflammatory arthritis was produced the same way as in cats. In these models, the elastoviscous hylan significantly reduced the discharge rate both in the healthy joints (movement over the normal range of motion) and inflamed joints (ongoing and movement-produced). Nonelastoviscous hylan had no effect [Pawlak et al., 2002]. The same rat knee model was used to evaluate the analgesic effect of three commercially available hyaluronan preparations for viscosupplementation [Gomis et al., 2002]. In this study, 55 nociceptive afferent fibers of the medial articular nerve were investigated, both in normal and in acutely inflamed joints of 50 adult rats. The joints were exposed to standardized movement stimuli in the normal range of motion of the joint as well as beyond the normal range. The stimuli were applied for 50 s, every 5 min for 2–3 h. The three hyaluronan-based preparations for viscosupplementation had very different elastoviscous properties due to the differences of the average molecular weights of the polymers used. The average molecular weight of Hyalgan® (Fidia) was 0.5 million, of Orthovisc® (Anika Therapeutics) 1.2 million, and of Synvisc 6 million (Biomatrix, now Genzyme Biosurgery). Phosphate-buffered physiological saline solution, the solvent of all three preparations, was also tested in this model. Hyalgan injected into the normal joint led to a slight (20%) increase in the rate of discharges, which, 100 min after application, became significantly greater than the control level. Hyalgan showed no effect on the rate of discharges in the inflamed joint by the end of the experi-
Analgesic Effect of Hyaluronan
Fig. 1. Comparison of the effect of elastoviscous and nonelastoviscous hylan solutions injected into the intact (healthy) knee of cats on the impulse discharges evoked by extending the joint beyond the normal range of motion. Average frequency values are shown during interstimulus period (ongoing activity, g) and during noxious movement-evoked nerve activity (i). a Mean frequency values before and after intra-articular injection of elastoviscous hylan G-F 20 (Synvisc) solution. b Mean frequency values before and during various periods of time after the injection of nonelastoviscous hylan G-F 20 solution. The only significant differences (p ! 0.05) could be found between the ongoing and movement-evoked impulses before injection, and 70–120 min after injection of elastoviscous hylan G-F 20 (Synvisc) [from Belmonte et al., 1998].
ment. Orthovisc had no effect in the normal joint, but significantly decreased the discharge rate in inflamed joints. This decrease was approximately 30% compared to control, but lasted only until 20 min after the injection. In contrast, Synvisc, with much greater elastoviscosity, decreased the discharges significantly during the entire re-
Cells Tissues Organs 2003;174:49–62
51
cording time, reaching significance 40 min after injection. A 70% reduction of discharge rate was observed both in normal and inflamed joints compared to control, and this reduction lasted to the end of the experiment [Gomis et al., 2002]. These electrophysiological studies confirmed the findings of the behavioral study that showed that the elastoviscosity of the preparation for viscosupplementation determines the level of the analgesic effect. Since elastoviscosity is the product of concentration and average molecular weight, the molecular weight is the determining factor in solutions of similar concentration. In these three commercial preparations, the concentration of the polymer was in the range of 0.8–1.5%, while the average molecular weight ranged from 0.5 to 6 million. These neurophysiological studies on normal and inflamed cat and rat knees clearly showed that the rheological properties (elasticity and viscosity) of the polymer (hyaluronan or hylan) used determine its analgesic effect. The neurophysiological findings were further confirmed in an in vitro cellular model of stretch-activated ion channels. So-called patch clamp recordings in intact frog oocytes (Xenopus laevis) and excised membrane patches, in outside-out and inside-out configuration, were used [de la Peña et al., 2002b]. Both the outside surface and the inside cytoplasmic surface of the oocyte were exposed to control salt solution (nonelastoviscous) or to elastoviscous hylan or nonelastoviscous hylan solutions. The concentration of hylan polymer was the same in both cases (0.8%) , but the average molecular weight of hylan in the elastoviscous solutions was 6 million, in contrast with 96,000 in the nonelastoviscous solution. The cell membranes of the oocytes were mechanically deformed by gradually increased suction through a micropipette. This suction was applied either on the surface of the intact oocytes or on the outside (external side) or the inside (cytoplasmic side) of the cell membrane patch. The hylan or the control salt solution bathed the entire external cell surface (in intact oocytes) or the opposite side (external or cytoplasmic) of the membrane patches where the suction pipette was attached. The suction caused deformation of the membrane, thereby causing mechanical opening of the channels. Electrical parameters were recorded as pulses, while a range of negative pressures was applied through the micropipette attached to the cell membrane or its patches. The activity of channels, that is their opening caused by the mechanical stimuli, can be recorded by the electrical activity, which is proportional to the number of channels opened in the deformed membrane patch. The activity of the stretch-activated channels was signifi-
52
Cells Tissues Organs 2003;174:49–62
cantly reduced when a cell membrane was bathed in elastoviscous hylan solution. The nonelastoviscous hylan and the control salt solutions had no such effect. To test that this effect was not due to a specific interaction between the high-MW hyaluronan chain and a cell surface receptor, elastoviscous solutions of DNA were also used in this experiment. The DNA solution, with rheological properties similar to those of hylan, had the same activity-reducing effect. It was confirmed in this cell model that the mechanosensitivity of stretch-activated ion channels decreased significantly when the cell membranes were exposed to elastoviscous solutions. One must conclude that the rheological properties, and not the chemical nature of such solutions, are responsible for this desensitization. The fact that this desensitization is also achieved from the cytoplasmic side of the cell membrane suggests that more than one type of mechanoreceptor may be involved in this process. One would expect that the mechanoreceptors have different structural definitions on the outside surface of the cell and on the cytoplasmic surface (fig. 2) [de la Peña et al., 2002]. While the oocyte experiments clearly demonstrate coupling between elastoviscous molecular structures and mechanoreceptors, one must keep in mind that experiments in inflamed cat and rat joints clearly indicate that elastoviscous hyaluronan solutions also have effects on ongoing nerve activities that are not induced by movement of the joint. This effect can be explained only by an interaction between the high-MW hyaluronan or hylan and a pain stimulatory chemical in the inflamed joint. Animal behavioral studies also indicated such interactions between high-MW hyaluronan and bradykinin. It is suggested, therefore, that hyaluronan, depending on its concentration and MW, acts as a detoxicating or scavenging molecule in the intercellular matrix of the joint. This activity of the molecule is based on its polyanionic character, that is binding cationic molecules by charge interactions or more specifically, through a hyaluronan-specific binding site of proteins (link module). Such an interaction between a great variety of proteins and hyaluronans has been widely reported during the last decade [see for review Day and Sheehan, 2001]. To support the hypothesis that elastoviscous hyaluronan can regulate the sensitivity of afferent sensory terminals, one must establish that these nerve terminals are exposed directly to the hyaluronan-containing intercellular matrix. While there is not much known about the function of innervations of the TMJ, one can assume that it is similar to that of other synovial joints in the body [Grigg, 2001]. Based on studies of the sensory nerve endings of
Balazs
Fig. 2. Response of oocyte cell membrane patches to elastoviscous and nonelastoviscous solutions. The mechanical stimulation (suction) through a micropipette is applied to the inside (cytoplasmic) of the patch while the test solutions cover the outside of the cell membrane patch (outside-out configuration). The micropipette with the attached cell membrane patch could be moved between the test chamber’s two compartments. Compartment a contained the control solution (physiological salt solution, Barth buffer), compartment b the test solutions. A central compartment filled with mineral oil separated the two compartments. The pipette with the attached patch was moved from one compartment to the other through the oil compartment. The negative pressure (in mm Hg) applied during the experi-
ments is shown on the left side of each recording. (Pipette voltage was fixed during all experiments.) a Ion channel opening activity induced by suction when the outside (external) of the patch was exposed first to the control solution (compartment a) and then to the elastoviscous Synvisc solution (compartment b). No activity was observed during exposure to the elastoviscous solution even when the negative pressure applied was nearly doubled. b Response to suction in the presence of control solution (compartment a) and response to nonelastoviscous Synvisc (compartment b). c Response to suction in the presence of control solutions (compartment a) and response to elastoviscous DNA solution (compartment b) [composite from de la Peña et al., 2002a, and from unpublished data from the authors].
human [Favia and Mairoano, 1995] and sheep TMJ [Tahmasebi-Sarvestani et al., 1996], one can assume that their microstructures are similar to those of the knee joint. Therefore, I believe that the ultrastructural work on the free nerve endings in the knee can be applied to TMJ. Work by Heppelmann et al. [1990] on the three-dimensional ultrastructure of sensory nerve endings very clearly demonstrated the relationship of the intercellular matrix and sensory terminals. They studied the ultrastructure of free nerve endings of sensory afferent nerve fibers in the
synovial tissue of the cat knee. They also created a threedimensional reconstruction of the unmyelinated group IV afferent sensory fiber terminal based on serial semi- and ultrathin sections photographed using the transmission electron microscope. This unique study established that the sensory endings of these fibers formed terminal trees that extended into dense connective tissue consisting of collagen fibers, mostly running parallel with the nerve fibers. The bundles of collagen fibers were always separated from the nerve elements by lightly stained intercel-
Analgesic Effect of Hyaluronan
Cells Tissues Organs 2003;174:49–62
53
Hyaluronan Content of the Healthy and Pathological TMJ
Fig. 3. The cross section of the sensory endings of two axons (group IV). On the micrograph, B1 is the sensory end bulb of one axon and A2 designates a distal part of the second axon. Note that these cross sections of the terminal sensory segments of both axons have cell membrane (axolemma) areas that are not covered by Schwann cells. The axolemma of axon A2 is directly exposed to the hyaluronan molecules represented in this picture by the lightly stained granular material between the cell membrane and the darkly stained, mostly cross-sectioned collagen fibers. The cell membrane of the B1 sensory end bulb that is in direct contact with the interfibrillar elastoviscous matrix is marked with a transverse arrow. !75,000. Bar = 0.5 Ìm [from Heppelmann et al., 1990, fig. 10].
lular space, which contained amorphous granulated elements. This is the space in which I believe that hyaluronan and other proteoglycans are located. Most importantly, these unmyelinated sensory endings (200–300 Ìm length) were not covered by the associated Schwann cells. Consequently, the cell membrane (called axolemma) was directly exposed to the hyaluronan-rich intercellular matrix (fig. 3). This is then the location where changes in the elastoviscosity of hyaluronan (changes in concentration and molecular weight) can influence the sensitivity of the sensory nerve endings. In a later work, Elfvin et al. [1998] confirmed the ultrastructural findings of Heppelmann et al. Based on the above ultrastructural work, I believe that these multiple receptive sites extending to the surface layer of the synovium are embedded in the elastoviscous hyaluronan solution that is entrapped between connective tissue cells and collagen fibers. The elastoviscosity of this perineural hyaluronan envelope represents a mechanical and chemical regulatory environment for the sensory terminal elements. Changes in the concentration and MW distribution of hyaluronan have a direct effect on the sensitivity of these terminals to mechanical and chemical stimuli occurring both during the resting and moving states of the joint in health and disease.
54
Cells Tissues Organs 2003;174:49–62
Smith et al. [1989] were the first to demonstrate the presence of hyaluronan in healthy adult human TMJ. In healthy fluids, only hyaluronan was detected, but in pathological fluids, chondroitin sulfate was also identified. This study stimulated further investigation [Shibata et al., 1998] using more sensitive methods to detect glycosaminoglycans (chondroitin sulfate, dermatan sulfate and keratin sulfate) in fluids collected from 28 female patients with various pathologies [osteoarthritis, rheumatoid arthritis, displaced disc with (DDR) or without (DDN) reduction]. In the aspirated fluids, hyaluronan concentration varied greatly without any relationship to the various pathologies. All fluids contained small amounts of chondroitin sulfate and dermatan sulfate (44– 53 Ìg/ml). The presence of a very low concentration of hyaluronan (F0.17 mg/ml) and an even lower level of chondroitin sulfate (F2 Ìg/ml) and dermatan sulfate (F4 Ìg/ml) was reported in TMJ synovial fluids collected from 15 patients with ‘internal derangement’ [Murakami et al., 1998]. The concentration of both sulfated glycosaminoglycans increased as the arthroscopically observed changes became more severe. There are no reports in the literature about the hyaluronan concentration of the synovial fluid in healthy adult human TMJ. It was, however, determined in cattle [Hatton and Swann, 1986] and sheep [Ishimaru et al., 2001]. In both species, the concentration is similar (2– 3 mg/ml).
Viscosupplementation of TMJ in Humans
Hyaluronan Preparations with Very High Elastoviscosity The first clinical trial of viscosupplementation of the TMJ was performed by Kopp and Wenneberg [1979] at the University of Lund (Sweden) in the late 1970s. They used the highly purified fraction of high MW (2–3 million) hyaluronan (NIF-NaHA) that was introduced under the trade name Hylartil as a therapeutic agent in veterinary medicine primarily for the treatment of pain (lameness in horses) in traumatic arthritis [see for review Balazs and Denlinger, 1985]. By that time, NIF-NaHA had also been introduced under the trade name Healon (Pharmacia) as a protective viscosurgical device to facilitate surgical procedures in the anterior segment of the eye and in vitreoretinal surgery [see for review Balazs, 1986]. Kopp
Balazs
et al. [1985, 1987] reported the comparative effectiveness of two Hylartil injections (0.5 ml, 2 weeks apart) into the superior compartment of the TMJ with the same volume of corticosteroid betamethasone (Celestone® bifas, Schering-Plough) applied in two injections. This singleblinded (patient) study was carried out between 1975 and 1981 on 33 patients (18 Hylartil, 15 corticosteroid) with a 2-year follow-up. Patients were selected on the basis of having pain of at least 6 months’ duration localized to the TMJ, and for failure of previous conservative treatment. The clinical parameters measured were pain, difficulties in opening the mouth and joint sound. The authors concluded: ‘Hyaluronate and corticosteroids by intra-articular route have a similar long-term effect on chronic arthritis of the TMJ. However, intra-articular sodium hyaluronate might be the best alternative due to reduced-risk for side effects’. According to the investigators: ‘There seems to be a subgroup of patients with pain and dysfunction probably of inflammatory nature that benefits from this intra-articular treatment.’ Kopp et al. [1991] carried out another clinical study comparing the efficacy of the same hyaluronan (Hylartil) with corticosteroid (Depo-Medron, Upjohn), but this time using a saline control (14 hyaluronan, 14 corticosteroid and 13 saline control patients). The patients had ‘classical or definitive RA’ involving the TMJ with pain at rest, during chewing, and upon maximum mouth opening. All patients received 0.7 ml of the test substances, twice intra-articularly 2 weeks apart, and were then followed for 4 weeks. The authors concluded that both hyaluronan and steroid treatment produced a significant improvement of the patient’s subjective symptoms, and that both treatments significantly increased the maximum voluntary mouth opening. However, the comprehensive clinical dysfunction score was reduced significantly in all three groups, including saline injection. Two case report studies were published using Healon (Pharmacia), a product identical to Hylartil but labeled for ophthalmic viscosurgical use. In a 24-year-old female patient suffering from painful nonreducing closed-lock disc displacement, the mandibular translation was assessed through electrognathography and electrovibratography (EVG). Healon (1.8 ml) was injected into the superior joint cavity. The use of electrognathography and electrovibratography showed the short-term beneficiary effect of this treatment [Fader et al., 1993]. The results of one Healon injection were also reported in 9 patients with anterior meniscal displacement with reduction and no evidence of meniscal perforation. During 8 weeks’ followup on half of the patients, temporary improvements (1–4
weeks’ duration) of symptoms (click, limited translation, and pain) were observed, but none of these improvements lasted for 8 weeks. The authors point out that the lack of effect, among other factors, may be attributed to the ‘inadequate amount of drug’ (one injection, 0.55 ml) [Edwards, 1994]. A case report was published using Synvisc (hylan G-F 20) for the treatment of painful TMJ with moderate erosion of the articular surfaces and osteophyte formation of both TMJ. After intra-articular injection of local anesthetic 1 ml of Synvisc was injected into the superior joint space. This was repeated twice at 2-week intervals. Clinical symptoms, such as crepitus, lateral excursive movements and pain, were greatly reduced for the 4-month follow-up [Yustin et al., 1995].
Analgesic Effect of Hyaluronan
Cells Tissues Organs 2003;174:49–62
Hyaluronan Preparation with Medium Elastoviscosity In 1993, Bertolami et al. [1993] published the results of a study using a hyaluronan preparation of the same concentration (1%) as previously reported in other studies, but with considerably lower elastoviscosity due to its lower average MW (Orthovisc, 1.5–2.0 million; Anika Research, Waltham, Mass., USA). In this randomized, double-blind, placebo-controlled study 121 patients entered the study with degenerative joint disease (41% of patients), DDR (47% of patients) or DDN (12% of patients). The test substances were injected once unilaterally in the superior joint space: 80 patients received hyaluronan and 41 patients physiological saline solution. Depending on the patients’ diagnostic criteria (degenerative joint disease, DDR or DDN), various symptoms and clinical signs (pain, tenderness over condyle, clicking, limited opening and range, lateral movements) and radiographic findings were recorded during 6 months at monthly visits. According to the authors, the primary beneficial effect (6 months’ duration) was observed for the DDR patients. The dysfunctional, the anamnestic and the improvement scores showed statistically and clinically significant differences in favour of hyaluronan injections throughout the 6month follow-up. Most importantly, only 3% of the patients with DDR relapsed to pretreatment levels compared to 31% with the control treatment. The authors concluded that ‘their study suggests that recovery (for at least 6 months) for patients with DDR became possible when the obstruction/trauma cycle is broken by a single injection of sodium hyaluronate’ [Bertolami et al., 1993]. Alpaslan and Alpaslan [2001] repeated the clinical trial with the same hyaluronan preparation. The test material was injected into the upper joint compartment after administration of an intra-articular anesthetic solution
55
(F3.5 ml) and extensive irrigation of the joint space with 200–300 ml sterile saline solution. Then 26 joints of 23 patients were injected with hyaluronan and 15 were not injected. All were evaluated in 10 visits during 24 months. The authors concluded that although the arthrocentesis with this extensive lavage alone was therapeutic, the postlavage injection of hyaluronan improved the results. In joints with a closed-lock variant of internal derangement hyaluronan injection resulted in statistically significant improvement over control, assessed by maximal mouth opening, lateral movement, pain, noise and jaw function. According to the authors, ‘Although patients benefited from both technologies (‘‘arthrocentesis with lavage with or without hyaluronan injection’’) arthrocentesis with injection of sodium hyaluronate seemed to be superior to arthrocentesis alone especially in patients with closed lock.’ A third study using Orthovisc for viscosupplementation was recently reported [Hepguler et al., 2002]. Thirty-eight patients with DDR were injected unilaterally with hyaluronan or saline solution (0.5 ml twice, 1 week apart). Evaluations before treatment and 1 and 6 months after treatment were made using a clinical dysfunction index (Helcima) and by the intensity of joint vibration during opening and closing measured by accelerometer. In the control group, only the pain intensity decreased significantly at 1and 6-month follow-up. Most importantly, the hyaluronan-treated group, at both follow-ups, showed significant improvement in all clinical parameters. The author concluded that this treatment (2 injections, 1 week apart) is an ‘effective treatment for a reducing displaced disc’. The use of a hyaluronan preparation with 1.2 million average MW (1% solution; Hunzghou Gallop Biological Products, China) was reported by Zhiyuan et al. [1998]. The TMJ cases used were displacement with or without reduction. The 69 TMJ cases were randomly distributed between test and control groups. All 69 TMJs were irrigated first with 5 ml of saline solution. After all the irrigation fluid was removed in 45 TMJs, 0.5–1.0 ml hyaluronan was injected into the lower cavity and 1 ml into the upper cavity. In 16 TMJs, a second hyaluronan injection was made 1 week later, and in 5 TMJs the injection was repeated 1 week later for a third time. In 24 control TMJs, after the irrigation process, 1 ml of 2% lidocaine was injected. The clinical evaluation at 1 week, 2 weeks and 1 month after the last injection was based on pain, decrease of click and degree of mouth opening. The authors came to the conclusion that the improvement was greater in the test group and ‘the difference between the test group and control group was statistically significant (p ! 0.01). How-
56
Cells Tissues Organs 2003;174:49–62
ever, the difference between disc displacement with reduction and without reduction was not significant’. Hyaluronan Preparations with Very Low Elastoviscosity Sato et al. [1997, 1999] used a very different hyaluronan preparation with a much lower average MW (0.7 million) in 1% solution and consequently with much lower elastoviscous properties than the previously used preparations (Artz® or Artzal®, Seikagaku, Kogyo, Japan). In two publications, they reported the results of studies on 76 patients with DDN of the TMJ. In the first study, the superior joint space in 26 patients was injected and withdrawn a few times with 1 ml xylocaine, and then the injected xylocaine was withdrawn. Finally, all fluid was removed from the joint before 1 ml of hyaluronan was injected. This procedure, called by the authors ‘lavage and hyaluronan injection’, was repeated once a week for 5 weeks. The control group consisted of 50 patients (49 female) that did not receive any treatment. At 3- and 6month follow-up, it was found that the experimental group had significantly lower levels of tenderness and a greater range of motion without pain. These authors later reported the results of a similar therapeutic strategy on a larger group of patients [Sato et al., 2001a, b]. Before each hyaluronan injection, 1 ml xylocaine (1%) was injected and withdrawn 5 times. This ‘lavage’ with xylocaine and the subsequent injection of 1 ml hyaluronan that distended the joint space was called by the authors the ‘pumping injection of hyaluronan’. The control group was not treated, but observed, during the same period of time. This retrospective cohort study confirmed the previous observation that the ‘pumping injection of hyaluronan’ into the TMJ is an effective treatment for nonreducing disc replacement. Recently Bertolami [2001], in a discussion of Alpaslan and Alpaslan’s paper [2001] on the treatment of internal derangement with arthrocentesis with or without viscosupplementation, pointed out that all placebo (physiological saline)-controlled studies published as of today indicate that while arthrocentesis, alone or in combination with lavage using physiological salt solution, is therapeutically effective, viscosupplementation with or without lavage ‘achieved statistical significance as assessed by maximal mouth opening, lateral movement, joint pain, noise and jaw function’. The question is whether the statistical significance represents clinical significance as well. Bertolami points out that 24 months postoperatively there was a 7.97-mm difference in maximal opening for the closelock variant of internal derangement in favor of the hyal-
Balazs
uronan-treated group. However, this difference is smaller when adjustment is made for baseline (pretreatment) values. Bertolami concludes that, after proper adjustment, a 3- to 4-mm statistically significant benefit still exists, and whether or not this difference is clinically significant, only the clinician and patient can determine the clinical importance of such improvement.
Viscosurgery in TMJ
The viscosurgical use of elastoviscous hylan fluids during arthroscopic surgery of the TMJ was reported by McCain et al. [1989]. At this time, elastoviscous hyaluronan solutions were being used worldwide in ophthalmic surgery to protect sensitive tissue surface from mechanical damage (corneal endothelium, iris, retina), to create space during surgical procedures (cataract surgery) and to move and manipulate sensitive tissues during surgery (retinal detachment) [for review, see Arshinoff, 2002]. These authors used 3 ml of hylan A fluid (0.5%) with very high elastic and viscous properties (average MW F6 million) for distention of the capsule and to protect the surfaces of the menisci, cartilage and soft tissues from mechanical damage during diagnostic arthroscopy. In a randomized study, 55 patients with various joint pathologies were assigned to a control group and to the viscosurgical group. Surgical scores to evaluate visualization, control of debris, bleeding, amount of scuffing of the cartilage, tissue debridement and control of synovium were established. According to the authors, the viscosurgical use of hylan A provided control of tissue in the joint space and, most significantly, protection of joint tissues and reduction of iatrogenic damage during surgery. Postoperative pain, mouth opening and range of motion were not affected by the viscosurgical use of this high-MW hyaluronan derivative.
Effect of Viscosupplementation on Synovial Fluid
There are very few reports on the analysis of the hyaluronan concentration in synovial fluids from human knee joints obtained before and after viscosupplementation therapy. Even fewer are reported on the analysis of inflammation-causing agents in human arthritic joints after viscosupplementation. Two such reports on the joint wash of the TMJ diagnosed with DDR and DDN are reviewed here. Synovial fluid washes were obtained by
Analgesic Effect of Hyaluronan
injecting 2 ml of saline solution into the upper joint space (after local anesthesia), aspiration and reinjection of the same solutions 3 times [Alpaslan et al., 2000]. After the cells were removed, the wash was analyzed for total nitrite, nitrate and lipid peroxidation products. These chemicals in the joint fluid of animals with experimental arthritis have been associated with the presence of inflammatory mediators, arachidonic acid and pain-causing molecules, and they have been found in the synovial fluid of the TMJ with various pathologies. Synovial wash was collected before treatment and 15 days after treatment. The levels of total nitrite, nitrate and lipid peroxidation products were significantly reduced 15 days after the injection of 1 ml hyaluronan (Orthovisc) in the TMJ. The authors make the statement, but do not support with data, that ‘decrease in these levels was not significant in patients receiving arthrocentesis only’. It is important to point out that the differences in the change in levels of these chemicals in the synovial fluid did not correlate with the level of improvement of clinical symptoms. Jaw function and pain scores were significantly reduced from pretreatment to 15 days after treatment in both groups; however, there was no difference between the levels of reduction in the two groups [Alpaslan et al., 2000]. The effect of viscosupplementation with Artz on clinical symptoms (pain, noise, mouth opening), on arachinoid acid metabolites (LTC4, 6-keto-PGF-· and PGF2-·) and interleukin-1ß (IL-1ß) in the synovial fluid wash was reported [Hirota, 1998]. The TMJ of 15 patients with painful unilateral derangement without radiological evidence of condylar degeneration was lavaged with 5 ml of physiological saline solution and then 1 ml Artz was injected twice (1 week apart) into the superior compartment of the joint. Lavage was performed also after the completion of the study. After random selection in the lavage fluid of 9 patients the arachinoid acid metabolites and in 6 fluids ILs were determined. The hyaluronan treatment caused a significant decrease of all arachidonic acid metabolites that were detectable before treatment. Similarly, in 6 patients, the total IL-1ß was significantly reduced. Hyaluronan treatment also significantly reduced the mean pain and noise scores and mouth opening measurements in all but 3 patients. The author remarked that ‘these findings suggest that inflammation plays a part in internal derangement of the TMJ and injection of antiinflammatory substance may be beneficial to such patients’. Based on literature data, the author assumes that in this case, hyaluronan had an anti-inflammatory effect as scavenger of free radicals.
Cells Tissues Organs 2003;174:49–62
57
Viscosupplementation of TMJ in Animal Models
Several studies reported the use of hyaluronan injection in sheep TMJ after artificially created wounds. Neo et al. [1997] created intra-articular wounds by perforating the central part of the disc (3.8 mm) and removing the fibrocartilaginous condylar surface over the entire condyle in the TMJ of 6 sheep. Five hyaluronan injections (3 days apart; Artz, 1%, average MW 0.8 million) on one side and the same number of saline injections into the contralateral TMJ were performed. One and 3 months after injection, the joints were examined macroscopically and histologically. Both endpoints showed significant differences between the two treatments. The hyaluronantreated joints showed ‘less deviation from normal condylar morphology and in general for all histological scores there was a significant difference between the two groups’. The authors concluded that hyaluronan treatment ‘minimizes the extent of osteoarthritic changes when compared with control joint’. Bilateral intra-articular wounds were made in 6 sheep TMJ by perforating discs and by separating subchondral condylar surfaces. One day later, hyaluronan (1%, average MW 3.0 million) was injected 5 times at 3-day intervals. Physiological saline solution was injected in the contralateral joint at the same time intervals. One and 3 months after surgery, the joints were investigated radiographically and histopathologically. The results indicated that the treatment with viscosupplementation ‘inhibited the progression of OA in the ovine TMJs by inducing the development of articular cartilage and by reducing the proliferation of fibrous tissues’ [Kim et al., 2001]. These two studies suggest that repeated hyaluronan injections during the healing of sterile intra-articular wounds are beneficial to the healing process.
The Future of Viscosupplementation in TMJ
It has been more than three decades since the highly purified elastoviscous sodium hyaluronan was first tested for analgesic effect in the joints of osteoarthritic patients and racehorses, and two decades have passed since similar clinical tests were performed in human TMJ using the same preparations. Today, viscosupplementation as a treatment of osteoarthritis of the knee is used worldwide, and millions of patients receive this treatment with less frequent and less severe adverse events than any other known chronic drug or surgical treatment of this disease.
58
Cells Tissues Organs 2003;174:49–62
Two types of hyaluronan preparations are available worldwide for viscosupplementation of the knee and, in some countries, for other joints: preparations with low rheological properties (elastoviscosity) similar to those of the synovial fluid in arthritic joints and those with high rheological properties similar to those of the healthy synovial fluid. Despite the fact that 15 clinical studies using four different hyaluronan preparations have been reported, hyaluronan is not available for treatment of the TMJ. The explanation must be found in the fact that the elastoviscosity of the hyaluronan preparations used and the design of the clinical protocols were so different that comparisons of treatment outcomes were difficult. The acceptance of viscosupplementation as a new therapeutic modality was considerably slowed down by the confusion caused by these variables. The rheological differences in hyaluronan preparations, caused primarily by the differences in average MW and the polydispersity of the molecules, are demonstrated in figure 4, where the elastic and viscous properties of most of the hyaluronan products used in TMJ are compared. Hyaluronan molecules in nature and in purified preparations are always present in a polydisperse population. This means that the MW expresses only the statistical average mass of the molecules present in solution and does not give any information about the distribution of the molecules with respect to their individual masses (MW). Figure 5 demonstrates the polydispersity of the molecules with respect to their molecular masses in healthy and pathological human knee synovial fluids. Figure 6 shows the polydispersity of the hyaluronan preparations used in clinical trials for viscosupplementation of TMJ. By comparing the two figures, it is obvious that only two hyaluronan preparations used in these clinical studies have a molecular distribution similar to hyaluronan in healthy synovial fluid. The concept of viscosupplementation was developed on the theoretical consideration that the exchange of the synovial fluid of arthritic joints with hyaluronan solutions that resemble the fluid in healthy joints in MW distribution and, therefore, elastoviscous properties, will have an analgesic effect and promote the healing of intra-articular wounds. The clinical significance of this hypothesis became obvious at the time of the first human clinical studies. First of all it was observed that the exchange of the pathological synovial fluid with physiological salt solutions (as control) results in a short-term analgesic effect. Repeated exchange or extensive exchange, such as lavage in diagnostic arthroscopy, improved the analgesic effect. Thus, viscosupplementation presented two modes of ac-
Balazs
Fig. 4. Proportion of elasticity and viscosity
in hyaluronan preparations used in TMJ studies as a function of the frequency at which the measurement of the elastic (G)) and viscous (G))) dynamic moduli is made. The percent of elasticity was calculated as follows: % elasticity =
G) !100 G) + G))
[see for details Bothner-Wik and Wik, 1998].
0.12
Optical density
Relative mobility
0.10
Normal
Hylan A Hylartil
0.08
Orthovisc
0.06
Artzal
0.04
Osteoarthritic 0.02
6
4
3
2 1 0.6 Molecular mass × 10 –6
0.4 0.3
0 1
×
10 7
1
×
10 6
1
×
10 5
Molecular mass
Fig. 5. The polydispersity of hyaluronan in the synovial fluid of the knee of a healthy and an osteoarthritic patient demonstrated in agarose gel electrophoresis according to Lee and Cowman [1994]. The relative mobility was obtained from densitometry measurements and MW distribution is based on electrophoretic analysis of hyaluronan standards with known average molecular weight. The variation in polydispersity of hyaluronan from healthy knee fluids is not substantial and contains only a very small fraction of molecules with ! 2 million molecular mass. In contrast, the polydispersity profile of hyaluronan in the osteoarthritic fluid is greatly variable; however, it always contains a small, but variable amount of high molecular mass (1 2 million) fractions.
Fig. 6. Polydispersity of hyaluronan products used for viscosupplementation of the TMJ are demonstrated by agarose gel electrophoresis (see fig. 5 for details). W W W W W W = Hyaluronan in the synovial fluid from a healthy knee. Hylan A is the fluid component of Synvisc representing 80% (per volume) of the preparation. The average molecular weights of these products are as follows: hylan A 6 million, Hylartil 2.5 million, Orthovisc 1.2 million, and Artz® or Artzal® 0.7 million.
Analgesic Effect of Hyaluronan
Cells Tissues Organs 2003;174:49–62
59
tion to explain therapeutic efficacy. One is based on the exchange of the pathological synovial fluid with a physiological salt solution (the solvent of hyaluronan) that dilutes or washes out the pain- and inflammation-causing agents in the painful joint by repeated injections. This dilution or washing-out effect can be more effective using low-viscosity hyaluronan solutions instead of physiological salt solutions. This is because hyaluronan of any molecular size can have a detoxicating effect by absorbing or simply interacting with and thereby inactivating painand inflammation-causing agents in the joint. The second mode of action is related to the elastoviscosity of the hyaluronan solutions. The analgesic studies both in animals and in vitro systems clearly demonstrate that the higher the elastoviscosity, the greater the analgesic effect. From the time of the first clinical trials with hyaluronan solutions of low elastoviscosity, it was obvious that, in order to obtain clinical efficacy, one must repeat the weekly injections 5–10 times. On the other hand, using hyaluronan preparations of greater elastoviscosity, clinical efficacy compared to physiological salt solutions could be obtained after two or three applications [Scale et al., 1994; Wobig et al., 1999]. This underlines the significance of the diluting, wash-out and detoxicating effects of viscosupplementation as well as the critical importance of the rheological properties of the solution [Peyron, 1993; Adams, 1998]. One must ask why various pharmaceutical companies in the late 1980s introduced hyaluronan preparations with very low elastoviscosity when several nonproprietary technologies were available to prepare hyaluronan of much greater average MW from two tissue sources (rooster comb, umbilical cord). It can be that they overlooked or disregarded the conceptual basis of viscosupplementation. On the other hand, it was well known from 1971 that the technology to produce high MW, highly purified, sterile, pyrogen-free hyaluronan is demanding and expensive. Whatever the reason, between 1987 and 1994, only very low average MW (!1 million), very polydisperse hyaluronan preparations were available for viscosupplementation. In other words, viscosupplementation was introduced into medical practice with hyaluronan products injected 5–10 times, 1 week apart, that could have been designed for washing out or lavage of the joint. These hyaluronan solutions were less elastoviscous than the pathological exudate present in the joint and, therefore, did not resemble the healthy synovial fluid in any aspect. In reality, the concept of viscosupplementation was disregarded and replaced with a procedure that can be properly characterized as a wash-out of the joint with a hyaluronan
60
Cells Tissues Organs 2003;174:49–62
solution with very low rheological properties, and with similar or lower elastoviscosity than that of the pathological exudate itself. In the mid-1990s, for the first time a hyaluronan solution with a similar or even better molecular profile (average MW, polydispersity and rheology) than that of the healthy synovial fluid was introduced as a real viscosupplementation product [Adams, 1993; Balazs and Denlinger, 1993; Adams et al., 1995]. The essential differences between the properties of this new and the old hyaluronan products and their recommended treatment schedules were often disregarded. The concept of viscosupplementation and its use, and also the interpretation of clinical results have been and still are often confusing to the treating physician. It was emphasized by the inventors of this new therapeutic modality that before the viscosupplement is injected into the joints, the pathological fluid, especially if it is present in large volume (effusion), must be removed. This procedure presents extra work and cost and therefore was not always followed. Many clinical trials were performed without the application of these essential procedures and therefore the potential for efficacy was decreased. Viscosupplementation is not a drug treatment like corticosteroid injection, and therefore, the elastoviscous solution must be injected into the joint space that has no or very little pathological fluid in it. A number of studies demonstrated that this essential step is often missed. In an important number of cases, the partial or completely extra-articular injection resulted in lack of efficacy. It is no surprise to me that the great variation in the rheological quality of hyaluronan preparations available today and the diverse treatment recommendations by the marketing companies, such as those related to the removal of exudate, confused not only the physician who had to choose between the various preparations, but also the clinical scientists who critiqued this relatively new and often misunderstood therapeutic application. One hopes that the pioneers of viscosupplementation for the TMJ will benefit from their own early experience and those obtained during decades of clinical trials on the knee joint. In this regard, the fact that various types of lavage were frequently used in TMJ studies prior to application of the viscosupplement is extremely encouraging for the future.
Balazs
References Adams, M.E. (1993) An analysis of clinical studies of the use of crosslinked hyaluronan, hylan, in the treatment of osteoarthritis. J Rheumatol 20(suppl 39): 16–18. Adams, M.E. (1998) Viscosupplementation as articular therapy; in Laurent, T. (ed): The Chemistry, Biology and Medical Applications of Hyaluronan and Its Derivatives (Proceedings of the Wenner-Gren Foundation International Symposium held in honor of Endre A. Balazs, September 18–21, 1996, Stockholm, Sweden). London, Portland Press, pp 243–253. Adams, M.E., M.H. Atkinson, A. Lussier, J.I. Schulz, K.A. Siminovitch, J.P. Wade, M. Zummer (1995) The role of viscosupplementation with hylan G-F 20 (Synvisc®) in the treatment of osteoarthritis of the knee: A Canadian multicenter trial comparing hylan G-F 20 alone, hylan G-F 20 with non-steroidal anti-inflammatory drugs (NSAIDs) and NSAIDs alone. Osteoarthritis Cartilage 3: 213–226. Aihara, S., N. Murakami, R. Ishii, K. Kariya, Y. Azuma, K. Hamada, J. Umemoto, S. Maeda (1992) Effects of sodium hyaluronate on the nociceptive response of rats with experimentally induced arthritis. Folia Pharmacol Jpn 100: 359–365. Alpaslan, G.H., C. Alpaslan (2001) Efficacy of temporomandibular joint arthrocentesis with and without injection of sodium hyaluronate in treatment of internal derangements. J Oral Maxillofac Surg 59: 613–618. Alpaslan, C., A. Bilgihan, G.H. Alpaslan, B. Güner, Ö. Yis, D. Erbas¸ (2000) Effect of arthrocentesis and sodium hyaluronate injection on nitrate, nitrate, and thiobarbituric acid-reactive substance levels in the synovial fluids. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 89: 686–690. Arshinoff, S.A. (2002) The use of ophthalmic viscosurgical devices in cataract surgery; in Kennedy, J.F., G.O. Phillips (eds): Hyaluronan 2000. Cambridge, Woodhead Publishing, vol 2, pp 119–128. Balazs, E.A. (1971) Hyaluronic Acid and Matrix Implantation. Arlington, Biotrics. Balazs, E.A. (1986) The development of sodium hyaluronate (Healon) as a viscosurgical material in ophthalmic surgery; in Eisner, G. (ed): Ophthalmic Viscosurgery – A Review of Standards, Techniques and Applications. Bern, Medicopea, pp 3–19. Balazs, E.A., S. Al-Assaf, G.O. Phillips (1999) A new biomaterial tissue allograft: Hylan, a hyaluronan derivative; in Phillips, G.O., D.M. Strong, R. von Versen, A. Nather (eds): Advances in Tissue Banking. New York, World Scientific Publishing, vol 3, pp 355–397. Balazs, E.A., J.L. Denlinger (1985) Sodium hyaluronate and joint function. J Equine Vet Sci 5: 217–219, 222–228. Balazs, E.A., J.L. Denlinger (1993) Viscosupplementation: A new concept in the treatment of osteoarthritis. J Rheumatol 20(suppl 39): 3–9.
Analgesic Effect of Hyaluronan
Balazs, E.A., E.A. Leshchiner (1989) Hyaluronan, its crosslinked derivative – hylan – and their medical applications; in Inagaki, H., G.O. Phillips (eds): Cellulosics Utilization: Research and Rewards in Cellulosics. Proceedings of Nisshinbo International Conference on Cellulosics Utilization in the Near Future. New York, Elsevier Applied Science, pp 233–241. Belmonte, C., M.A. Pozo, E.A. Balazs (1998) Modulation by hyaluronan and its derivatives (hylans) of sensory nerve activity signaling articular pain; in Laurent T. (ed): The Chemistry, Biology and Medical Applications of Hyaluronan and Its Derivatives (Proceedings of the Wenner-Gren Foundation International Symposium held in honor of Endre A. Balazs, September 18–21, 1996, Stockholm, Sweden). London, Portland Press, pp 205–217. Bertolami, C.N. (2001) Efficacy of temporomandibular joint arthrocentesis with and without injection of sodium hyaluronate in treatment of internal derangements. J Oral Maxillofac Surg 59: 618–620. Bertolami, C.N., T. Gay, G.T. Clark, J. Rendell, V. Shetty, C. Liu, D.A. Swann (1993) Use of sodium hyaluronate in treating temporomandibular joint disorders: A randomized, doubleblind, placebo-controlled clinical trial. J Oral Maxillofac Surg 51: 232–242. Bothner-Wik, H., O. Wik (1998) Rheology of hyaluronan; in Laurent, T. (ed): The Chemistry, Biology and Medical Applications of Hyaluronan and Its Derivatives (Proceedings of the Wenner-Gren Foundation International Symposium held in honor of Endre A. Balazs, September 18–21, 1996, Stockholm, Sweden). London, Portland Press, pp 25–32. Day, A.J., J.K. Sheehan (2001) Hyaluronan: Polysaccharide chaos to protein organisation. Curr Opin Struct Biol 11: 617–622. de la Peña, E., S. Sala, J.C. Rovira, R.F. Schmidt, C. Belmonte (2002a) Elastoviscous substances with analgesic effects on joint pain reduce stretch-activated ion channel activity in vitro. Pain 99: 501–508. de la Peña, E., S. Sala, R.F. Schmidt, C. Belmonte (2002b) Effects of elastoviscous hylans on mechanotransduction; in Kennedy, J.F., G.O. Phillips (eds): Hyaluronan 2000. Cambridge, Woodhead Publishing, vol 2, pp 407–418. Denlinger, J.L. (1998) Hyaluronan and its derivatives as viscoelastics in medicine; in Laurent, T. (ed): The Chemistry, Biology and Medical Applications of Hyaluronan and Its Derivatives (Proceedings of the Wenner-Gren Foundation International Symposium held in honor of Endre A. Balazs, September 18–21, 1996, Stockholm, Sweden). London, Portland Press, pp 235–242. Denlinger, J.L., E.A. Balazs (1980) Replacement of the liquid vitreus with sodium hyaluronate in monkeys. I. Short-term evaluation. Exp Eye Res 31: 81–99. Edwards, J.L. (1994) Sodium hyaluronate for temporomandibular joint disorders. Ann R Aust Coll Dent Surg 12: 217–222.
Elfvin, L.-G., K. Holmberg, J. Johansson, H. Aldskogius (1998) The innervation of the synovium of the knee joint in the guinea pig: An immunohistochemical and ultrastructural study. Anat Embryol 197: 293–303. Fader, K.W., D.C. Grummons, R. Maijer, L.V. Christensen (1993) Pressurized infusion of sodium hyaluronate for closed lock of the temporomandibular joint. I. A case study. J Craniomandib Practice 11: 68–72. Favia, G., E. Maiorano (1995) Presence of neuroreceptors in normal and diseased temporomandibular joints. Boll Soc Ital Biol Sper 71: 201– 212. Gomis, A., M. Pawlak, E.A. Balazs, R.F. Schmidt, C. Belmonte (2003) Effects of different molecular weight viscoelastic hyaluronan solutions on articular nociceptive afferents. Arthritis Rheum (in press). Gotoh, S., K. Miyazaki, J. Onaya, T. Sakamoto, K. Tokuyasu, O. Namiki (1988) Experimental knee pain model in rats and analgesic effect of sodium hyaluronate (SPH). Folia Pharmacol Jpn 92: 17–27. Grigg, P. (2001) Properties of sensory neurons innervating synovial joints. Cells Tissues Organs 169: 218–225. Hatton, M.N., D.A. Swann (1986) Studies on bovine temporomandibular joint synovial fluid. J Prosthet Dent 56: 635–638. Hepguler, S., Y.S. Akkoc, M. Pehlivan, C. Ozturk, G. Celebi, A. Saracoglu, B. Ozpinar (2002) The efficacy of intra-articular sodium hyaluronate in patients with reducing displaced disc of the temporomandibular joint. J Oral Rehabil 29/1: 80–86. Heppelmann, B., K. Messlinger, W.F. Neiss, R.F. Schmidt (1990) Ultrastructural three-dimensional reconstruction of group III and group IV sensory nerve endings (‘Free nerve endings’) in the knee joint capsule of the cat: Evidence for multiple receptive sites. J Comp Neurol 292: 103–116. Hirota, W. (1998) Intra-articular injection of hyaluronic acid reduces total amounts of leukotriene C4, 6-keto-prostaglandin F1alpha, prostaglandin F2alpha and interleukin-1beta in synovial fluid of patients with internal derangement in disorders of the temporomandibular joint. Br J Oral Maxillofac Surg 36: 35–38. Ishimaru, J.-I., N. Ogi, S. Mizuno, A.N. Goss (2001) Quantitation of chondroitin-sulfates, disaccharides and hyaluronan in normal, early and advanced osteoarthritic sheep temporomandibular joints. Osteoarthritis Cartilage 9: 365–370. Kim, C.H., B.J. Lee, J. Yoon, K.M. Seo, J.H. Parks, J.W. Lee, E.S. Cho, J.J. Hong, Y.S. Lee, J.H. Park (2001) Therapeutic effect of hyaluronic acid on experimental osteoarthrosis of ovine temporomandibular joint. J Vet Med Sci 63: 1083–1089. Kopp, S., S. Åkerman, M. Nilner (1991) Short-term effects of intra-articular sodium hyaluronate, glucocorticoid and saline injections on rheumatoid arthritis of the temporomandibular joint. J Craniomandib Facial Oral Pain 5: 231–238.
Cells Tissues Organs 2003;174:49–62
61
Kopp, S., G.E. Carlsson, T. Haraldson, B. Wenneberg (1987) Long-term effect of intra-articular injections of sodium hyaluronate and corticosteroid on temporomandibular joint arthritis. J Oral Maxillofac Surg 45: 929–935. Kopp, S., B. Wenneberg (1979) Injection av healosid è käkled. En preliminary rapport. Rep Ser No 22, August 1979. Göteborg, Department of Stomatognathic Physiology, Göteborg University. Kopp, S., B. Wenneberg, T. Haraldson, G.E. Carlsson (1985) The short-term effect of intra-articular injections of sodium hyaluronate and corticosteroid on temporomandibular joint pain and dysfunction. J Oral Maxillofac Surg 43: 429–435. Lee, H.G., M.K. Cowman (1994) An agarose gel electrophoretic method for analysis of hyaluronan molecular weight distribution. Anal Biochem 219: 278–287. McCain, J.P., E.A. Balazs, H. de la Rua (1989) Preliminary studies on the use of a viscoelastic solution in arthroscopic surgery of the temporomandibular joint. J Oral Maxillofac Surg 47: 1161–1168. Miyazaki, K., S. Gotoh, H. Ohkawara, T. Yamaguchi (1984) Studies on analgesic and anti-inflammatory effects of sodium hyaluronate (SPH). Pharmacometrics 28: 1123–1135. Murakami, T., H. Higaki, Y. Sawae, N. Ohtsuki, S. Moriyama, Y. Nakanishi (1998) Adaptive multimode lubrication in natural synovial joints and artificial joints. Proc Inst Mech Eng 212: 23–35. Neo, H., J.-I. Ishimaru, K. Kurita, A.N. Goss (1997) The effect of hyaluronic acid on experimental temporomandibular joint osteoarthrosis in the sheep. J Oral Maxillofac Surg 55: 1114–1119.
62
Pawlak, M., A. Gomis, S. Just, B. Heppelmann, C. Belmonte, R.F. Schmidt (2002) Mechanoprotective actions of elastoviscous hylans on articular pain receptors; in Kennedy, J.F., G.O. Phillips (eds): Hyaluronan 2000. Cambridge, Woodhead Publishing, vol 2, pp 341–351. Peyron, J.G. (1993) Intra-articular hyaluronan injections in the treatment of osteoarthritis: State-of-the-art review. J Rheumatol 20(suppl 39): 10–15. Pozo, M.A., E.A. Balazs, C. Belmonte (1997) Reduction of sensory responses to passive movements of inflamed knee joints by hylan, a hyaluronan derivative. Exp Brain Res 116: 3– 9. Rydell, N., E.A. Balazs (1971) Effect of intra-articular injection of hyaluronic acid on the clinical symptoms of osteoarthritis and on granulation tissues formation. Clin Orthop 80: 25–32. Rydell, N.W., J. Butler, E.A. Balazs (1970) Hyaluronic acid in synovial fluid. VI. Effect of intra-articular injection of hyaluronic acid on the clinical symptoms of arthritis in track horses. Acta Vet Scand 11: 139–155. Sato, S., S. Goto, T. Kasahara, H. Kawamura, K. Motegi (2001a) Effect of pumping with injection of sodium hyaluronate and the other factors related to outcome in patients with nonreducing disk displacement of the temporomandibular joint. Int J Oral Maxillofac Surg 20: 194–198. Sato, S., S. Oguri, K. Yamaguchi, H. Kawamura, K. Motegi (2001b) Pumping injection of sodium hyaluronate for patients with non-reducing disc displacement of the temporomandibular joint: Two-year follow-up. J Craniomaxillofac Surg 29: 89–93. Sato, S., M. Ohta, H. Ohki, H. Kawamura, K. Motegi (1997) Effect of lavage with injection of sodium hyaluronate for patients with nonreducing disk displacement of the temporomandibular joint. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 84: 241–244.
Cells Tissues Organs 2003;174:49–62
Sato, S., M. Sakamoro, H. Kawamura, K. Motegi (1999) Disc position and morphology in patients with nonreducing disc displacement treated by injection of sodium hyaluronate. Int J Oral Maxillofac Surg 28: 253–257. Scale, D., M. Wobig, W. Wolpert (1994) Viscosupplementation of osteoarthritic knees with hylan: A treatment schedule study. Curr Ther Res 55/3: 220–232. Shibata, T., K.-I. Murakami, E. Kubota, H. Maeda (1998) Glycosaminoglycan components in temporomandibular joint synovial fluid as markers of joint pathology. J Oral Maxillofac Surg 56: 209–213. Smith, T.J., R.S. Bahn, C.A. Gorman (1989) Hormonal regulation of hyaluronate synthesis in cultured human fibroblasts: Evidence for differences between retroocular and dermal fibroblasts. J Clin Endocrinol Metab 69: 1019– 1023. Tahmasebi-Sarvestani, A., R.A. Tedman, A. Gross (1996) Neural structures within the sheep temporomandibular joint. J Orofac Pain 10: 217– 231. Wobig, M., G. Bach, P. Beks, A. Dickhut, J. Runzheimer, G. Schwieger, G. Vetter, E.A. Balazs (1999) The role of elastoviscosity in the efficacy of viscosupplementation for osteoarthritis of the knee: A comparison of hylan G-F 20 and a lower-molecular-weight hyaluronan. Clin Ther 21: 1549–1562. Yustin, D., B. Kryshtalskyj, A. Galea (1995) Use of Hylan G-F 20 for viscosupplementation of the temporomandibular joint for the management of osteoarthritis: A case report. J Orofac Pain 9: 375–379. Zhiyuan, G., W. Qiulian, Z. Yingxi, Z. Zhenkang, S. Kaihua (1998) Viscosupplementation therapy in internal derangement of temporomandibular joint. Chin Med J 111: 656–659.
Balazs
Cells Tissues Organs 2003;174:63–72 DOI: 10.1159/000070575
Induction of Early Growth Response Gene Egr2 by Basic Calcium Phosphate Crystals through a Calcium-Dependent Protein Kinase CIndependent p44/42 Mitogen-Activated Protein Kinase Pathway Xiao R. Zeng a Yubo Sun a Leonor Wenger a Herman S. Cheung b, c a Department of Medicine, University of Miami School of Medicine and b Research Service and Geriatric Research, Education and Clinical Center, Veterans Administration Medical Center, Miami, Fla., and c Department of Biomedical Engineering, University of Miami, Coral Gables, Fla., USA
Key Words Basic calcium phosphate crystals W Early growth response gene W Transcriptional factors W Gene expression
Abstract Using the reverse transcriptase-polymerase chain reaction we examined the effect of basic calcium phosphate (BCP) crystals on the induction of the early growth response gene Egr2 transcription and the signal transduction pathway involved. The results showed that BCP
crystals induced Egr2 transcription up to 8-fold, peaking at 24 h after treatment. The induction of Egr2 was confirmed by transient transfection assays using an Egr2 promoter/luciferase reporter construct and could be inhibited by the p44/42 mitogen-activated protein kinase (MAPK)-specific inhibitor U0126, or by calcium chelator TMB-8, but not by the SAPK2/p38 MAPK inhibitor SB202190 or by the protein kinase C inhibitor bisindolylmaleimide I (Bis-I). Using the Mercury Pathway Profiling System (Clontech, Palo Alto, Calif., USA) we further showed that induced Egr2 could stimulate the activities of several transcription factors that are associated with cell proliferation, such as c-fos, SRF and c-myc. Copyright © 2003 S. Karger AG, Basel
Abbreviations used in this paper
Introduction BCP Bis-I CPPD EGF MAPK PC PKC RT-PCR SEAP
basic calcium phosphate bisindolylmaleimide I calcium pyrophosphate dihydrate epidermal growth factor mitogen-activated protein kinases phosphocitrate protein kinase C reverse transcriptase-polymerase chain reation secreted alkaline phosphatase
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0063$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
Crystalline calcium pyrophosphate dihydrate (CPPD) and basic calcium phosphate (BCP) are the two most common forms of pathologic articular mineral. Each occurs frequently in osteoarthritic joints, and each may be phlogistic, causing acute attacks of pseudogout in the case of CPPD crystals and acute calcific periarthritis in the case of BCP crystals [Halverson and McCarty, 1997; Ryan and
Herman S. Cheung, PhD Department of Medicine University of Miami School of Medicine Miami, FL 33101 (USA) Tel. +1 305 324 4455, ext. 3646, Fax +1 305 324 3365, E-Mail
[email protected]
McCarty, 1997]. There is compelling in vitro evidence that these crystals engender multiple biological effects that promote joint degeneration, and clinical observations support a relationship of both crystal types to osteoarthritis. Synovial hyperplasia is commonly observed in the joints among patients with calcium-containing crystal deposition [Schumacher, 1968]. It has been reported that calcium-containing crystals, such as CPPD and BCP crystals, can induce macrophages/monocytes and synovial fibroblasts around the joints to produce a number of mitogenic cytokines which can promote synovial lining proliferation [Feldman et al., 1990]. Previously, we have shown that BCP crystals, by themselves, can stimulate mitogenesis of cultured human skin fibroblasts and canine synovial fibroblasts in a concentration-dependent fashion [Cheung et al., 1984]. Addition of BCP crystals to the growth medium yielded an immediate 10-fold rise in the intracellular calcium level mainly through extracellular Ca2+ influx [Halverson et al., 1998]. This early rise in Ca2+ triggered a fast membrane-associated event that, acting in conjunction with phospholipase C and protein kinase C (PKC) activation, led to induction of the early response genes c-fos and c-myc [Mitchell et al., 1989] which may play important roles in promoting cell proliferation. In addition, BCP crystals induced a second rise in intracellular calcium levels starting at 60 min after stimulation and continuing to increase for up to 3 h. This second increase in Ca2+ is believed to be derived from intracellular dissolution of phagocytosed crystals and leads to a variety of late calcium-dependent events resulting in further cell mitogenesis [Halverson et al., 1998]. So far, the molecular events involved in the second slow intracellular calcium rise have not been well elucidated. Early growth response (Egr) genes were identified originally based on their rapid induction of gene expression in quiescent fibroblasts stimulated by serum [Lemaire et al., 1988]. The amino acid sequences of four Egr gene family members are distinct, except that they all contain three highly conserved Cyc2/His2 types of ‘zinc finger’ regions, which allow them to interact with the same Sp1-type of DNA target element GCGGGGGCG [Crosby et al., 1991]. Egr proteins serve as sensors of extracellular signaling pathways that play key roles in regulating cell proliferation, differentiation and function. Egr gene expression can be rapidly and transiently induced by a variety of stimuli such as growth factors [Lemaire et al., 1988], protein kinase A and PKC [Mechta et al., 1989], electroconvulsive shock [Bhat et al., 1992] and fine spherical or fibrous particles [Hirano et al., 2000] in various cell types.
64
Cells Tissues Organs 2003;174:63–72
Because Egr proteins are transcription factors, early activation of Egr gene expression will trigger a cascade of late gene expression. Previously, we have shown that BCP crystals can stimulate cell proliferation in human fibroblasts [Cheung et al., 1984]. We are particularly interested in the effect of BCP crystals on the expression of genes involved in regulating cell proliferation. Therefore, we undertook the present study to examine the effect of BCP crystals on the expression of the Egr gene family. Among the four Egr gene family members, we found that the mRNA of Egr4 was undetectable in human fibroblasts, while the other three family members were detectable and inducible by BCP crystals. Egr1 and Egr3 transcripts responded to BCP crystals in a similar fashion as growth factors such as epidermal growth factor (EGF), which was rapid and transient [unpubl. data]. Interestingly, the induction of Egr2 transcripts by BCP crystals increased gradually and persistently up to 24 h. The induction pattern of Egr2 transcripts suggests that it may relate to the second progressively increased intracellular calcium levels of human fibroblasts treated with BCP crystals we reported previously [Halverson et al., 1998]. Therefore, in the present study we focused our investigation on the induction of Egr2 transcription by BCP crystals, and the signal transduction pathways involved. The results presented here may provide important insights into the mechanism linking early response gene expression and BCP crystalinduced mitogenesis.
Materials and Methods Cell Cultures of Human BJ1 Cells and Rabbit Synovial Fibroblasts Human fibroblast cell line BJ1 was purchased from Clontech (Palo Alto, Calif., USA). The cells were routinely cultured in DMEM supplemented with 10% FBS, penicillin and streptomycin in a 5% CO2 37 ° C incubator. Rabbit synovial fibroblasts were isolated by enzymatic dispersion of rabbit synovial tissues. Briefly, the tissues were minced and incubated with 0.1% hyaluronidase for 10 min at 37 ° C, washed with Hanks’ medium, and digested with 0.5% of trypsin, 0.2% collagenase and 0.05% collagenase for 10 min, 45 min and 3 h, respectively. The cells were cultured in DMEM supplemented with 10% FBS, penicillin and streptomycin in a 5% CO2 37 ° C incubator overnight. Nonadherent cells were removed, and adherent cells were cultivated in fresh DMEM with 10% FBS. The cells used for transfection experiments were third to twelfth passage cells. BCP Crystals and Phosphocitrate Preparation BCP crystals were prepared by a modification of previously published methods [Evans et al., 1984]. These crystals have a calcium/ phosphate ratio of 1.59 and contain partially carbonate-substituted hydroxyapatite mixed with octacalcium phosphate as determined by
Zeng/Sun/Wenger/Cheung
Fourier transform infrared spectroscopy. The crystals were crushed and sieved to yield 10- to 10-Ìm aggregates, which were sterilized and rendered pyrogen-free by heating at 200 ° C for at least 90 min. Phosphocitrate (PC) was prepared as the sodium salt by phosphorylation of tribenyl citrate followed by deprotection through hydrogenation and crystallization from water [Pankowaki et al., 1994]. Quantification of Egr2 mRNA Levels in Human BJ1 Cells by Reverse Transcriptase-Polymerase Chain Reaction Method Total RNA was isolated using the Trizol reagent following the manufacturer’s instructions (Life Technologies, Rockville, Md., USA). One microgram of each sample was reversed-transcribed at 50 ° C for 40 min, followed by enzyme inactivation at 85 ° C for 5 min. The resulting cDNAs were amplified by polymerase chain reaction (PCR) using Egr2 mRNA-specific primers (table 1). The PCR amplifications were carried out for 28 cycles by denaturing at 95 ° C for 30 s, annealing at 55 ° C for 30 s and extending at 72 ° C for 45 s. As an internal control, a 353-base-pair-long fragment of the constitutively expressed housekeeping gene, ß-actin, was also amplified using the same PCR settings. The PCR products were analyzed by electrophoresis on 2% agarose gel-containing ethidium bromide. Finally, the relative Egr2 mRNA levels were quantified by a 2dimensional spot densitometry tool of AlphaEaseTM software (Alpha Innotech, San Leandro, Calif., USA). Generation of a Luciferase Reporter Expression Vector Driven by Egr2 Promoter A fragment of the Egr2 gene promoter sequence was amplified by a two-step nested PCR method. The PCR primers were designed based on published sequences [Rangnekar et al., 1990]. The first-step PCR was performed using the Egr2 first pair of primers (table 1) for an initial 6 cycles of amplification by denaturing at 95 ° C for 30 s, annealing at 68 ° C for 30 s for the first cycle, decreased 1 ° C each cycle, and extending at 72 ° C for 2 min and 30 s, followed by another 10 cycles of amplifications which include denaturing at 95 ° C for 30 s, annealing at 62 ° C for 30 s and extending at 72 ° C for 2 min and 30 s. Four microliters of the first-step PCR product was used as a template for the second-step PCR. The second-step PCR was performed using the second pair of primers (table 1) for an initial 6 cycles of amplification by denaturing at 95 ° C for 30 s, annealing at 68 ° C for 30 s for the first cycle, decreased 1 ° C every cycle, and extending at 72 ° C for 1 min, followed by another 30 cycles of amplification which include denaturing at 95 ° C for 30 s, annealing at 68 ° C for 30 s and extending at 72 ° C for 1 min. The final PCR product and pGL3-Basic vector DNA (Promega, Madison, Wisc., USA) were both digested with restriction enzymes NheI and SacI at 37 ° C overnight. After gel purification, two pieces of DNA were ligated together by T4 DNA ligase and transformed into DH5·-competent cells. Positive clones were identified by mini-prep screening and confirmed by DNA sequencing analysis. Construction of Egr2 Expression Vector pcDNA3.1-Egr2 A 1,452-base-pair-long DNA fragment which contains the full length of the Egr2 coding sequence was amplified by the reverse transcriptase (RT)-PCR method using a pair of primers (table 1) based on the Egr2 mRNA sequence from Gene Bank database accession No. XM-005883. One microgram of total RNA isolated from human BJ1 cells using the Trizol Reagent (Life Technologies) was used for cDNA synthesis using the ThermoScript RT-PCR System (Life Technologies). Two microliters of the resulting cDNA product
BCP Crystal Activation of Egr2 Gene Is PKC-Dependent
was used as a template for 35 cycles of PCR amplification by denaturing at 95 ° C for 30 s, annealing at 59 ° C for 30 s and extending at 72 ° C for 1 min and 30 s. After purification by QIAquick PCR purification column, the PCR product was directly subcloned into pGEMT Easy vector (Promega) following the manufacturer’s instructions and positive clones were identified by mini-prep screening. The 1,452-base-pair-long fragment was then cut out of the vector by EcoRI digestion, and ligated with the pcDNA3.1 vector pretreated with the same restriction enzyme. Plasmid DNA Transient Transfection Plasmid DNA transient transfections were performed using the LipofectAMINE reagent (Life Technologies) following the manufacturer’s instructions. Exponentially growing cells were plated at a density of 5 ! 105/well in six-well cluster plates in 2 ml of DMEM with 10% FBS and grown until 100% confluent. The cells were then washed and placed in 1 ml of fresh DMEM medium in the presence of 2 Ìg of reporter plasmid and 5 Ìl of LipofectAMINE reagent/well and incubated at 37 ° C for 18 h. Luciferase/SEAP Assay For luciferase assays, Promega’s Luciferase Assay Systems was used. After transient transfection of 2 Ìg of reporter plasmid using the LipofectAMINE method, the cells were placed in fresh DMEM medium containing stimuli/inhibitors for an additional 24 h before being harvested by lysis in reporter lysis buffer (Promega). The luciferase activities were measured on an EG&G Berthold Autolumat LB953 Rack Luminometer. For secreted alkaline phosphatase (SEAP) assays, 1 Ìg of Egr2 expression vector, pcDNA3.1-Egr2, and 1 Ìg of each cis-acting SEAP reporter vector of the Mercury Pathway Profiling System (Clontech) were cotransfected into rabbit synovial fibroblasts using the LipofectAMINE method. After 18 h, the cells were placed in fresh DMEM medium with 10% FBS for an additional 24 h before performing SEAP assay using Great EscAPe SEAP Detection Kit reagent (Clontech). The SEAP activities were measured on an EG&G Berthold Autolumat LB953 Rack Luminometer.
Results
Induction of Egr2 mRNA by BCP Crystals and EGF A 24-hour time course of BCP crystal- and EGFinduced Egr2 transcript levels was performed using an RT-PCR method (fig. 1). Figure 1b shows that the induction of Egr2 transcripts following EGF treatment peaked early at 1 h after stimulation, then gradually decreased, and finally rose to a small peak at 24 h after stimulation. In contrast, the response of Egr2 transcripts to BCP crystals was initially low, increased progressively, and peaked at 24 h after treatment (fig. 1a). The induction of Egr2 transcripts isolated 24 h after treatment with BCP crystals was BCP crystal concentration-dependent (fig. 2). At a concentration of 50 Ìg/ml, BCP crystal-induced Egr2 mRNA levels rached a plateau. PC (1 mM), a specific inhibitor of BCP crystal-mediated biological effects,
Cells Tissues Organs 2003;174:63–72
65
Fig. 1. Time course of Egr-2-relative mRNA levels in response to BCP crystals and EGF in human BJ1 cells. Confluent BJ1 cells in 100-mm2 dishes were starved with 6 ml of DMEM medium for 24 h followed by addition of BCP crystals (50 ÌM/ml) or EGF (20 ng/ml). Total RNAs were isolated at time points as indicated. RT-PCR was performed as described in Methods and the PCR products, induced by BCP crystals (a) and by EGF (b), were visualized on 2% agarose gel stained with ethidium bromide. c The relative intensities of bands were scanned and the relative Egr2 mRNA levels were quantified by a 2-dimensional spot densitometry tool of AlphaEaseTM software (Alpha Innotech), and plotted using ß-actin bands as internal control for normalization. The results are expressed as means B SEM.
Table 1. PCR primers designed for Egr2
mRNA quantification and for Egr2 promoter and coding region cloning
Primer name
Nucleotide sequence of the primers
RT-PCR Egr2–5 Egr2–3
5)-CGG TGA CCA TCT TTC CCA ATG C-3) 5)-CTG TAC CAT GTA GGT CAC TCT G-3)
Egr2 promoter cloning First step 1-pro-5 1-pro-3 Second step 2-pro-5 2-pro-3
5)-ACG TGG TAA GAA CGC AGC TAT C-3) 5)-CCT GGG ATG GTA TCT CCT TTT G-3) 5)-AAA GAG CTC TCA GCT TCC GTG AAT GCA TG-3) 5)-AAA GCT AGC TCG CTC AGT TAG ACG GAA AG-3)
Egr2 coding region sequence cloning cod-5 5)-ATG ATG ACC GCC AAG GCC GTA C-3) cod-3 5)-GTG TAT CAG CCT GAG TCT CAT C-3)
66
Cells Tissues Organs 2003;174:63–72
Zeng/Sun/Wenger/Cheung
Fig. 2. Induction of Egr2 transcription de-
pends on BCP crystal concentration. BJ1 cells were treated the same as in figure 1. The concentrations of BCP crystals and PC (1 mM ) were added as indicated. Total RNAs were isolated 24 h afterward. The relative Egr2 mRNA levels were determined by the RT-PCR method the same as in figure 1.
blocks 95% of BCP crystal-induced Egr2 mRNA (fig. 2, lane 7). The Responsiveness of Egr2 Promoter to BCP Crystal Induction We have constructed an Egr2 promoter-luciferase reporter vector, pEgr2-luc, which contains 564 base pairs of the Egr2 promoter upstream of a luciferase reporter gene. The Egr2 promoter was cloned by PCR, using human genomic DNA as a template. To increase the PCR specificity, we designed two sets of Egr2 5) flanking region primers (table 1), based on the published sequence [Ranknekar et al., 1990], for a two-step nested PCR method as described in Methods. The PCR product was directly subcloned into pGEM-T Easy vector (Promega). After DNA sequencing the positive clone from both directions, we confirmed that the clone contained 14 base pairs of Egr2’s exon 1 5) noncoding sequence and 564 base pairs of its 5) flanking region sequence, and that there was a 100% match with the published data. The Egr2 promoter contains a TATA box and a number of important cis-DNA binding elements near the transcription start region, including a CRE and an SRE (fig. 3). To measure promoter
BCP Crystal Activation of Egr2 Gene Is PKC-Dependent
activity, we selected rabbit synovial fibroblasts as host cells rather than human fibroblast BJ1 cells because of their significantly higher transient transfection efficiency. About 5 ! 105 rabbit synovial fibroblasts were transiently transfected with 2 Ìg of pEgr2-luc using the LipofectAMINE method. The results in figure 3 showed that in rabbit cells BCP crystals and EGF induced luciferase activities driven by the Egr2 promoter by 13- and 5.4fold, respectively, in comparison to a negative control (fig. 3, lane 1). PC blocked 40% crystal-induced reporter activity (fig. 3, lane 4). Suppression of EGF or BCP Crystal-Induced Egr2 Transcription by Protein Kinase Inhibitors The protein kinase inhibitors bisindolylmaleimide I (Bis-I), U0126 and SB202190 were used to test whether members of the mitogen-activated protein kinase (MAPK) family of kinases are involved in EGF or BCP crystal-induced Egr2 transcription. Total RNAs from BJ1 cells were isolated at 1 and 24 h after treatment with EGF and BCP crystals, respectively, at the peaks of Egr2 induction times according to the time course study results. In figure 4, it was shown that SB202190, a specific inhibitor
Cells Tissues Organs 2003;174:63–72
67
Fig. 3. Induction of Egr2 promoter activity by EGF and BCP crystals. Rabbit synovial fibroblasts were plated at a density of 5 ! 105/well in six-well cluster plates in 2 ml of DMEM with 10% FBS. After attachment, the cells were starved with 2 ml of DMEM for 24 h, then transfected with 2 Ìg of pEgr2luc using lipofectAMINE for 18 h. The medium was replaced with 2 ml of fresh DMEM plus EGF, BCP crystals or PC as indicated. After 24 h, cells were lysed and the luciferase activity was assayed.
of SAPK/p38 MAPK, failed to block the induction of Egr2 transcription either by BCP crystals or by EGF (fig. 4a, b, lane 7). However, U0126, a specific inhibitor of the MAPK-ERK kinase, blocked both stimuli completely (fig. 4a, b, lane 6). This demonstrates that both EGF- and BCP-mediated Egr2 activation are regulated by the p44/ 42 MAPK pathway. Interestingly, these two inductions are quite different in terms of PKC dependency. Using Bis-I, a highly selective cell-permeable PKC inhibitor, we could specifically inhibit the induction of Egr2 mRNA by EGF in a Bis-I dose-dependent manner (fig. 4b, lanes 3–5). In contrast, Bis-I shows a stimulatory instead of an inhibitory effect on BCP crystal-induced Egr2 transcription (fig. 4a, lanes 3–5). This suggests that BCP crystalinduced Egr2 transcription is a PKC-independent process, in contrast to EGF-induced Egr2 transcription which is PKC-dependent. This is similar to what is seen in the cases of EGF versus BCP crystal-induced MMP-1 and MMP-3 expression [Brogley et al., 1999].
68
Cells Tissues Organs 2003;174:63–72
Calcium Dependency of Egr2 mRNA Induction by BCP Crystals We have shown that the BCP crystal-induced MMP-1 and MMP-3 expression is governed by a calcium-dependent PKC signal pathway [Hirano et al., 2000]. Next we wanted to test whether calcium plays a role in the BCPmediated induction of Egr2 mRNA. Cellular calcium levels are regulated by either influx of extracellular calcium from the tissue culture medium, or by mobilization of calcium from intracellular stores through a phospholipase C/IP3-DAG/PKC pathway. We examined the effects of an intracellular calcium chelator, TMB-8, on the induction of Egr2 mRNA by BCP crystals. The results in figure 5 show that TMB-8 (100 ÌM ) could completely block the induction of Egr2 mRNA by BCP crystals (lane 3). In addition, ammonium chloride, an inhibitor of intracellular BCP crystal dissolution, at a concentration of 10 and 25 mM, blocked the BCP crystal-induced Egr2 transcription by 80 and 95%, respectively (lanes 4 and 5). These
Zeng/Sun/Wenger/Cheung
Fig. 4. Inhibition of EGF or BCP crystalinduced Egr2 transcription by protein kinase inhibitors. BJ1 cells were treated the same as in figure 1. Cells were pretreated with the protein kinase inhibitors Bis-I, U0126 or SB202190, respectively, for 30 min before being stimulated with EGF or BCP crystals except lane 1 which served as negative control. Total RNAs were isolated 24 h afterward. The relative Egr2 mRNA levels, induced by BCP crystals (a) and EGF (b), were determined by the RT-PCR method the same as in figure 1. c The relative intensities of bands were scanned and plotted using ßactin bands as internal control for normalization. The results are expressed as means B SEM.
results suggest that the induction of Egr2 mRNA by BCP crystals is governed by a calcium-dependent PKC-independent p44/42 MAPK pathway. Transcriptional Factors Can Be Activated by the Induction of Egr2 Expression To further evaluate the consequences of the induction of Egr2 expression by BCP crystals, we used Clontech’s cis-acting enhancer element vectors in the Mercury Pathway Profiling System to test potential genes which could
BCP Crystal Activation of Egr2 Gene Is PKC-Dependent
be activated by a high level of Egr2 gene expression. Each Mercury vector contains a specific cis-acting DNA binding sequence located upstream to a TATA box, and a SEAP gene as a reporter. These promoters provide optimal induction of the reporter with a very low background. After each of these vectors was cotransfected with either pcDNA3.1-Egr2, an Egr2 expression vector driven by a strong CMV promoter, or the pcDNA3.1 vector by itself without insert as a negative control, the ability of the transcription factors AP-1, SRE, CREB, NFkB, MYC and
Cells Tissues Organs 2003;174:63–72
69
Fig. 5. Repression of BCP crystal-induced
Egr2 transcription by Ca2+ chelator TMB-8 and crystal dissolution inhibitor ammonium chloride. BJ1 cells were pretreated with the intracellular Ca2+ chelator TMB-8 (100 ÌM ) or crystal dissolution inhibitor ammonium chloride (10 or 25 mM ) for 30 min before being stimulated with BCP crystals. Determinations of the relative Egr2 mRNA levels were the same as in figure 1.
Fig. 6. Transcriptional factors activated by
Egr2 expression. Rabbit synovial fibroblasts were cotransfected with 1 Ìg of pAP1SEAP, pCRE-SEAP, pNFkB-SEAP, pMYCSEAP or pSRE-SEAP together with 1 Ìg of Egr2 expression vector pcDNA3.1-Egr2 or pcDNA3.1 as control for 18 h. The medium was then replaced with fresh DMEM with 10% FBS for another 24 h before taking 100 Ìl of the conditioned medium for SEAP assay. The results shown are representative of three independent experiments.
70
Cells Tissues Organs 2003;174:63–72
Zeng/Sun/Wenger/Cheung
MMP1 to mediate Egr2-dependent activation was determined by measuring the SEAP or luciferase reporter activity. The results in figure 6 showed that induction of Egr2 expression resulted in an increase in AP-1, MYC or SRF-dependent reporter activities by 6-, 11-, and 14-fold, respectively (lanes 1, 4 and 5), compared to a negative control. In contrast, activities of CRE and NFkB-dependent reporters did not change as much (lanes 2 and 3). These results suggest that transcription factors AP-1, MYC and SRF may be activated by BCP crystal-induced Egr2 expression but CRE and NFkB are not.
Discussion
Egr2, also known as Krox-20 in mouse [Joseph et al., 1988], encodes a 406-amino acid protein. The amino acid sequences in the three zinc finger regions are 92% identical to those of Egr1, also known as Krox 24, Zif268 and NGFI-A. McCubrey et al. [1991] showed that both Egr1 and Egr2 mRNAs had half-lives of 10–20 min after PMA treatment, which suggests that short mRNA half-lives contribute to the transient nature of expression of these genes. It has been reported that fetal bovine serum or purified growth factors can transiently induce Egr2 expression [Lemaire et al., 1988]. In the present study, we have shown that BCP crystals strongly stimulate Egr2 transcription in a different fashion compared to the stimulation of a growth factor such as EGF. The induction increased gradually and persistently up to 24 h. Furthermore, we show that the induction of Egr2 transcripts by BCP crystals follows the p44/42 MAPK pathway and is a PKC-independent process. The p44/42 MAPK-specific inhibitor U0126 can completely block BCP crystalinduced Egr2 transcription (fig. 4a, lane 6), while the PKC inhibitor Bis-I actually enhanced the induction of Egr2 transcripts by BCP crystals in a Bis-I concentrationdependent manner (fig. 4a, lanes 3–5). These results suggested that although both p44/42 MAPK and PKC pathways could be independently activated by BCP crystals and contributed to the induction of MMP1 [Brogley et al., 1999], only the p44/42 MAPK pathway was required for the induction of Egr2 by BCP crystals. Recently, we have shown that inhibition of BCP crystal-activated PKC with Bis-I did not block the activation of p44/42 MAPK [Reuben et al., 2002], which indicated that activation of p44/42 MAPK by BCP crystals is through a PKC-independent pathway. Here we provide another example which shows that BCP crystal-induced Egr2 expression also occurs through a PKC-independent p44/42 MAPK
BCP Crystal Activation of Egr2 Gene Is PKC-Dependent
pathway. In contrast, EGF induces Egr2 transcription transiently, peaking at 1 h after stimulation (fig. 1b). The induction can be completely blocked by both the p44/42 MAPK inhibitor U0126 (fig. 4b, lane 6) and the PKC inhibitor Bis-I (fig. 4b, lanes 3–5). Thus, depending upon the stimulus presented, p44/42 MAPK and Egr2 could both be activated either through a PKC-independent mechanism by BCP crystals, or through a PKC-dependent mechanism by EGF. It has been reported that BCP crystals can act like growth factor PDFG to exert a number of Ca2+-dependent early cell cycle events [Cheung et al., 1986]. Mazer et al. [1991] demonstrated that induction of Egr2 gene expression is a Ca2+-dependent process. Platelet-activating factor, a rapid inducer of free cytosolic Ca2+ concentration through both intracellular release and extracellular influx, can significantly induce Egr2 expression in a concentration-dependent manner in human B cells. Here we demonstrate that the induction of Egr2 transcripts by BCP crystals in fibroblasts is also a Ca2+-dependent process. The induction of Egr2 transcripts by BCP crystals could be totally blocked by TMB-8, an intracellular Ca2+ chelator. Ammonium chloride, an inhibitor of intracellular crystal dissolution, at concentrations of 10 and 25 mM could also block the Egr2 transcription induced by BCP crystals (fig. 5, lanes 3 and 4) and strongly suggests that the rise of intracellular Ca2+ concentration due to the BCP crystal dissolution is maybe the key that causes the induction. Previously, we have shown that EGF can stimulate an early intracellular calcium rise mainly through extracellular calcium influx [Halverson et al., 1998]. It is clear that the EGF-induced early rise of intracellular calcium plays an important role in activating Egr2 transcription because the induction can be blocked by the Ca2+-dependent PKC-specific inhibitor Bis-I (fig. 4b). In contrast, BCP crystals can induce a biphasic intracellular Ca2+ response in human fibroblasts [Halverson et al., 1998]. The first early intracellular calcium rise is believed to be caused by an extracellular calcium influx and the second caused by intracellular crystal dissolution. It was a surprise to us that the BCP crystal-induced early rise of intracellular calcium failed to induce an early Egr2 transcription like the EGF. These results suggested that early events triggered by BCP crystals and EGF were similar but not completely the same. Although the induction of EGR2 transcription is a Ca2+-dependent process, the BCP crystal-induced early rise of intracellular calcium is not sufficient to induce Egr2 gene expression per se.
Cells Tissues Organs 2003;174:63–72
71
We have demonstrated previously that PC can block a number of BCP crystal-induced events such as MMPs expression, p44/42 MAPK activation and mitogenesis in human fibroblasts [McCubrey et al., 1991; Cheung et al., 1996]. Again, using RT-PCR, we have shown here that PC can block 95% of BCP crystal-induced Egr2 transcription (fig. 2, lane 7), but only 40% of the BCP crystalinduced luciferase activity driven by the Egr2 promoter (fig. 3, lane 4). The different PC inhibitory efficiency could be attributed to the difference in mRNA stability and the half-life of luciferase. The half-life of firefly luciferase mRNA in mammalian cells is 6 h, which is about 18–36 times longer than that of Egr2 mRNA [Thompson et al., 1991].
Finally, using the Mercury signaling profiling reporter system we demonstrated that the induced Egr2 expression could in turn activate AP-1, MYC and SRF transcription factors (fig. 6, lanes 1, 4 and 5), all of which, when activated, are associated with cell proliferation. It has been reported that Egr2 can also promote gene expression of a number of growth factors and cytokines, such as fibroblast growth factor II [Svaren et al., 1998], human insulinlike growth factor II [Sussenbach et al., 1993] and IL-3 [Koyano-Nakagawa et al., 1994]. Taken together, our data have raised the possibility that the induction of Egr2 expression by BCP crystals may have a profound impact on cell proliferation, and therefore contribute to the progression of BCP crystal-associated osteoarthritis.
References Bhat, R.V., P.F. Worley, A.J. Cole, J.M. Baraban (1992) Activation of the zinc finger encoding gene krox-20 in adult rat brain: Comparison with zif268. Brain Res Mol Brain Res 13: 263– 266. Brogley, M.A., M. Cruz, H.S. Cheung (1999) Basic calcium phosphate crystal induction of collagenase I and stromelysin expression is dependent on a p42/44 mitogen-activated protein kinase signal transduction pathway. J Cell Physiol 180: 215–224. Cheung, H.S., J.D. Sallis, J.A. Struve (1996) Specific inhibition of basic calcium phosphate and calcium pyrophosphate crystal-induction of metalloproteinase synthesis by phosphocitrate. Biochim Biophys Acta 1315: 105–111. Cheung, H.S., M.T. Story, D.J. McCarty (1984) Mitogenic effects of hydroxyapatite and calcium pyrophosphate dihydrate crystals on cultured mammalian cells. Arthritis Rheum 27: 668–674. Cheung, H.S., J.J. Van Wyk, W.E. Russell, D.J. McCarty (1986) Mitogenesis activity of hydroxyapatite: Requirement for somatomedin C. J Cell Physiol 128: 143–148. Crosby, S.D., J.J. Puetz, K.S. Simburger, T.J. Fahrner, J. Milbrandt (1991) The early response gene NGFI-C encodes a zinc finger transcriptional activator and is a member of the GCGGGGGCG (GSG) element-binding protein family. Mol Cell Biol 11: 3835–3841. Evans, R.W., H.S. Cheung, D.J. McCarty (1984) Cultured human monocytes and fibroblasts solubilize calcium phosphate crystals. Calcif Tissue Int 36: 645–650. Feldman, M., F.M, Brennan, D. Chantry, C. Haworth, M. Turner, E. Abney, G. Buchan, K. Barrett, D. Barkley, A. Chu, M. Field, R.N. Maini (1990) Cytokine production in the rheumatoid joint implications for treatment. Ann Rheum Dis 49: 480–486.
72
Halverson, P.B., D.J. McCarty (1997) Basic Calcium Phosphate (Apatite, Octacalcium Phosphate, Tricalcium Phosphate) Crystal Deposition Disease: Calcinosis; in Koopman, W.J. (ed): Arthritis and Allied Conditions. Baltimore, Williams & Wilkins, p 2127. Halverson, P.B., A. Greene, H.S. Cheung (1998) Intracellular calcium response to basic calcium phosphate crystals in fibroblasts. Osteoarthritis Cartilage 6: 324–329. Hirano, S., C.D. Anuradha, S. Kanno (2000) Transcription of krox-20/egr-2 is upregulated after exposure to fibrous particles and adhesion in rat alveolar macrophages. Am J Respir Cell Mol Biol 23: 313–319. Joseph, L.J., M.M. Le Beau, G.A. Jamieson Jr., S. Acharya, T.B. Shows, J.D. Rowley, V.P. Sukhatme (1988) Molecular cloning, sequencing and mapping of EGR2, a human early growth response gene encoding a protein with ‘zincbinding finger’ structure. Proc Natl Acad Sci USA 85: 7164–7168. Koyano-Nakagawa, N., J. Nishida, D. Baldwin, K. Arai, T. Yokota (1994) Molecular cloning of a novel human cDNA encoding a zinc finger protein that binds to the interleukin-3 promoter. Mol Cell Biol 14: 5099–5107. Lemaire, P., O. Revelant, R. Bravo, P. Charnay (1988) Two mouse genes encoding potential transcription factor with identical DNA-binding domains are activated by growth factors in culture cells. Proc Natl Acad Sci USA 85: 5691–4695. Mazer, B., J. Domenico, H. Sawami, E.W. Gelfand (1991) Platelet-activating factor induces an increase in intracellular calcium and expression of regulatory genes in human B lymphoblastoid cells. J Immunol 146: 1914–1920. McCubrey, J.A., L.S. Steelman, J.P. McKeaun (1991) Interleukin-3 and phorbol esters induce different patterns of immediate-early gene expression in an interleukin-3 dependent cell line. Oncogene Res 6: 1–12.
Cells Tissues Organs 2003;174:63–72
Mechta, F., J. Piette, S.I. Hirai, M. Yaniv (1989) Stimulation of protein kinase C or protein kinase A mediated signal transduction pathways shows three modes of response among serum inducible genes. New Biol 1: 297–304. Mitchell, P.G., W.J. Pledger, H.S. Cheung (1989) Molecular mechanism of basic calcium phosphate crystal-induced mitogenesis. J Biol Chem 26: 14071–14077. Pankowski, A.H., J.D. Meehan, J.D. Sallis (1994) Synthesis via cyclic dioxatrichlorophosphorane of 1,3-dibenzyl-2-phosphonooxy citrate. Tetrahedron Lett 35: 927–930. Rangnekar, V.M., A.C. Aplin, V.P. Sukhatme (1990) The serum and TPA responsive promoter and intron-exon structure of Egr2, a human early growth response gene encoding a zinc finger protein. Nucleic Acids Res 18: 2749–2757. Reuben, P.M., M.A. Brogley, Y. Sun, H.S. Cheung (2002) Molecular mechanism of the induction of metalloproteinases 1 and 3 in human fibroblasts by basic calcium phosphate crystals. J Biol Chem 277: 15190–15198. Ryan, L.M., D.J. McCarty (1977) Calcium pyrophosphate crystal deposition disease, pseudogout, and articular chondrocalcinosis; in Koopman, W.J. (ed): Arthritis and Allied Conditions. Baltimore, Williams & Wilkins, p 2103. Schumacher, H.R. Jr. (1968) The synovitis of pseudogout: Electron microscopic observations. Arthritis Rheum 11: 426–435. Sussenbach, J.S., R.J. Rodenburg, W. Schejper, P. Holthuizen (1993) Transcriptional and posttranscriptional regulation of the human IGF-II gene expression. Adv Exp Med Biol 343: 63– 71. Svaren, J., B.R. Sevestson, T. Golda, J.J. Stanton, A.H. Swirnoff, J. Milbranst (1998) Novel mutants of NAB corepressors enhance activation by Egr transactivators. EMBO 17: 6010–6019. Thompson, J.F., L.S. Hayes, D.B. Lloyed (1991) Modulation of luciferase stability and impact on studies of gene regulation. Gene 103: 171– 177.
Zeng/Sun/Wenger/Cheung
Cells Tissues Organs 2003;174:73–86 DOI: 10.1159/000070576
Specialized Cranial Muscles: How Different Are They from Limb and Abdominal Muscles? James J. Sciote a Michael J. Horton a Anthea M. Rowlerson b Jason Link a a Department b Applied
of Orthodontics, School of Dental Medicine, University of Pittsburgh, Pittsburgh, Pa., USA; Biomedical Research Group, Guy’s, Kings and St. Thomas’ School of Biomedical Sciences, London, UK
Key Words Muscle W Myosin W Fiber types W Reverse transcription-polymerase chain reaction
Abstract Mammalian skeletal muscle fibers can be classified into functional types by the heavy chain (MyHC) and light chain (MyLC) isoforms of myosin (the primary motor protein) that they contain. Most human skeletal muscle contains fiber types and myosin isoforms I, IIA and IIX. Some highly specialized muscle fibers in human extraocular and jaw-closing muscles express either novel myosins or unusual combinations of isoforms of unknown functional significance. Extrinsic laryngeal mus-
Abbreviations used in this paper
ML/s MyHC MyLC PCR RT TMD
muscle lengths/second myosin heavy chain myosin light chain polymerase chain reaction reverse transcription temporomandibular disorders
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0073$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
cles may express the extraocular MyHC isoform for rapid contraction and a tonic MyHC isoform for slow tonic contractions. In jaw-closing muscles, fiber phenotypes and myosin expression have been characterized as highly unusual. The jaw-closing muscles of most carnivores and primates have tissue-specific expression of the type IIM or ‘type II masticatory’ MyHC. Human jaw-closing muscles, however, do not contain IIM myosin. Rather, they express myosins typical of developing or cardiac muscle in addition to type I, IIA and IIX myosins, and many of their fibers are hybrids, expressing two or more isoforms. Fiber morphology is also unusual in that the type II fibers are mostly of smaller diameter than type I. By combining physiological and biochemical techniques it is possible to determine the maximum velocity of unloaded shortening (Vo) of an individual skeletal muscle fiber and subsequently determine the type and amount of myosin isoform. When analyzed, some laryngeal fibers shorten at much faster rates than type II fibers from limb and abdominal muscle. Yet some type I fibers in masseter show an opposite trend towards speeds 10fold slower than type I fibers of limb muscle. These unusual shortening velocities are most probably regulated by MyHC isoforms in laryngeal fibers and by MyLC isoforms in masseter. For the jaw-closing muscles, this finding represents the first case in human muscle of physiological regulation of kinetics by light chains. To-
James Sciote Department of Orthodontics Salk Hall, 3501 Terrace Street Pittsburgh, PA 15261 (USA) Tel. +1 412 648 8419, Fax +1 412 648 8817, E-Mail
[email protected]
gether, these results demonstrate that, compared to other skeletal muscles, cranial muscles have a wider repertoire of contractile protein expression and function. Molecular techniques for reverse transcription of mRNA and amplification by polymerase chain reaction have been applied to typing of single fibers isolated from limb muscles, successfully identifying pure type I, IIA and IIX and hybrid type I/IIA and IIA/IIX fibers. This demonstrates the potential for future studies of the regulation of gene expression in jaw-closing and laryngeal muscles, which have such a variety of complex fiber types fitting them for their roles in vivo. Copyright © 2003 S. Karger AG, Basel
Introduction
Mammalian skeletal muscles may be characterized by their motor neuron, motor unit and muscle fiber properties. Physiological characteristics of these structures include motor neuron axon conduction velocity, recruitment order of skeletal muscle motor units, muscle tension, muscle activity pattern (either ‘tonic’ activity or ‘phasic’ activity) and fatigue resistance. A common way to describe muscle tissue diversity is through classification of skeletal muscle fiber types which link protein composition and metabolism to contractile performance. Mammalian muscle is classified into slow (type I) and fast (type II) groups, with subclassification of fast fibers including types IIA, IIB and IIX. Motor axon diameter, motor unit size, muscle fiber diameter, fusion frequency, contraction speed and tetanic force increase by fiber type in the order of I ! IIA ! IIX ^ IIB. Recruitment of motor units however is ordered from the slowest (in which slow-conducting motor axons innervate slow-contracting, type I, muscle fibers) to the fastest (in which fast-conducting motor axons innervate the largest and fastest, type IIB, muscle fibers) [Farina et al., 2002]. A large range of functional diversity is possible in skeletal muscle tissue by variations in fiber type distribution between muscle type, species differences and individual variation. Current evidence suggests that the IIB fiber type is rare or entirely absent in larger mammals including man [Snow et al., 1982; Horton et al., 2001], due probably to differences in isometric tension cost. Isometric tension cost, i.e. the ratio between energy expended and tension generated (PO), is lowest for type I fibers in rat muscle (0.66) and highest for type IIB fibers (2.90). Likewise in humans, tension cost increases from type I fibers (0.56) to type IIX fibers (1.76) [Stienen et al., 1996]. These ratios support the hypothesis that large
74
Cells Tissues Organs 2003;174:73–86
mammals have a lower energy cost for activation of skeletal muscles than smaller animals. The absence of type IIB fibers in larger mammals is probably a necessary adaptation for size. Variability beyond these general parameters is evidenced in ‘specialized’ cranial muscles as exemplified by the jaw-closing, intrinsic laryngeal and extraocular muscles. Unlike their counterparts in the limb and abdominal areas these groups have evolved to serve functions that are distinct from postural tonicity and locomotion. Two of these groups have evolved characteristic fiber types containing unique isoforms of myosin, the main motor protein of muscle. Type II masticatory (‘type IIM’) myosin heavy chain (MyHC) isoform is found in jaw-closing muscles of some species [Rowlerson et al., 1983], and the extraocular MyHC isoform in extraocular muscles [Wieczorek et al., 1985]. Data has suggested the possibility of novel myosins in intrinsic laryngeal muscles [DelGaudio et al., 1995], but definitive molecular evidence does not currently exist. Further, at least in rabbit muscle, it has been demonstrated with molecular techniques that the extraocular MyHC is expressed in laryngeal muscle [Briggs and Schachat, 2000]. The pattern of expression of these characteristics suggests that they are important functional specializations, of which extraocular and jaw-closing muscle groups can be regarded as extreme examples. The extraocular muscles have a similar function between species and therefore limited variability in anatomy and fiber type compositions. In contrast, the jaw-closing muscles have functions, anatomy and fiber types that vary tremendously between species. Laryngeal muscles have more modest variability in anatomy, perhaps more significant variability in function and their fiber type properties are currently under investigation [Brandon et al., 2003]. Adult type I and II fibers typically express only their specific MyHC isoform, yet in cranial muscles myosin coexpressions are an important and common feature. Myosin coexpression in these fibers has been shown to be relatively consistent and probably does not represent an orderly transition from one stable fiber type to another. Extraocular muscles are anatomically subdivided into global and orbital layers. The global layer contains fibers that may coexpress extraocular, slow, tonic and other fast MyHC isoforms. Orbital region fibers may coexpress tonic, extraocular, embryonic, neonatal or other fast MyHC isoforms [Sartore et al., 1987]. Hence, although many fast fibers in extraocular muscles are characterized by the presence of some ‘extraocular’ MyHC, most are
Sciote/Horton/Rowlerson/Link
Fig. 1. Immunohistochemical staining of two serial sections cut in the transverse plane from an area associated with the posterior mandible, close to the temporomandibular joint in an adult Monodelphis domestica (South American opossum). Heads were hemidissected then rapidly fixed in methacarnoy to preserve antigenicity prior to staining [Sciote and Rowlerson, 1998]. a Reactivity of anti-type IIM MyHC antibody for fibers in masseter, a jaw-closing muscle. c Reactivity of anti-type II (IIA, IIX and IIB) antibody for fibers in digastric, a jaw-opening muscle. b Craniofacial skeleton of Monodelphis, with line indicating the relative area stained in a and c.
likely to be hybrids rather that the homogenous types expressing a single myosin (as commonly found in limb muscles). Jaw-closing muscles of carnivores, primates [Rowlerson et al., 1983] and some marsupial mammals [Sciote et al., 1995; Sciote and Rowlerson, 1998] contain the IIM fiber type since many of their fibers express only type IIM MyHC and myosin light chains (MyLC) [Rowlerson et al., 1981] (summarized in fig. 1). Laryngeal muscles have some normal type I and II fibers, and some fibers with apparent stable coexpressions of myosin in-
cluding extraocular MyHC [DelGaudio et al., 1995] and tonic MyHC [Han et al., 1999]. Of special clinical interest are adaptations found in human jaw-closing muscles. Although a member of the primate order, human jaw-closing muscles do not contain type IIM fibers. Rather there is expression of a variety of myosins including type I, IIA, IIX, neonatal and atrial (·cardiac) [Sciote et al., 1994] and some embryonic MyLC [Butler-Browne et al., 1988]. Although some fibers express only one isoform, many have hybrid myosin content
Skeletal Muscle Biology Findings in Human Tissue
Cells Tissues Organs 2003;174:73–86
75
in which almost all possible combinations of two or more isomyosins may be found [Sciote et al., 1994; Monemi et al., 1996, 1998]. Hybrid fibers may also be identified in limb muscles, but in most cases represent cells in transition from one stable phenotype to another. In limb muscles, transitions between the stable fiber types I, IIA and IIX are found on an orderly continuum: I } (I + IIA) } IIA } (IIA + IIX) } IIX. Another unusual feature of human jaw closing muscles is the relatively small diameter of fibers containing fast (type II) myosins. These fast fibers are of such diminished size that their appearance is suggestive of disuse atrophy found in limb muscles (fig. 2). In considering the role specialized cranial muscles may play in musculoskeletal disorders it is important to consider their intrinsic differences from other skeletal muscle tissues. This report seeks to describe some of these differences by presenting new data and, in addition, reviewing recent findings principally from jaw-closing and laryngeal muscles.
Scientific Methods and Techniques
Histochemistry and Immunohistochemistry of Tissue Sections Tissue level investigations of skeletal muscle have sought to characterize the structure as an estimate of overall organ performance and contractility. Features such as fascicle organization, capillarity, fiber organization and fiber type classification are useful indicators of overall muscle tension, contraction speed and fatigability. Although opportunity is limited, where possible investigators have sampled muscle from living humans, usually as a needle biopsy in training and performance experiments or as a larger sample when muscle is resected as part of a standard surgical procedure. In the present study masseter muscle samples were taken from patients undergoing orthognathic surgical procedures for jaw repositioning. Although these patients demonstrated modest abnormalities in craniofacial growth, there is no indication of muscle abnormalities. Samples taken from similar subjects show characteristics of fiber type distribution and size
Fig. 2. Myofibrillar ATPase histochemical reactivity of serial sections of human sartorius muscle after incubation in buffer at pH 10.2 (a), 4.3 (b) and 4.6 (c). Type II fibers are reactive in alkali pH, type I in acid pH, and type IIX fibers moderately reactive in mild acid pH. [ = Type I fiber; $ = type IIX fiber; * = type IIA fiber. Magnification !200.
76
Cells Tissues Organs 2003;174:73–86
Sciote/Horton/Rowlerson/Link
[Daniel et al., 2001] resembling those described for subject populations including individuals with normal carniofacial growth [Thornell et al., 1984; Sciote et al., 1994; Monemi et al., 1998]. Laryngeal muscle samples were obtained from patients undergoing laryngectomy usually for treatment of neoplasm of epithelial origin, and in whom laryngeal muscle function was normal. Samples which represent normal limb muscle were collected subsequent to resections for treatment of orthopedic oncology. All surgical biopsies were obtained according to University Institutional Review Board guidelines for ethical use of human subjects in research. A standard preparation of a muscle biopsy is immediate snap-freezing of the tissue block at –80 ° C (or colder), followed by serial sectioning in a cryostat. Avoiding fixation assures appropriate tissue reactivity for histochemical and immunohistochemical staining. For determination of fiber type by myofibrillar ATPase histochemistry in human muscles a histochemical stain similar to that of Brooke and Kaiser [1970] is often used. Our technique is a modification of this original method [as described previously in Snow et al., 1982], but with reaction incubation times doubled to increase the intensity of staining for human muscle. Immunohistochemical staining is conducted using an indirect immunoperoxidase method after incubation of tissue sections in MyHCspecific antibodies. The following MyHC-specific antibodies were used: anti-type I monoclonal (slow skeletal MyHC – Sigma Aldrich clone NOQ7), antifast monoclonal (fast skeletal IIA, IIB and IIX MyHC – Sigma Aldrich clone MY-32), anti-IIA monoclonal (fast IIA – American Type Culture Collection clone SC-71), anti-IIB (American Type Culture Collection clone BF-F3), antineonatal polyclonal [Scapolo et al., 1991], antitonic polyclonal [Mascarello and Rowlerson, 1992], anticardiac monoclonal (cardiac ·MHC – Sera Lab clone MAS-366 and BAG5; American Type Culture Collection Clone BA-G5), and antiIIM polyclonal [Sciote et al., 1995]. Subsequent to staining serial sections with a panel of antibodies and a series of histochemical stains after various acid and alkali preincubations, biopsies are characterized for the proportions, size and types of fibers present. These fiber-type characteristics are used to infer physiological activity of the whole muscle. Single Fiber Physiology There are many physiological tests which may be utilized to determine muscle activity. Perhaps the most common is electromyographic activity as determined by surface electrodes placed on the skin overlying a muscle belly
Skeletal Muscle Biology Findings in Human Tissue
[Svensson et al., 2001], but this technique only describes general gross activity in whole muscle. Of interest, especially in cranial muscles, are the fine movements and motor control in performance of highly specialized tasks. In jaw-closing muscles, bite force transducers are used to determine muscle activity as it is translated into force produced at the occlusal surfaces of teeth [Dechow and Carlson, 1983]. Working under the constraint of biopsy samples, it is still possible to directly assess muscle cell performance using fiber bundle [Zuurbier et al., 1995] and single fiber [Edman, 1979] preparations. Single fiber preparations allow direct assessment of cell physiology to be correlated with contractile protein expression and fiber type. This is particularly important in specialized cranial muscles where novel myosins or unusual protein coexpressions have been identified. In data presented here we have used the slack-test protocol [Edman, 1979] to assess unloaded shortening velocity (Vo) of chemically skinned single fibers. Single fibers as short as 0.5–1.5 mm may be dissected from a skinned muscle fiber bundle and mounted via small stainless steel clamps between a force transducer and linear motor of the physiological testing equipment (Scientific Instruments, Heidelberg) in relaxing solution. Initial sarcomere length is adjusted to approximately 2.3–2.5 Ìm, as calculated by a first order laser diffraction pattern. The fiber is then activated to resting maximal tension using activating solution with an activating solution of FpCa2+ (4.5). The slack test protocol measures repeated ramp shortenings of the activated fiber and the amount of time necessary to remove introduced slack by sarcomeric shortening as described previously [Sciote and Kentish, 1996; Morris et al., 2001; Sciote et al., 2002]. The test determines Vo, in muscle lengths/second (ML/s). Subsequent to physiological testing, fiber cells are prepared for contractile protein isolation and identification. Identification of MyHC isoforms requires an adapted electrophoresis protocol with low acrylamide concentrations and addition of glycerol [Blough et al., 1996]. A single fiber provides sufficient protein load for glycerol gel electrophoresis and standard discontinuous SDS-PAGE. Standard SDS-PAGE allows separation and identification of actin and other contractile proteins including MyLC isoforms, but does not separate MyHC isoforms (which require glycerol gels). All gels are silver stained and optically scanned with Quantity-1 software (BioRad) to quantify band intensities for calculation of protein content. Where necessary, Western blots were used for isoform identification according to the method of LaFramboise et al. [1990].
Cells Tissues Organs 2003;174:73–86
77
Table 1. Primers for RT-PCR of MyHC genes
MyHC
Sense primer
Antisense primer
ß IIA IIX
5)-AGTAGCTTTGCCACATCTTGATCT-3) 5)-GAGCAGAGAAGGAGGAAAAGTGAC-3) 5)-CTGCAAGCAAAGGTGAAATCCT-3)
5)-TTCCTCCCAAGGAGCTGTTACACA-3) 5)-CTGCATAACGTTCTTTGAGGTTGT-3) 5)-TGGTCACCTTTCAGCAGTTTAGATAA-3)
Identification of MyHC Gene Message in Single Fibers As described above, transitions along a continuum from slow to fast MyHC isoforms occur in response to diverse physiological states that accompany changes in muscle fiber phenotypes. The changes between myosin isoforms are reversible, often resulting in mismatches between message and protein content within transitional fibers [Pette and Staron, 2001], as well as unusual patterns of coexpression for slow and fast myosin mRNAs [Horton et al., 2001] and the presence of some negative fibers that lack detectable myosin message [Smerdu et al., 1994; Ennion et al.,1995]. We are using reverse transcription (RT) and amplification by polymerase chain reaction (PCR) together with protein data from histochemistry and immunohistochemistry to investigate the plasticity of MyHC gene expression in tissue extracts and isolated single fibers of limb, laryngeal, extraocular and jaw-closing skeletal muscle complexes. For single fiber studies, muscle biopsies are quickly separated into thin longitudinal strips, snap frozen in isopentane-dry ice and lyophilized in sterile microfuge tubes. Individual fibers that range in length from 2 to 8.5 mm are dissected from the freezedried strips. RNA is isolated from the single fibers with guanidinium thiocyanate using RNAqueous™-4PCR kits (Ambion) and digested for 30 min at 37 ° C with 1 Ìg DNAse. Approximately 1.5% of the total RNA from each fiber is reverse-transcribed at 50 ° C for 1 h in the presence of 0.4 ÌM oligo dT(18) and 20 pg/Ìl random hexamer primers and amplified using Titanium™ One-Step RTPCR kits (Clontech Laboratories) in an MJ Research PTC-200 thermal cycler. PCR reactions include an initial denaturation at 94 ° C for 5 min followed by 30 cycles of denaturation (30 s at 94 ° C), annealing (30 s at 60 ° C) and elongation (1 min at 72 ° C). Final elongation is done for 5 min at 72 ° C. Oligonucleotide primers for PCR (table 1) were selected from GenBank mRNA sequences for MyHC-I/ß (accession XM-033374), MyHC-IIA (accession XM-012618), MyHC-IIX (accession XM-052590) and MyHC-IIB (accession XM-017815) using Jellyfish software (LabVelocity) and synthesized by a commercial
78
Cells Tissues Organs 2003;174:73–86
source (Invitrogen). QuantumRNA™ 18S internal standards (Ambion) were amplified using a 7:3 primer to competimer ratio. Electrophoretic analyses of PCR products were done on 5-Ìl aliquots out of 15-Ìl reaction volumes in 3% NuSieve® 3:1 agarose gels containing 0.5 Ìg/ ml ethidium bromide. Gels were photographed on a UV light box using an AlphaImager 2000 system (Alpha Innotech) and a densitometric analysis of bands was performed with Quantity One 4.1 (Bio-Rad Laboratories) quantitation software.
Results and Discussion
Fiber Types and Their Variability in Human Jaw-Closing Muscles The jaw-closing muscles of carnivores and primates contain an additional fast fiber type, with a notably rapid ATPase activity and fast twitch speed, termed IIM or ‘type II masticatory’ [Rowlerson et al., 1981]. This fiber type, which contains a distinct isoform of myosin, is thought to be characteristic of carnivorous behavior or aggressive biting [Rowlerson et al., 1983]. We have recently extended these observations to marsupial mammals including the North American [Sciote et al., 1995] and South American [Sciote and Rowlerson, 1998] opossum (fig. 1) and marsupial mammals of Australia including the microbat and kangaroo [Hoh et al., 1996]. These marsupial species demonstrate dramatic differences between fiber type composition among type II fibers when jaw-opening and jaw-closing muscles are compared. Jawclosing muscles as exemplified by masseter are composed predominately of type IIM fibers, while the jaw-opening muscles (such as anterior digastric) contain predominately limb type II fibers. The jaw-closing muscles of man are notably different, containing fibers of an exceptionally wide range of fiber diameters and MyHC composition, but no IIM fiber types as found in other primates. Before describing human jawclosing muscles it is necessary to provide insight into
Sciote/Horton/Rowlerson/Link
human skeletal muscle in general since there are some species differences. In the limb muscles of small mammal types IIB and IIX may be found in abundance, but as animal size increases the relative proportion of IIB decreases substantially. In human muscle the IIB fiber type classified by ATPase histochemistry muscle actually contains the IIX homologue and no IIB myosin or transcripts [Smerdu et al., 1994; Ennion et al., 1995]. Hence, human limb and abdominal muscles of normal health and function are composed of type I, IIA and IIX fibers (fig. 2). Human jaw-closing muscles contain many fiber types, several of which coexpress MyHC isoforms [Sciote et al., 1994; Monemi et al., 1996, 1998]. The fiber type composition of human masseter is also distinctive compared to typical limb muscle in that its type II fibers are typically small in diameter compared to the normally sized type I fibers (fig. 3a–c). This unusual size of the type II fibers is reminiscent of a limb muscle that is subject to disuse, producing preferential atrophy and diameter decrease in type II fibers (fig. 3f). Another extreme phenotype occurs in masseter hypertrophy where type II fibers approach the average fiber diameter of their counterparts in limb muscle (fig. 3d). The subject we sampled with the condition of masseter hypertrophy possessed a long-term neurologic condition causing spasm of the masseter resulting in extreme muscle hyperfunction. Unlike our subject, masseter hypertrophy usually occurs as a gradual progressive increase in size of the muscle in the area located over the ramus of the mandible not accompanied by spasm or parotid gland involvement [Buchner et al., 1979]. Furthermore masseter hypertrophy may be associated with symptoms commonly described in temporomandibular disorders (TMD) such as temporal headache, jaw tension, trismus, indefinable masseter pain and joint pain [Beckers, 1977], and is probably the best example of direct muscle involvement in TMD. Jaw-closing muscles do contain type I, IIA and IIX fibers with homogeneous isoform content, but type I, IIA and IIX MyHC isoforms can all be coexpressed in various combinations with neonatal and ·-cardiac MyHCs. We have distinguished two fiber types in jaw-closing muscles as ‘neonatal’ and ‘atrial’ on the basis of their MyHC expression; these types are found quite often in human jaw-closing muscles but rarely, if ever, in other skeletal muscles. The ‘neonatal’ fiber type expresses the neonatal MyHC in various combinations of type I, IIA and IIX, but is a mature fiber in other respects. Likewise, the ‘atrial’ fiber type expresses the cardiac ·-MyHC in combination with type I, IIA and/or IIX even though it is a normal skeletal muscle fiber. In jaw-closing muscles of human
Skeletal Muscle Biology Findings in Human Tissue
subjects, these fiber types have characteristic sizes (fig. 4). These unusual fiber types characteristic of human jawclosing muscles are currently the subject of physiological investigation since there are no known animal models that closely mimic this phenotype. An added complication of extrapolating physiological results obtained at single fiber level to the whole jaw-closing muscles is the level of heterogeneity that exists in these muscles across all organizational levels from individual fibers to the whole muscle in man. Although type I fibers predominate overall in human jaw-closing muscles, there are both systematic differences in the percentage of type I fibers in different parts of the muscles, and highly variable compositions of remaining fiber types (fig. 5) [Ringqvist et al., 1982; Eriksson and Thornell, 1983]. Thus, we are currently investigating the question of whether fiber type diameter and percent composition differ significantly in subject populations with varying craniofacial morphologies. A preliminary answer (published in abstract form) can be found in a report on a group of orthodontic patients from whom a masseter muscle biopsy was taken during an orthognathic surgical procedure to reposition one or both jaws as part of overall orthodontic treatment [Daniel et al., 2001]. There appear to be significant differences in fiber types between extremes of facial types, but the variability in fiber type in these orthodontic groups probably requires sampling from a larger patient population to develop firm conclusions. Single Fiber Physiology In mammalian skeletal muscle, type I fibers containing type I MyHC are associated with a slower unloaded shortening velocity (Vo), and type II fibers containing type II MyHCs with a faster Vo, as measured by the slack test [Reiser et al., 1988; Bottinelli et al., 1991]. Fibers heterogeneous for type I and type II isoforms have intermediate shortening speeds, with Vo proportional to the amount of type II isoform present [Reiser et al., 1985]. The correlation of shortening velocity to MyHC isoform content within fast-contracting type IIA, IIB and IIX fibers is less certain when compared across species. In the rabbit [Sweeney et al., 1988] and human [Larsson and Moss, 1993], IIX fibers have a faster Vo than IIA fibers, which may be correlated to MyHC isoform content, but in the rat this correlation has been questioned [Bottinelli et al., 1994a]. The relative content of essential light chains (the actual amount of MyLC1f vs. MyLC3f in a muscle fiber) in type II fibers of rabbit may modify MyHC-dependent shortening speed, producing a greater range of Vo for each fiber type [Sweeney et al., 1988]. But in the rat, the rela-
Cells Tissues Organs 2003;174:73–86
79
3
80
Cells Tissues Organs 2003;174:73–86
Sciote/Horton/Rowlerson/Link
Fig. 4. Morphometric analysis of average fiber type diameter after biopsy sampling from a young adult population (jaw closing, jaw opening, abdominal and facial). Number of subjects: masseter n = 28, rectus (abdominal muscle) n = 4, buccinator n = 3 and mylohyoid n = 3. A portion of this data is taken from Sciote et al. [1994].
Fig. 5. Percent composition for each fiber type in masseter muscle biopsies sampled in figure 4 presented as boxplots. The boxes contain the median and the fourths of data and resist gross distortion of outlying values, here demonstrating the masseter muscle’s striking variability in % fiber type composition.
tive essential light chain content may modify Vo to the extent that some IIB fibers may shorten more slowly than IIA fibers [Bottinelli et al., 1994b]. This strong correlation of MyLC content to Vo was not demonstrated in human type II limb muscle fibers [Larsson and Moss, 1993]. Similar experiments were first conducted by us in rabbit masseter muscle, since this species contains fibers that homogeneously express ·-cardiac MyHC like that present in some human masseter fibers. When tested for Vo, ·-fibers have a mean Vo of 0.78 ML/s, which was intermediate in speed between masseter and soleus type I and type II fibers [Sciote and Kentish, 1996]. The only slightly faster Vo of ·-cardiac fibers relative to type I fibers is surprising since in cardiac muscle the ·-cardiac MyHC is thought of as fast-contracting, in the range of 4–5 ML/s [Sweitzer and
Moss, 1993]. However, shortening velocities in cardiac myocytes are regulated by other factors in addition to myosin proteins [Ford, 1991], so direct comparison to skeletal fibers is probably not appropriate. A further complication exists when comparing the results between cardiac and skeletal muscle cells. In cardiac myocytes ·-cardiac MyHC is typically combined with the atrial MyLCs, MyLC1A and MyLC2A, while in ·-cardiac fibers of masseter, ·-cardiac MyHC is combined with the slow skeletal MyLCs, MyLC1s and MyLC2s [d’Albis et al., 1993]. We have now tested shortening velocity in chemically skinned fibers from human limb, masseter and two intrinsic laryngeal muscles, the posterior cricoarytenoid and thyroarytenoid, in relation to the type and amount of myosin isoforms present [Morris et al., 2001]. Sampling from a variety of limb or abdominal muscles, we were able to determine unloaded shortening velocities for fibers that were comparable to previous studies of human fibers and with average velocity increasing in the order of type I ! IIA ! IIX. This data, presented in figure 6, shows a threedimensional bar chart with single fiber data separated into bins according to shortening velocity, fiber type and MyHC isoform content. The observed data for type I fibers of masseter and type II fibers of intrinsic laryngeal muscles exemplify the differences between cranial and limb muscle. Masseter fibers homogeneous for type I MyHC have shortening velocities that span almost the
Fig. 3. Staining of biopsy sections taken from the masseter muscle of a young healthy adult (serial sections a–c), a middle aged adult with masseter hypertrophy secondary to muscle spasm (serial sections d, e) and the vastus lateralis of a middle aged man with preferential type II fiber atrophy subsequent to inactivity (section f). a Reactivity of anti-type I MyHC antibody. b Reactivity of anti-type II MyHC antibody. c Reactivity for anti-type IIA MyHC antibody. Myofibril-
lar ATPase reactivity is shown after incubation in buffer at pH 10.2 (d, f) and after pH 4.6 (e). [, P = Type I fiber; ), $ = type IIA fiber; * = fibers coexpressing both type I and II MyHC isoforms. Magnification !600.
Skeletal Muscle Biology Findings in Human Tissue
Cells Tissues Organs 2003;174:73–86
81
Fig. 6. Summary of Vo (unloaded shortening velocity) values for chemically skinned single fibers from selected limb, masseter and laryngeal muscles. Fibers are positioned on lanes based upon muscle of origin and MyHC isoform composition, and into bins based upon shortening velocity. Please note that the Vo scale changes at a rate of 0.2 ML/s in the range of 0.0–2.2, then in 1.0 ML/s changes from 3.0 to 5.0.
entire range of shortening values found in type I and type II limb fibers, so factors other than MyHC isoform content must help regulate contraction. Within this group of type I fibers there is a subpopulation of fibers that contracted at a much lower range of shortening speeds than that of limb muscle. This spread of data for six fibers with an unusually low Vo (!0.1 ML/s) is illustrated in masseter lane, bin 1 of figure 6. Subsequently we isolated and identified MyLC content in addition to MyHC content in these fibers. Through standard SDS-PAGE and Western blotting with MyLC-specific antibodies we were able to identify an additional MyLC isoform present in masseter fibers with unusually slow Vo, namely MyLC embryonic, but absent in the majority of fibers studied (fig. 7a). There is only one other report of changes in MyLC isoform content modifying shortening velocity in human muscle
82
Cells Tissues Organs 2003;174:73–86
[D’Antona et al., 2002]. In skinned human thyroarytenoid muscle fibers they found relatively large variability in Vo of type IIA fibers. Faster IIA fibers contained decreased amounts of MyLC2f (the regulatory MyLC usually associated with fast MyHC isoforms) and IIA fibers with slower Vo contained increased amounts of MyLC2s (the regulatory MyLC usually associated with the type I or slow MyHC isoform). Further physiological techniques, such as the in vitro motility assay could help confirm these results. Some type II fibers from intrinsic laryngeal muscles have an exceptionally fast shortening velocity. In figure 6, bins 5–7 contain fibers with velocities ranging from approximately 3- 5 ML/s that are very fast in comparison to the fastest IIX fiber bin in limb muscle with approximately 2.2 ML/s. Electrophoretic isolation of MyHC iso-
Sciote/Horton/Rowlerson/Link
Fig. 7. a Electrophoretic analysis of MyLC isoenzymes in standard 14% acrylamide discontinuous SDS-PAGE with single human muscle fiber cells sampled from poster cricoarytenoid (PCA), masseter (Mass.), triceps, gastrocnemius (Gastroc) and thyroarytenoid (TA) muscles. Two slow and three fast MyLC isoforms are resolved with the addition of one embryonic MyLC species found in some masseter fibers. The masseter fiber in lane 3 has a Vo value found in bin 1 of the masseter type I fiber lane (fig. 6). b Electrophoretic analysis of MyHC isoenzymes in modified 9% polyacrylamide gels containing
0.1% SDS and 6% glycerol and visualized by silver staining. The control marker was taken from a whole biopsy sample homogenate of human tibialis muscle (lane 1). Other muscle samples are labeled by lane: atria, fetal, extraocular (EO), thyroarytenoid (TA) and posterior cricoarytenoid (PCA) muscles. Relative mobility of known protein species was neonatal 1 I 1 IIA 1 · 1 IIX. Two additional bands (labeled MHCa and MHCb) may be identified in EO, TA and PCA muscles only.
forms from these fibers has been inconclusive since myosin content is heterogeneous, including variable expression of two additional species (fig. 7b). This information suggests that further work is necessary to determine MyHC isoform content in human laryngeal muscle before final physiological observations are made.
typically comprise the continuum of fiber-type transitions in skeletal muscle. The RT-PCR products of four such representative fibers are shown in figure 8, which is a composite from different agarose gels that were run under identical conditions of electrophoresis. Each panel in the figure displays a data set from different single fibers whose RNA was consecutively reverse-transcribed and amplified under the same conditions, but in separate reaction tubes containing 18S, MyHC-I/ß, MyHC-IIA or MyHC-IIX-specific PCR primers. Type I, I/IIA, IIA and IIA/IIX fibers were obtained from lyophilized strips of
MyHC Gene Message in Single Fibers Single-fiber methods for the analysis of MyHC-RNA by RT-PCR were used to detect individual representatives of homogeneous and hybrid gene expressions that
Skeletal Muscle Biology Findings in Human Tissue
Cells Tissues Organs 2003;174:73–86
83
Fig. 8. Representative human muscle fiber types analyzed by agarose gel electrophoresis of RT-PCR products from single fibers. RNA from isolated single fibers of sartorius (I, I/IIA, IIA and IIA/IIX types) and semitendinous (IIX type) muscles was reverse-transcribed and amplified using primers for 18S rRNA internal control (a), MyHC-I/ß (b), MyHC-IIA (c) and MyHC-IIX (d). Amplified cDNA products were 489 bp (18S), 89 bp (IIB), 176 bp (IIA) and 224 bp (IIX).
human sartorius whereas the type IIX fiber came from human semitendinous muscle. Equivalent volumes between samples were maintained throughout for RNA isolation, RT-PCR reactions and agarose gel electrophoresis. However, concentrations of RNA in the single fiber samples were not adjusted for equivalency, so that the intensities of the ethidium bromide-stained 18S control bands, as determined by densitometric analysis using Quantity One™ software, differ between the data sets. Although the dissected fibers were approximately the same length, variations in their diameter and weight, as well as some losses encountered during isolation steps could all contribute to differences in RNA recoveries. Nevertheless, the intensities of bands for RT-PCR of 18S rRNA were approximately equal in the type I, I/IIA and IIA/IIX fibers whereas intensities of the bands in the IIA and IIX sets were, respectively, 75 and 23% of maximum values. Because of the possible inequalities between amounts of RNA and efficiencies of RT-PCR for the MyHC species examined here, comparisons between data sets are primarily qualitative, with the identification of MyHC-ß/I, IIA and IIX expressing pure fibers and I/IIA and IIA/IIX expressing hybrid fibers being the foremost observation. Secondarily, comparisons of the intensities between 18S internal controls and PCR product bands were done to determine a relative abundance of MyHC RNA within the data set for each fiber. Analysis shows that amounts of I/ß, IIA and
84
Cells Tissues Organs 2003;174:73–86
IIX message, relative to 18S rRNA, were 74, 58 and 211%, respectively, in the pure fibers. In the I/IIA hybrid fiber, the relative abundance of MyHC-I/ß was 62% and IIA was 9% of 18S. In the IIA/IIX hybrid fiber, IIA was 64% and IIX was 101% of 18S. Both hybrid fibers may be in transition to the faster phenotype. Also, according to the ‘nearest neighbor rule’ proposed for combinations of MyHC isoforms in fibers [Pette and Staron, 2001], the relative quantities of RNA detected here indicate that the first hybrid fiber contained F6.7! more MyHC-I/ß than IIA and it is properly classified I/IIA or IC. The relative abundance of MyHC-IIX RNA was F1.6-fold greater than IIA in the second hybrid fiber, which is more correctly termed type IIX/IIA. Of course this classification only applies if there is directly proportionate translation of message into protein. In future studies it will be interesting to directly compare message to protein amounts, especially in cases of dysfunction or at least altered function. The power of single fiber techniques and quantitative RTPCR [Wright et al., 1997] for analysis of gene expression in muscle has previously been demonstrated. The further application of single fiber methods presented here with modifications for quantitative analysis of mRNA in relation to protein isoform composition will be extremely important for characterization of the diverse fiber types and transitional states in muscles of the temporomandibular complex.
Sciote/Horton/Rowlerson/Link
Applications for Skeletal Muscle Biology in Temporomandibular Joint Research
TMD may encompass a wide variety of more specific diseases or disorders which often result in pain, tenderness and other symptoms in the joint, surrounding connective structures and/or skeletal muscle. In the extreme, pathological changes such as jaw-closing muscle inflammation and hypertrophy may be imaged and described, most commonly with magnetic resonance imaging [Yang et al., 2001]. Masticatory muscles are directly involved in positioning of the joint and disc, especially the lateral pterygoid muscle which inserts directly into the anteromedial portion of the articular disc. The articular disc also attaches to the fascia of the masseter and temporalis muscles. Beyond these direct connections afferent information from muscle spindles from all of the masticatory muscles is used to provide overall jaw posture including positioning of the temporomandibular joint [Schmolke, 1994]. Although it is likely that muscle tissue does not ‘cause’ TMD this does not exclude the possibility that it plays a role in exacerbating or prolonging joint problems. Although recent investigations involving cell biology and physiology of cranial muscles have provided interesting information, they have not so far provided evidence for muscle tissue involvement in the genesis of TMD disorders. At present we are still in the initial stages of understanding the range of metabolic and physiological diversi-
ty found in human cranial muscle tissue. The greatest challenge for the immediate future is a precise description of this diversity, given a number of practical and ethical constraints. Any investigation of cranial muscle as an organ must be conducted on living subjects undergoing clinical treatments. Cadaver specimens which allow greater volumes of tissue sampling are of limited value, as they generally do not allow direct investigation of physiological functions. Hence findings in ongoing work often depend upon relatively small sample sizes in comparison to the overall size of muscle. An additional complication with regard to jaw-closing muscles is the tremendous amount of variability observed between different subjects, even when samples are collected from the same local area of muscle by one surgeon and processed and analyzed using the same techniques by the researcher. One major problem for the future will be to determine the extent to which individual variations in normal jaw function are related to these differences in fiber type composition. Another will be identifying criteria which allow distinction between normal variation and pathological changes. This will require larger, multifaceted clinical studies of form and function in human subjects in whom adequate clinical diagnosis is known together with the results of skeletal muscle biology research. Ultimately the question remains: Are there clinical markers for developing TMD in muscle tissue, which may be incorporated into diagnosis and treatment of these complicated problems?
References Beckers, H.I. (1977) Masseter hypertrophy and its intraoral surgical correction. J Maxillofac Surg 5: 28–35. Blough, E.R., E.R. Rennie, F. Zhang, P.J. Reiser (1996) Enhanced electrophoretic separation and resolution of myosin heavy chains in mammalian and avian skeletal muscles. Anal Biochem 233: 31–35. Bottinelli, R., R. Betto, S. Schiaffino, C. Reggiani (1994a) Maximum shortening velocity and coexistence of myosin heavy chain isoforms in single skinned fast fibers of rat skeletal muscle. J Muscle Res Cell Motil 15: 413–419. Bottinelli, R., R. Betto, S. Schiaffino, C. Reggiani (1994b) Unloaded shortening velocity and myosin heavy chain and alkali light chain isoform composition in rat skeletal muscle fibers. J Physiol 478: 341–349. Bottinelli, R., S. Schiaffino, C. Reggiani (1991) Force-velocity relations and myosin heavy chain isoform compositions of skinned fibres from rat skeletal muscle. J Physiol 437: 655– 672.
Skeletal Muscle Biology Findings in Human Tissue
Brandon, C.A., C. Rosen, G. Georgelis, M.J. Horton, J. Sciote (2003) Muscle fiber type composition and effects of vocal fold immobilization on the two compartments of the human posterior cricoarytenoid: A case study of four patients. J Voice, in press. Briggs, M.M., F. Schachat (2000) Early specialization of the superfast myosin in extraocular and laryngeal muscles. J Exp Biol 203:2485–2494. Brooke, M.H., K.K. Kaiser (1970) Muscle fiber types: How many and what kind. Arch Neurol 23: 369–379. Buchner, A., R. David, D. Temkin (1979) Unilateral enlargement of the masseter muscle. Int J Oral Surg 8: 140–148. Butler-Browne, G.S., P.O. Eriksson, C. Laurent, L.E. Thornell (1988) Adult human masseter muscle fibers express myosin isozymes characteristic of development. Muscle Nerve 11: 10– 20. d’Albis, A., M. Anger, A.M. Lompre (1993) Rabbit masseter muscle expresses the cardiac · myosin heavy chain gene. Evidence from mRNA sequence analysis. FEBS Lett 324: 178–180.
Daniel, Y., J. Ferri, T. Krivosic-Horber, F. McDonald, G. Raoul, H. Reyford, A. Rowlerson (2001) Masseter muscle fibre types in relation to craniofacial form. J Physiol 531: 154–155P. D’Antona, G., A. Megighian, S. Bortolotto, M. Antonietta Pellegrino, R. Marchese Ragona, A. Staffieri, R. Bottinelli, C. Reggiani (2002) Contractile properties and myosin heavy chain isoform composition in single fibre of human laryngeal muscles. J Muscle Res Cell Motil 23: 187–195. Dechow, P.C, D.S. Carlson, (1983) A method of bite force measurement in primates. J Biomech 16: 797–802 DelGaudio, J.M., J.J. Sciote, W.R. Carroll, R.M. Escalmado, (1995) Atypical myosin heavy chain in rat laryngeal muscle. Ann Otol Rhinol Laryngol 104: 237–245. Edman, K.A. (1979) The velocity of unloaded shortening and its relation to sarcomere length and isometric force in vertebrate muscle fibres. J Physiol 291: 143–159.
Cells Tissues Organs 2003;174:73–86
85
Ennion, E., P.J. Sant’Ana, A.J. Sargeant, A. Young, G. Goldspink (1995) Characterization of human skeletal muscle fibres according to the myosin heavy chains they express. J Muscle Res Cell Motil 16: 35–43. Eriksson, P.O., L.E. Thornell (1983) Histochemical and morphological muscle-fibre characteristics of the human masseter, the medial pterygoid and the temporalis muscles. Arch Oral Biol 28: 781–795. Farina, D., M. Fosci, R. Merletti (2002) Motor unit recruitment strategies investigated by surface EMG variables. J Appl Physiol 92:235–247. Ford, L.E. (1991) Mechanical manifestations of activation in cardiac muscle. Circ Res 68: 621– 637. Han, Y., J. Wang, D.A. Fischman, H.F. Biller, I. Sanders (1999) Slow tonic muscle fibers in the thyroarytenoid muscles of human vocal folds; a possible specialization for speech. Anat Rec 256: 146–157. Hoh, J.F., P. Hugues, Q. Thomas (1996) Masticatory myosin expression in marsupial mammals of Australia. Basic Appl Myol 18: 32–45. Horton, M.J., C.A. Brandon, T.J. Morris, T.W. Braun, K.M. Yaw, J.J. Sciote (2001) Abundant expression of myosin heavy-chain IIB RNA in a subset of human masseter muscle fibres. Arch Oral Biol 46: 1039–1050. LaFramboise, W.A., M.J. Daood, R.D. Guthrie, P. Moretti, S. Schiaffino, M. Ontell (1990) Electrophoretic separation and immunological identification of type 2X myosin heavy chain in rat skeletal muscle. Biochim Biophys Acta 1035:109–112. Larsson, L., R.L. Moss (1993) Maximum velocity of shortening in relation to myosin isoform composition in single fibres from human skeletal muscles. J Physiol 472: 595–614. Mascarello, F., A.M. Rowlerson (1992) Myosin isoform transitions during development of extraocular and masticatory muscles in the fetal rat. Anat Embryol 185: 143–153. Monemi, J., P.O. Eriksson, I. Dubail, G.S. ButlerBrowne, L.E. Thornell (1996) Fetal myosin heavy chain increases in human masseter muscle during aging. FEBS Lett 386: 87–90. Monemi, J., P.O. Eriksson, A. Eriksson, L.E. Thornell (1998) Adverse changes in fibre type composition of the human masseter versus biceps brachii muscle during aging. J Neurol Sci 154: 35–48. Morris, T.J., C.A. Brandon, M.J. Horton, J.J. Sciote (2001) Maximum shortening velocity and myosin heavy-chain isoform expression in human masseter fibers. J Dent Res 80: 1845– 1848.
86
Pette, D, R.S. Staron (2001) Transitions of muscle fiber phenotypic profiles. Histochem Cell Biol 115: 359–372. Reiser, P.J., C.E. Kasper, M.L. Greaser, R.L. Moss (1988) Functional significance of myosin transitions in single fibers of developing soleus muscle. Am J Physiol 254:C605–C613. Reiser, P.J., R.L. Moss, G.G. Giulian, M.L. Greaser (1985) Shortening velocity in single fibers from adult rabbit soleus muscles is correlated with myosin heavy chain composition. J Biol Chem 260:9077–9080. Ringqvist, M., I. Ringqvist, P.O. Eriksson, L.E. Thornell (1982) Histochemical fibre-type profile in the human masseter muscle. J Neurol Sci 53: 273–282. Rowlerson, A., F. Mascarello, A. Veggetti, E. Carpene, (1983) The fibre-type composition of the first branchial arch muscles in carnivora and primates. J Muscle Res Cell Motil 4: 443–472. Rowlerson, A., B. Pope, J. Murray, R.B. Whalen, A.G. Weeds (1981) A novel myosin present in cat jaw-closing muscles. J Muscle Res Cell Motil 2: 415–438. Sartore, S., F. Mascarello, A. Rowlerson, l. Gorza, S. Ausoni, M. Vianello, S. Schiaffino (1987) Fibre types in extraocular muscles: A new myosin isoform in the fast fibers. J Muscle Res Cell Motil 8: 161–172. Scapolo, P.A., A. Rowlerson, F. Mascarello, A. Veggetti (1991) Neonatal myosin in bovine and pig tensor tympani muscle fibers. J Anat 178: 255– 263. Schmolke, C. (1994) The relationship between the temporomandibular joint capsule, articular disc and jaw muscles. J Anat 184: 335–345. Sciote, J.J., J.C. Kentish (1996) Unloaded shortening velocities of rabbit masseter muscle fibres expressing skeletal or alpha-cardiac myosin heavy chains. J Physiol 492: 659–667. Sciote, J., T. Morris, C. Brandon, M. Horton, C. Rosen (2002) Unloaded shortening velocity and myosin heavy chain variations in human laryngeal fibers. Ann Otol Rhinol Laryngol 111:120–127. Sciote, J.J., A. Rowlerson (1998) Skeletal fiber types and spindle distribution in limb and jaw muscles of the adult and neonatal opossum, Monodelphis domestica. Anat Rec 251: 548– 562. Sciote, J.J., A.M. Rowlerson, D.S. Carlson (1995) Myosin expression in the jaw-closing muscles of the domestic cat and American opossum. Arch Oral Biol 40:405–413.
Cells Tissues Organs 2003;174:73–86
Sciote, J.J., A.M. Rowlerson, C. Hooper, N.P. Hunt (1994) Fibre type classification and myosin isoforms in the human masseter muscle. J Neurol Sci 126: 15–24. Smerdu, V., I. Karsch-Mizrachi, M. Campione, L. Lienwand, S. Schiaffino (1994) Type-IIx myosin heavy chain transcripts are expressed in type-IIb fibers of human skeletal muscle. Am J Physiol 267:C1723-C1728. Snow, D.H., R. Billeter, F. Mascarello, E. Carpene, A. Rowlerson, E. Jenny (1982) No classical type-IIB fibres in dog skeletal muscle. Histochemistry 75: 53–65. Stienen, G.J., J.L. Kiers, R. Bottinelli, C. Reggiani (1996) Myfibrillar ATPase activity in skinned human skeletal muscle fibres: Fibre type and temperature dependence. J Physiol 493: 299– 307. Svensson, P., A. Burgaard, S. Schlosser (2001) Fatigue and pain in human jaw muscles during a sustained, low-intensity clenching task. Arch Oral Biol 46: 773–777. Sweeney, H.L., M.J. Kushmerick, K. Mabuchi, F.A. Sreter, J. Gergely (1988) Myosin alkali light chain and heavy chain variations correlate with altered shortening velocity of isolated muscle fibers. J Biol Chem 263:9034–9039. Sweitzer, N.K., R.L. Moss (1993) Determinants of loaded shortening velocity in single cardiac myocytes permeabilized with ·-hemolysin. Circ Res 73: 1150–1162. Thornell, L.E., R. Billeter, P.O. Eriksson, M. Ringqvist (1984) Heterogeneous distribution of myosin in human masticatory muscle fibres as shown by immunohistochemistry. Arch Oral Biol 29: 1–5. Wieczorek, D.F., M. Periasamy, G.S. ButlerBrowne, R.G. Whalen, B. Nadal-Ginard, (1985) Coexpression of multiple myosin heavy chain genes, in addition to a tissue-specific one, in extraocular musculature. J Cell Biol 101: 618–629. Wright, C., F. Haddad, A.X. Qin, K.M. Baldwin (1997) Analysis of myosin heavy chain mRNA expression by RT-PCR. J Appl Physiol 83: 1389–1396. Yang, X.Y., H. Pernu, J. Pyhtinen, P.A. Tiilikainen, K.S. Oikarinen, A.M. Raistoa (2001) MRI findings concerning the lateral pterygoid muscle in patients with symptomatic TMJ hypermobility. J Craniomand Pract 19: 260–268. Zuurbier, C.J., J.W. Heslinga, M.B. Lee-de Groot, W.J. Van der Laarse (1995) Mean sarcomere length-force relationship of rat muscle fibre bundles. J Biomech 28: 83–87.
Sciote/Horton/Rowlerson/Link
Cells Tissues Organs 2003;174:87–96 DOI: 10.1159/000070577
Sex Differences in Rabbit Masseter Muscle Function Arthur W. English a Charles G. Widmer b a Department
of Cell Biology, Emory University, Atlanta, Ga., and b Department of Orthodontics, University of Florida, Gainesville, Fla., USA
Key Words Temporomandibular W Masseter W Motoneurons W Sexual dimorphism W Testosterone
Abstract The proportions of fibers of different phenotypes in the rabbit masseter muscle differ strikingly in adult males and females. Muscles from females contain similar proportions of small fibers that express both the slow/ß and cardiac · myosin heavy chain (MyHC) isoforms and larger fibers containing the IIa MyHC isoform. In muscles from males, nearly 80% of fibers are of the IIa phenotype. To evaluate the functional significance of these sex differences, we used finely graded intramuscular microstimulation to study the contractile properties of masseter motor units in 16-month-old male and female rabbits. Twitch forces and torques in males were significantly
greater in magnitude than those of females. Greater proportions of units that produced larger forces/torques were encountered in the males. The same motor units produced force or torque more rapidly in males than in females, principally because units in which twitch rise times were 1 22 ms were found only in females. The forces applied to the mandible and the torques generated about the right mandibular condyle were studied during cortically evoked rhythmic activation of the masticatory muscles. The overall range of torque rise times and the magnitudes of the peak torques did not differ between sexes. The mean rise time was significantly shorter and the mean peak torque was significantly greater in males. We conclude that sex differences in fiber phenotype proportions are reflected in sex differences in motor unit properties and in the function of these motor units during rhythmic activation. Copyright © 2003 S. Karger AG, Basel
Introduction Abbreviations used in this paper
ANOVA MyHC
analysis of variance myosin heavy chain
ABC
© 2003 S. Karger AG, Basel 1422–6405/03/1742–0087$19.50/0
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/cto
The muscles of mastication of several mammals are sexually dimorphic with respect to the proportions of fibers of different phenotypes [Gutmann et al., 1970; Bass et al., 1971; Maxwell et al., 1979; Lyons et al., 1986; d’Al-
Arthur W. English Department of Cell Biology, Emory University School of Medicine 405P Whitehead Research Bldg., 615 Michael Street Atlanta, GA 30322 (USA) Tel. +1 404 727 6250, Fax +1 404 727 3677, E-Mail
[email protected]
bis et al., 1992, 1993; English et al., 1999b; Eason et al., 2000a]. In the rabbit masseter muscle, nearly equal numbers of large fibers containing the IIa myosin heavy chain (MyHC) isoform or smaller fibers containing both the cardiac · and slow/ß (type I) MyHC isoforms are found in adult females and young adults of both sexes [English et al., 1999b]. Under the influence of testosterone, masseter muscles of adult male rabbits come to be composed overwhelmingly of fibers containing the IIa MyHC isoform [Eason et al., 2000b; English and Schwartz, 2002]. Thus masseter muscle fibers in adult male rabbits undergo a naturally occurring large-scale MyHC isoform switch from ·-slow/ß to IIa which is androgen-induced. The rabbit masseter muscle is compartmentalized. Based on the orientation of muscle fibers with respect to internal tendons and the pattern of branching of the nerve to the masseter, we recognized 20–25 different unique subvolumes of the muscle [Widmer et al., 1997]. Each of these subvolumes is thought to be innervated exclusively by a unique group of masseter motoneurons [English and Letbetter, 1982; Weijs et al., 1993; Widmer et al., 1997]. In most, but not all of these compartments sex differences in muscle fiber phenotype composition exist, as described above. In some, notably those in the posterior deep masseter, no such sex differences exist [Eason et al., 2000b; Reader et al., 2001]. The functional significance of sex differences in masseter muscle fiber phenotype is not known. Functional differences could be reflected in the contractile properties of whole muscles or neuromuscular compartments or in those of individual motor units, a single masseter motoneuron and the muscle fibers it innervates. In prior studies of contractile properties of rabbit masseter motor units, the sex of the animal subjects studied often is not treated as a variable or all of the subjects were young adult males [Kwa et al., 1995a, b; van Eijden and Turkawski, 2001; Turkawski and von Eijden, 2001]. Masseter motor units are reportedly universally fast-contracting [van Eijden and Turkawski, 2001], although a large range of contraction speeds is found [Kwa et al., 1995a, b]. A similarly large range of forces has been reported for rabbit masseter motor units [van Eijden and Turkawski, 2001]. One goal of the present study was to evaluate whether sex differences in contractile properties of rabbit masseter motor units might exist that correlate with sex differences in muscle fiber phenotype proportions. Based on the results of studies of the contractile properties of skinned muscle fibers, the contraction speed of a fiber is highly correlated with its MyHC isoform [Bottinelli et al., 1994a]. Fibers containing the slow/ß MyHC
88
Cells Tissues Organs 2003;174:87–96
isoform contract most slowly [Bottinelli et al., 1994a], those containing also the cardiac · MyHC isoform contract more rapidly [Sciote and Kentish, 1996], and those containing the IIa isoform contract even more rapidly [Bottinelli et al., 1994a]. Even though a substantial variability exists in contraction speeds within populations of fibers containing the same MyHC isoform [Bottinelli et al., 1994a; Sciote et al., 2002], and even though this variability might be explained by differences in the composition of myosin light chains [Bottinelli et al., 1994b; Moss et al., 1995; Sciote et al., 2002], we would predict that motor units in the masseter muscles of adult male rabbits would contract more rapidly than those in masseter muscles of females. The force produced by a motor unit is proportional to the total cross-sectional area of its muscle fibers. Since the specific tension the force produced per unit of cross-sectional area of muscle fibers does not vary with MyHC isoform content [Bottinelli et al., 1991; Totosy de Zepetnek et al., 1992] larger motor units are composed of larger muscle fibers and, especially, larger numbers of muscle fibers (innervation ratio) [Bodine et al., 1988; Totosy de Zepetnek et al., 1992]. We have shown that rabbit masseter fibers containing the IIa MyHC isoform are much larger than those containing the · and slow/ß isoforms, and that these fibers are significantly larger in males than females [English et al., 1998]. Weijs and colleagues [Weijs et al., 1993; Kwa et al, 1995b] have shown that the innervation ratios of motor units containing these IIa fibers are larger than innervation ratios of motor units composed of slow/ß fibers. From these data, we would predict that motor units in the masseter muscles of male rabbits would produce larger forces than those of females. Even if the contractile properties of rabbit masseter motor units are significantly different in males and females, it will be important to know whether those differences are reflected in their normal function. We have described the forces and torques produced by rabbit masticatory muscles during cortically evoked rhythmic activation [Widmer et al., 2003]. Based on analysis of the kinetics and EMG activity recorded, this rhythmic activation of the masticatory muscles is strikingly similar to that described for unrestricted chewing of different natural foods [Weijs and Dantuma, 1981; Weijs et al., 1989]. However, sex differences in these biomechanical data were not analyzed. The final goal of this paper will be to evaluate sex differences in masticatory muscle mechanics during rhythmic activation.
English/Widmer
Methods Experiments were conducted using New Zealand white rabbits. Seven males and six females, all at least 6 months old, were used. Rabbits are sexually mature at 2 months old, but they continue to grow until they are 6 months old [Thompson and Berndtson, 1993]. All experiments were in accordance with NIH guidelines and the Institutional Animal Care and Use Committee of Emory University approved protocols for them. In each experiment, animals were preanesthetized with acepromazine (10 mg/kg) and xylazine (20 mg/kg), and after 45 min, they were given halothane (1–2.5%, in 95% O2-5% CO2) by pediatric mask until a surgical plane of anesthesia was achieved. An endotracheal tube was then inserted for the remainder of the experiment. The right masseter muscle was exposed and EMG electrodes were inserted into the center of a number of different neuromuscular compartments (fig. 1). A pair of electrodes was also introduced into the right digastric muscle. Each electrode consisted of a pair of fine (50 Ìm diameter), enamel-coated stainless steel wires (Stablohm 800a, California Fine Wire, Grover Beach, Calif., USA), on which the distal 0.5 mm had been removed by scraping with a scalpel blade. The tips of the wires were separated by approximately 0.5 mm. These wire electrodes were used to record muscle activity during rhythmic activation but also for intramuscular microstimulation of masseter motor units (see below). Once these electrodes had been implanted, all skin incisions were closed with wound clips. The rabbit was then mounted in a modified stereotactic device to stabilize the head. The anterior part of the skull was held secure by a clamp over the incisor teeth on the hard palate and the posterior part was held in place by metal pins attached to the mastoid processes. A multi-axis force-moment sensor (Gamma 65/5, Assurance Technologies) was attached to the mandible by two small screws driven into the base of the incisor teeth at the level of the alveolus (fig. 1). No mandibular movements were possible in this configuration. The force-moment sensor measured the force applied to its faceplate as three analog outputs, corresponding to the three rectilinear components of the applied force, as noted in figure 1. The device also measured three moments or components of a torque vector about a userdefined point in space. We set this global center at the point at which the device was attached to the mandible. Activation of the masticatory muscles could induce a torque about this point, which is known as a reaction moment. The torque output of the device was thus configured to measure such a reaction moment [see also English et al., 1999a]. The recorded reaction moment was subtracted vectorially from the torques calculated from recorded forces (see below). Once the rabbit had been attached to the transducer, recordings of masseter motor unit forces were begun. Using the implanted EMG wire electrodes, we activated motor units in different masseter compartments using intramuscular microstimulation. We applied short (0.05 ms duration) constant voltage pulses at progressively increasing amplitudes that covered the full range of stimulus intensities available. Application of such short pulses results in the activation of muscle fibers synaptically, by stimulation of axon branches in the vicinity of the electrode tips [Hultman et al., 1983; Popovic et al., 1991; Rhee et al., 1992]. Pulse amplitude was controlled through the 12-bit analog output of a laboratory computer system. This output was amplified to drive a conventional stimulus isolation unit over a voltage range of 0–12 V, meaning that the stimulus strength was changed in increments of 3 mV. Using such finely graded ramps of stimulus intensity and conventional signal averaging methods, we were able to
Sex Differences in Rabbit Masseter Muscle Function
Fig. 1. Experimental setup. After the rabbit’s skull was immobilized
in a rigid head holder, the mandible was attached to a multi-axis force moment sensor via two small screws placed into the site of eruption of the lower incisor teeth. Motor units were studied by electrical stimulation of the masseter compartments indicated. The system of rectilinear components of the forces recorded and the torques computed from them are shown in the box below. Arrows indicate the direction of positive values of each.
detect very small changes in force (as little as 2 mN), associated with the activation of single motor units. When stimuli were applied with such gradually increasing ramps in intensity, clear voltage thresholds for the recruitment of different motor units could be noted (fig. 2A). As stimulus intensity was increased, an all or nothing twitch was noted in the force record and this same twitch was maintained at several subsequent small increases in stimulus intensity, until a larger twitch was encountered. This pattern was repeated until several such force increments were noted. As long as these increments were modest and a clear threshold for the response could be established, we interpreted the changes in force as due to the recruitment of motor units. Successive twitches of different magnitude were then tempo-
Cells Tissues Organs 2003;174:87–96
89
1 Sec
A
100 mN
B
10 msec
C
rally aligned (fig. 2B) and vectorially subtracted from one another to reveal the twitch force produced by individual motor units (fig. 2C). These criteria for motor unit isolation could not always be applied. In some cases, distinct stimulus thresholds could not be detected. Successive small increases in the stimulus intensity resulted in large increases in twitch forces so that we could not establish a clear threshold in intensity. In other cases, very large increments in twitch force were encountered, suggestive of nearly synchronous recruitment of several motor units. In both such scenarios we con-
90
Cells Tissues Organs 2003;174:87–96
20 mN
Fig. 2. Motor unit extractions. Long trains of intramuscular stimuli of gradually increasing strength were delivered to different masseter compartments and twitch forces were recorded (A). Stepwise increases in twitch force (arrows) at distinct thresholds were interpreted as due to the recruitment of new motor units. When overlaid, these were noted as gradually increasing twitches (B). Individual motor unit twitches were extracted by digital subtraction of these waveforms (C).
10 msec
cluded that we were unable to isolate single motor units and we studied these force records no further. We used the recorded twitch forces to calculate the torque produced by single motor units about the right mandibular condyle. We favor torque as a measure of the mechanical properties of motor units because it includes in a single vector the force generating capability of the activated muscle fibers and the mechanical arrangement of those muscle fibers in the head. At the end of each experiment, we determined the position vector between the center of the force
English/Widmer
moment sensor and the center of the right mandibular condyle using precision calipers. The cross product of this position vector and the recorded force vector is the torque produced about the mandibular condyle [for details, see English et al., 1999a]. There are three rectilinear components of this torque vector: pitch, yaw and roll (fig. 1). For our analysis, we computed the resultant of these components. We chose to calculate the torque rather than measure it directly using the transducer because we used the transducer to measure any reaction moment (see above). We studied the rise time of the resultant of the twitch torque vector as a measure of contraction speed. The peak of the resultant of the torque vector was used to measure contraction strength or motor unit size. In addition to studying the torque magnitude, we analyzed the magnitude of the resultant of the recorded force vector, in an attempt to compare our results to those published by others [van Eijden and Turkawski, 2001]. The significance of sex differences in these properties was evaluated using a one-way analysis of variance (ANOVA). This method tests the hypothesis that the variance in either twitch rise time, peak force, or peak torque between sexes was significantly different from the variance between individual rabbits of each sex. Significance of differences was set at p ! 0.05. To evaluate sex differences in masticatory muscle function, we measured the forces applied to the mandible during cortically evoked rhythmic activation and then calculated the torques about the right mandibular condyle. In the halothane-anesthetized rabbits, an opening in the skull over the left sensorimotor cortex was made just to the left and anterior to bregma. The dura mater was incised to expose the cortical surface. At this point the rabbit was given an intravenous injection of urethane (15 mg/kg) and halothane was discontinued. The animal breathed 95% oxygen and 5% CO2 for the remainder of the experiment. Supplemental doses of urethane were administered, as necessary, to maintain deep anesthesia, as verified by the lack of a corneal blink or limb withdrawal reflexes. A metal microelectrode (CBCSH75, Frederic Haer, Bowdonton, Me., USA) was lowered 2– 3 mm below to cortical surface and used to deliver short (0.1 ms duration) constant voltage trains of stimuli (50 Hz, 330 ms, 1/s). If the electrode was placed in a favorable location, rhythmic activation of the masticatory muscles was induced. This was noted as alternating jaw opening and jaw closing forces and rhythmic bursts of EMG activity recorded from wire electrodes in the right masseter muscle. Once rhythmic activation was induced, the stimulation of the cortex was stopped and force, torque, and EMG data were recorded onto the disc of a laboratory computer system. Typically, rhythmic activation persisted for 2–5 min after the termination of electrical stimulation and could be reinduced by subsequent stimulation. To avoid fatigue, we waited at least 5 min between periods of inducing rhythmic activation. We analyzed these data in a cycle-by-cycle manner. A cycle was defined as the interval between the ends of burst of EMG activity recorded from the digastric muscle. During each cycle we determined the torques produced about the right mandibular condyle as pitch, yaw and roll components, as described above. Using these components we calculated the resultant of this torque vector and measured both its peak and its rise time. We evaluated the significance of sex differences in these two measures of muscle mechanics using ANOVA, as described above.
Sex Differences in Rabbit Masseter Muscle Function
Results
Masseter Motor Units We studied the mechanical properties of 54 masseter motor units in two male rabbits and in 95 motor units in three adult female rabbits. Only motor units from neuromuscular compartments where we have found sex differences in the proportions of fiber of different phenotypes were studied [English et al., 1999b]. In figure 3, sex differences in these motor unit properties are displayed as histograms. The median twitch contraction time of masseter motor units is considerably shorter in males than females (17.27 vs. 26.00 ms, fig. 3A). These differences are statistically significant (ANOVA, p ! 0.05). In males the range of twitch contraction times is more restricted than noted in females. The slowest contracting motor unit in males is faster than that of most motor units in females and no motor units in females generate force as rapidly as the fastest third of units in males. The magnitudes of masseter motor unit twitch torques and forces found in female rabbits were significantly smaller (ANOVA, p ! 0.05) than that found in males (fig. 3B, C). Much of the significance of these differences can be observed as the large proportion of very small motor units (!20 mN W cm or !6 mN) found in females relative to that of males (fig. 3B, C). Rhythmic Activation We evaluated the forces/torques produced during rhythmic activation of the masticatory muscles. In urethane-anesthetized animals, stimulation of the cortical masticatory region results in a rhythmic activation of the muscles of mastication [Schwartz et al., 1989; Liu et al., 1993; English and Widmer, 2002; Widmer et al., 2003] and the mandible is moved in trajectories similar to that observed in unrestrained rabbits chewing on natural foods [Weijs and Dantuma, 1981; Weijs et al., 1989]. Because we have restrained the mandible to record forces during this stimulation, chewing movements were not produced. Thus we will refer to the outcome of our paradigm as rhythmic activation rather than mastication. The three rectilinear components of the force vector recorded during an epoch of rhythmic activation are shown in figure 4A. The same data are plotted against one another in figure 4B. Each cycle of rhythmic activation begins with the application of a jaw closing (+Fz) force (fig. 4A, filled symbol). During the elaboration of jaw closing forces, both retrusive (+Fy) and balancing (+Fx) forces are applied, but the directions of these applied forces shift
Cells Tissues Organs 2003;174:87–96
91
rapidly as Fx changes from balancing toward working directions and Fy decreases. The closing force then returns to zero. Jaw opening (–Fz) forces are then applied. Along with this jaw opening force both a significant protrusive (–Fy) component of force and a balancing side (+Fx) component of force are applied to the mandible. The cycle ends with the start of the next jaw closing force. The magnitude and direction of all of these forces are compatible with the movements of the mandible described by others, either during cortically evoked rhythmic activation [Schwartz et al., 1989; Liu et al., 1993] or unrestrained chewing [Weijs and Dantuma, 1981]. In figure 4C, components of the torque vector about the right mandibular condyle are shown for the same cycles of rhythmic activity as figure 4A. These components of the torque vector produced during a typical cycle are what might be expected from the forces described above. During the period of jaw closing forces, the large negative (closing) pitch (Ùx) component of the torque vector dominates, but this is accompanied by a substantial negative (working) yaw (Ùz) torque and a small but variable positive (labial) roll (Ùy) component. During opening, a positive pitch is accompanied by a positive yaw, and slight negative roll. To assay whether the torques produced during rhythmic activation differ between male and female rabbits, we examined the resultant of the torque vector. An example of the resultant of the torque vector for a single cycle of rhythmic activity is shown in figure 5A. We measured the peak of this vector and the rise time of the vector during the application of jaw closing torque as measures of the magnitude and speed of muscle contraction, respectively. We analyzed the peak torque and the rise time of the peak torque from 774 cycles of rhythmic activity in three adult females and 1,900 cycles of activity in four adult males. The peak torque found in males was significantly greater than that of females (ANOVA, p ! 0.001). Similarly, a large range of rise times for this peak torque was encountered, and their variances were not equal. Rise times in males were significantly (ANOVA, p ! 0.05) shorter than those of females.
Fig. 3. Sex differences in rabbit masseter motor unit twitch properties. A Distribution of twitch contraction times are shown for 95
motor units recorded from female rabbit masseter muscles and for 54 motor units from male rabbits. B, C Similar histograms are used to show the distribution of peak torques about the right mandibular condyle and the peak forces recorded from the same motor units, respectively.
92
Cells Tissues Organs 2003;174:87–96
Discussion
Significant sex differences in the proportions of fibers of different phenotype have been described for the muscles of mastication of guinea pigs [Gutmann et al., 1970; Bass et al., 1971; Lyons et al., 1986], mice [Eason et al., 2000b], macaques [Maxwell et al., 1979] and rabbits [En-
English/Widmer
A
Fz
Close Open Retrusion
Fy
Protrusion Balancing
Fx
Working
B
Fz (N)
4 Close
2
0 Retrusion 1.2 0.8 0.4
Open
0 -0.4
Protrusion
-0.8 -0.2
-0.1
0
0.1
Fx (N)
g cin lan Ba
Fy (N)
or W g kin
Fig. 4. Mechanics of cortically evoked
rhythmic activation of the masticatory muscles. In urethane-anesthetized rabbits, stimulation of the cortical masticatory area evoked rhythmic activation of the masticatory muscles and forces were recorded in three dimensions as shown in figure 1. A Typical forces in the three different planes are shown; the scale bars are 0.5 N for Fx and 1.0 N for Fy and Fz. B These same component forces are plotted against each other. The cycle of rhythmic activation begins at the filled symbol and courses in the direction of the arrows. C The three rectilinear components of the torque vector about the right (working) mandibular condyle are shown for the same cycles of activitiy shown in A. Scale bars are 20 N W cm for pitch, 10 N W cm for yaw, and 2 N W cm for roll.
Sex Differences in Rabbit Masseter Muscle Function
C
Open
x
Close
Labial
y
Lingual Balancing
z
Working
1 sec
Cells Tissues Organs 2003;174:87–96
93
Fig. 5. Sex differences in rhythmic activation torques. A A trace from the resultant torque vector for a single cycle of rhythmic activation is shown to illustrate the measurements taken. B, C These measurements are plotted as histograms for 1,900 cycles in four males and 774 cycles in three females.
94
Cells Tissues Organs 2003;174:87–96
glish et al., 1999b]. In all of these species, androgens are implicated in shifting phenotype proportions to include more fibers thought to contract more rapidly. Despite these results, there is no available information as to whether these structural and molecular sex differences are reflected in differences in masticatory muscle function. The principal finding of this study is that significant sex differences exist in the mechanical properties of motor units in the rabbit masseter muscle, and that these sex differences are found also during cortically evoked rhythmic activation of the masticatory muscles. The masseter muscles of adult females and young adults of both sexes contain nearly equal numbers of fiber of two phenotypes: ·-slow/ß and IIa. In adult males, nearly 80% of the fibers in the masseter contain the IIa MyHC isoform [English et al., 1999b]. Based on the results of studies of single muscle fiber mechanics, fibers containing the IIa MyHC isoform contract more rapidly than those containing either the slow/ß or · MyHC isoforms [Bottinelli et al., 1994a]. Thus one might expect that motor units in masseter muscles of females would contract more slowly than those of males because of the smaller proportion of fibers of the IIa phenotype and larger proportion of ·-slow/ß fibers in the females. In our sample of masseter motor units from adult male and female rabbits, we found that twitch rise times of masseter motor units in males were significantly faster than those of females. Masseter muscles from males contained only very fast contracting units and none were observed with rise times 122 ms. Despite this predicted finding, none of the motor units studied could be considered slow twitch, even though nearly half of the fibers contain the slow/ß MyHC isoform in females. In one respect this finding is not unexpected. Others [van Eijden and Turkawski, 2001] have reported that motor units in the masseter muscles of young adult males, whose fiber composition resembles that of adult females [Bredman et al., 1992; Eason et al., 2000b], are nearly universally fast contracting. This designation was based both on twitch rise times and the ‘sag’ response to an unfused tetanus [Kwa et al., 1995b]. The reasons for this apparent paradox are not clear. It is possible that the presence of the slow/ß MyHC isoform in rabbit masseter muscle fibers does not confer on them the same slow twitch contractile properties found in rat soleus fibers [e.g. Gillespie et al., 1986]. Rome et al. [1990] have shown that the very same MyHC isoform expressed in muscle fibers of different species is associated with a wide range of different intrinsic shortening velocities. Rabbit masseter fibers containing the cardiac · MyHC isoform contract slightly more rapidly than the slow/ß-containing fibers of
English/Widmer
other muscles [Sciote and Kentish, 1996]. It is possible also that molecules other than the MyHC isoform regulate the speed of contraction of masseter muscle fibers. Different myosin light chain isoforms are known to have profound effects on contraction speed [Bottinelli et al., 1994b; Moss et al., 1995] and it is possible that they may contribute substantially to the speed of rabbit masseter motor units. In young adult male rabbits, Weijs and coworkers [Kwa et al., 1995a, b], using glycogen depletion of muscle fibers, showed that individual motor units are not homogeneous with respect to muscle fiber phenotype. Most single masseter motoneurons were shown to innervate fibers of more than one phenotype [Kwa et al., 1995b]. If this same unique pattern of innervations exists in adult rabbits, then it is possible that the fast contraction speeds found in adult masseter motor units also could be the result of such heterogeneous innervation. Muscle fibers containing the IIa MyHC isoform are larger than those of the ·-slow/ß phenotype and in adult males these fibers are nearly twice the size of those found in females [English et al., 1998, 1999b]. Given that males also contain many more of these larger fibers than females, one might expect that motor units would produce larger forces in males than females. Based on the sample of motor units we studied, this prediction is borne out. In males, median peak twitch torque and force are nearly twice that of females. It is difficult to compare our data directly to that reported for motor units in young adult males produced by others, since they have normalized their recorded force data to the ratio of mandibular length and an estimate of the moment arm of the active muscle fibers in each motor unit [Kwa et al., 1995b]. However, we recalculated the published values for 24 masseter motor units in young adult males [Kwa et al., 1995b], using their measure of mandibular length (6 cm), and the average of the moment arms they estimated [Kwa et al., 1995b], to make their data more easily comparable to ours. The mean twitch tension so determined was 12.87 mN. This size is remarkably similar to the median value we obtained in adult females (12.73 mN). Using the same transformed data, we found that more than half of these 24 units produced twitch forces that were !6 mN, the largest category of twitch tensions found in adult females in our study (fig. 3C). Given that the cross-sectional area of muscle fibers of different phenotype in adult females is very similar to that reported for young adult males [Kwa et al., 1995a; English et al., 1998], we interpret our findings to mean that androgens must increase the magnitude of the forces produced by rabbit masseter
motor units in adult males by increasing the cross-sectional areas of fibers of the IIa phenotype. Sex differences in the mechanical properties of rabbit masseter motor units are reflected in sex differences in the torques about the working side mandibular condyle during rhythmic activation. The rise time of the torque is faster and the peak torque is larger during rhythmic activation in males. The ranges of both speeds and peaks are similar in both sexes, but cycles of rhythmic activation producing large or fast torques occurred more often in males. However, differences between males and females in these mechanical measures found during rhythmic activation are more subtle than the sex differences found in the torques or forces produced by single motor units. The most reasonable explanation of this apparent disparity is related to the recruitment of motor units that occurs during rhythmic activation. Masseter motor unit torques and forces are much smaller in females because of the preponderance of small motor units in the female. During rhythmic activation, masseter motor units are recruited in a size-ordered manner [Sokoloff et al., 1999], so that the mechanical contribution of small motor units to the large torques encountered must be very small. The contributions of larger motor units (140 mN W cm or 120 mN) would dominate. Sex differences in the proportions of these larger motor units are less clear and more in the range of differences found during rhythmic activation. It is not clear why these functional sex differences exist. We know of no reports of sex differences in dietary preferences in rabbits or reports of behaviors such as dominance struggles that could be marshaled in support of a simple biological explanation for the differences that we, and others, have observed. Until such data are forthcoming, we are left to consider sex differences in masticatory muscle structure and function as examples of fascinating secondary sex characteristics of unknown adaptive significance.
Sex Differences in Rabbit Masseter Muscle Function
Cells Tissues Organs 2003;174:87–96
Acknowledgements This work was supported by grants DE11536 and HD32571 from the USPHS. The authors are indebted to Dr. Dario Carrasco and Dr. John Hermanson for their helpful comments on earlier versions of the manuscript, to Dr. Michael Kutner for advice on statistical analyses, and to William Goolsby, who designed and built the computercontrolled stimulator used and wrote the software to make it work.
95
References Bass, A., E. Gutmann, V. Hanzlikova, I. Syrovy (1971) Sexual differentiation of enzyme pattern and its conversion by testosterone in the temporal muscle of the guinea-pig. Physiol Bohemoslov 20: 423–431. Bodine, S.C., A. Garfinkel, R.R. Roy, V.R. Edgerton (1988) Spatial distribution of motor unit fibers in the cat soleus and tibialis anterior muscles: Local interactions. J Neurosci 8: 2142–2152. Bottinelli, R., R. Betto, S. Schiaffino, C. Reggiani (1994a) Maximum shortening velocity and coexistence of myosin heavy chain isoforms in single skinned fast fibers of rat skeletal muscle. J Muscle Res Cell Motil 15: 413–419. Bottinelli, R., R. Betto, S. Schiaffino, C. Reggiani (1994b) Unloaded shortening velocity and myosin heavy chain and alkali light chain isoform composition in rat skeletal muscle fibres. J Physiol (Lond) 478: 341–349. Bottinelli, R., S. Schiaffino, C. Reggiani (1991) Force-velocity relations and myosin heavy chain isoform compositions of skinned fibres from rat skeletal muscle. J Physiol (Lond) 437: 655–672. Bredman, J.J., W.A. Weijs, H.A. Korfage, P. Brugman, A.F. Moorman (1992) Myosin heavy chain expression in rabbit masseter muscle during postnatal development. J Anat 180: 263–274. d’Albis, A., R. Couteaux, C. Janmot, J.C. Mira (1993) Opposite regulations by androgenic and thyroid hormones of V1 myosin expression in the two types of rabbit striated muscle: Skeletal and cardiac. FEBS Lett 318: 53–56. d’Albis, A., C. Tobin, C. Janmot, R. Couteaux (1992) Effect of testosterone and thyroid hormone on the expression of myosin in the sexually dimorphic levator ani muscle of rat. J Biol Chem 267: 10052–10054. Eason, J., G. Schwartz, G. Pavlath, A. English (2000a) Sexually dimorphic expression of myosin heavy chain isoforms in adult mouse masseter muscles. J Appl Physiol 89: 251–258. Eason, J.M., G. Schwartz, K.A. Shirley, A.W. English (2000b) Investigation of sexual dimorphism in the rabbit masseter muscle showing different effects of androgen deprivation in adult and young adult animals. Arch Oral Biol 45: 683–690. English, A., D. Carrasco, C. Widmer (1999a) Torques produced by different compartments of the rabbit masseter muscle. J Appl Biomech 15: 347–359. English, A., J. Eason, G. Schwartz, A. Shirley (1999b) Sexual dimorphism in the rabbit masseter muscle: Myosin heavy chain composition of neuromuscular compartments. Cells Tissues Organs 164: 179–191. English, A.W., J. Eason, M. Pol, G. Schwartz, A. Shirley (1998) Different phenotypes among slow/beta myosin heavy chain-containing fibers of rabbit masseter muscle: A novel type of diversity in adult muscle. J Muscle Res Cell Mol 19: 525–535.
96
English, A.W., W.D. Letbetter (1982) Anatomy and innervation patterns of cat lateral gastrocnemius and plantaris muscles. Am J Anat 164: 67– 77. English, A.W., G. Schwartz (2002) Development of sex differences in the rabbit masseter muscle is not restricted to a critical period. J Appl Physiol 92: 1214–1222. English, A.W., C.G. Widmer (2003) Sex differences in rabbit masseter motoneuron firing behavior. J Neurobiol, in press. Gillespie, M.J., T. Gordon, P.R. Murphy (1986) Reinnervation of the lateral gastrocnemius and soleus muscles in the rat by their common nerve. J Physiol (Lond) 372: 485–500. Gutmann, E., V. Hanzlikova, Z. Lojda (1970) Effect of androgens on histochemical fibre type. Differentiation in the temporal muscle of the guinea pig. Histochemie 24: 287–291. Hultman, E., H. Sjoholm, I. Jaderholm-Ek, J. Krynicki (1983) Evaluation of methods for electrical stimulation of human skeletal muscle in situ. Pflügers Arch 398: 139–141. Kwa, S.H., J.A. Korfage, W.A. Weijs (1995a) Function-dependent anatomical parameters of rabbit masseter motor units. J Dent Res 74: 1649– 1657. Kwa, S.H.S., W.A. Weijs, P.J.W. Jüch (1995b) Contraction characteristics and myosin heavy chain composition of rabbit masseter motor units. J Neurophysiol 73: 538–549. Liu, Z.J., Y. Masuda, T. Inoue, H. Fuchihata, A. Sumida, K. Takada, T. Morimoto (1993) Coordination of cortically induced rhythmic jaw and tongue movements in the rabbit. J Neurophysiol 69: 569–584. Lyons, G.E., A.M. Kelly, N.A. Rubinstein (1986) Testosterone-induced changes in contractile protein isoforms in the sexually dimorphic temporalis muscle of the guinea pig. J Biol Chem 261: 13278–13284. Maxwell, L.C., D.C. Carlson, J.A. McNamara, J.A. Faulkner (1979) Histochemical characteristics of the masseter and temporalis muscles of the rhesus monkey (Macaca mulatta). Anat Rec 193: 389–402. Moss, R.L., G.M. Diffee, M.L. Greaser (1995) Contractile properties of skeletal muscle fibers in relation to myofibrillar protein isoforms. Rev Physiol Biochem Pharmacol 126: 1–63. Popovic, D., T. Gordon, V.F. Rafuse, A. Prochazka (1991) Properties of implanted electrodes for functional electrical stimulation. Ann Biomed Eng 19: 303–316. Reader, M., G. Schwartz, A.W. English (2001) Brief exposure to testosterone is sufficient to induce sex differences in the rabbit masseter muscle. Cells Tissues Organs 169: 210–217.
Cells Tissues Organs 2003;174:87–96
Rhee, E.K., A.P. Furnary, J.J. Elson, R.L. Kao (1992) Correlation of electrophysiological activation patterns to tension generation in stimulated latissimus dorsi muscle. Pacing Clin Electrophysiol 15: 1730–1739. Rome, L.C., A.A. Sosnicki, D.O. Goble (1990) Maximum velocity of shortening of three fibre types from horse soleus muscle: Implications for scaling with body size. J Physiol (Lond) 431: 173–185. Schwartz, G., S. Enomoto, C. Valiquette, J. Lund (1989) Mastication in the rabbit: A description of movement and muscle activity. J Neurophysiol 62: 273–287. Sciote, J.J., J.C. Kentish (1996) Unloaded shortening velocities of rabbit masseter muscle fibres expressing skeletal or alpha-cardiac myosin heavy chains. J Physiol (Lond) 492: 659–667. Sciote, J.J., T.J. Morris, C.A. Brandon, M.J. Horton, C. Rosen (2002) Unloaded shortening velocity and myosin heavy chain variations in human laryngeal muscle fibers. Ann Otol Rhinol Laryngol 111: 120–127. Sokoloff, A.J., S.G. Siegel, T.C. Cope (1999) Recruitment order among motoneurons from different motor nuclei. J Neurophysiol 81: 2485– 2492. Thompson, T.L., W.E. Berndtson (1993) Testicular weight, Sertoli cell number, daily sperm production, and sperm output of sexually mature rabbits after neonatal or prepubertal hemicastration. Biol Reprod 48: 952–957. Totosy de Zepetnek, J.E., H.V. Zung, S. Erdebil, T. Gordon (1992) Innervation ratio is an important determinant of force in normal and reinnervated rat tibialis anterior muscles. J Neurophysiol 67: 1385–1403. Turkawski, S.J., T.M. van Eijden (2001) Mechanical properties of single motor units in the rabbit masseter muscle as a function of jaw position. Exp Brain Res 138: 153–162. van Eijden, T.M., S.J. Turkawski (2001) Morphology and physiology of masticatory muscle motor units. Crit Rev Oral Biol Med 12: 76–91. Weijs, W., R. Dantuma (1981) Functional anatomy of the masticatory apparatus in the rabbit (Oryctylagus cuniculus, L.). Neth J Zool 31: 99– 147. Weijs, W.A., P. Brugman, C.A. Grimbergen (1989) Jaw movements and muscle activity during mastication in growing rabbits. Anat Rec 224: 407–416. Weijs, W.A., P.J. Juch, S.H. Kwa, J.A. Korfage (1993) Motor unit territories and fiber types in rabbit masseter muscle. J Dent Res 72: 1491– 1498. Widmer, C., D. Klugman, A. English (1997) Anatomical partitioning and nerve branch patterns in the adult rabbit masseter. Acta Anat 159: 222–232. Widmer, C.G., D.I. Carrasco, A.W. English (2003) Differential participation by rabbit masseter compartments in different oral behaviors. Exp Brain Res, in press.
English/Widmer
Author Index Vol. 174, No. 1–2, 2003
Balazs, E.A. 49 Chan, J. 26 Cheung, H.S. 63 Dionne, R. 26 English, A.W. 5, 87 Gregg, J. 17 Gruber, H.E. 17 Haqqi, T.M. 34 Horton, M.J. 73 Islam, N. 34 Link, J. 73 McCartney-Francis, N. 26
ABC Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2003 S. Karger AG, Basel
Accessible online at: www.karger.com/cto
Malemud, C.J. 34 Nitzan, D.W. 6 Orenstein, J.M. 26 Rowlerson, A.M. 73 Sciote, J.J. 73 Sun, Y. 63 Ta, L. 26 Wahl, S.M. 26 Wenger, L. 63 Widmer, C.G. 87 Zeng, X.R. 63
97
Subject Index Vol. 174, No. 1–2, 2003
Adhesive forces 6 Anchored disc phenomenon 6 Antiresorbing agents 17 Apoptosis 34 Arthritic pain 49 Arthritis 26 Basic calcium phosphate crystals 63 Bone resorption 17 Chondrocytes, human 34 Cytokine 34 Early growth response gene 63 Fiber types 73 Friction 6 Gene expression 63 Hyaluronan 49 Hylan 49 Inflammation 26 Internal derangement 6 Masseter 87 Motoneurons 87
ABC Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
© 2003 S. Karger AG, Basel
Accessible online at: www.karger.com/cto
Muscle 73 Myosin 73 Nitric oxide 26 Open lock 6 Osteoarthritis 6, 34 Osteoclast 17 Osteoporosis 17 Reverse transcription-polymerase chain reaction 73 Sexual dimorphism 87 Signaling pathways 34 Temporomandibular 87 – joint 6, 17, 49 – – disease 17 – – disorder 26 Testosterone 87 Transcriptional factors 63 Tumor necrosis factor 26 Viscosupplementation 49
Conference Calendar
19.6.–21.6.2003 Washington, D.C. USA
American Society for Artificial Internal Organs
Information: www.asaio.org/
28.6.–1.7.2003 Mystic, Conn. USA
XXVth Annual Meeting of the International Society for Heart Research (ISHR) (North American Section)
Information: ishr2003.uchc.edu
Regenerative Medicine for Cardiovascular Disease Information: FEPS 2003 Organizing Secretariat, Nice Acropolis, Congress Management Department, 1 espl Kennedy, BP 4083, 06302 Nice Cedex 4, France Tel. +33 (0)4 9392 8159/61, Fax +33 (0)4 9392 8338 E-Mail
[email protected]
28.6.–3.7.2003 Nice France
3rd FEPS Congress
30.6.–2.7.2003 Davos Switzerland
ECM IV: European Cells and Materials Conferences
20.7.–25.7.2003 Plymouth, N.H. USA
Biomaterials: Biocompatibility/Tissue Engineering
Information: E-Mail
[email protected]
13.8.–16.8.2003 Houston, Tex. USA
Advances in Tissue Engineering 2003
Information: E-Mail
[email protected]
20.8.–22.8.2003 Sydney Australia
6th International Conference on Cellular Engineering
Information: Laura Pool-Warren E-Mail
[email protected] www.celleng2003.org.au
20.8.–23.8.2003 Pittsburgh, Pa. USA
Pittsburgh Bone Symposium 2003
Information: Tricina Cash, Conference Planner, UPMC Health System, Center for Continuing Education in the Health Sciences, Medical Arts Building, Suite 220, 200 Lothrop Street, Pittsburgh, Pa., USA E-Mail
[email protected]
24.8.–29.8.2003 Basel Switzerland
11th European Congress on Biotechnology
Information: ECB11, c/o AKM Congress Service, P.O. Box, CH–4005 Basel, Switzerland Tel. +41 61 686 77 11, Fax +41 61 686 77 88 E-Mail
[email protected]
3.9.–6.9.2003 Genova Italy
Second Annual Meeting of the European Tissue Engineering Society
Information: Local Secretariat Office, Daniela Garbarino, Barbara Minuto, Istituto Nazionale per la Ricerca sul Cancro/CBA, Largo R. Benzi, 10, 16132 Genova, Italy Tel. +39 010 5737398, Fax +39 010 5737405 E-Mail
[email protected]
11.9.–14.9.2003 Rome Italy
5th International Malpighi Symposium
Information: http://www.anatomia.med.uniroma1.it/malpighi/ malpighi5.html
20.9.–24.9.2003 Dresden Germany
ELSO (European Life Sciences Organisation) 2003
Information: Ingeborg Fatscher, P.O. Box 1151, 69199 Sandhausen, Germany Tel. +49 6224 925613, Fax +49 6224 925610, 11, 12 E-Mail
[email protected]
21.9.–23.9.2003 Beverwijk The Netherlands
European Tissue Repair Society
Information: ETRS 2003, P.O. Box 1015, 1940 EA Beverwijk, The Netherlands
ABC
© 2003 S. Karger AG, Basel
Fax + 41 61 306 12 34 E-Mail
[email protected] www.karger.com
Federation of European Physiological Societies
Bone Tissue Engineering
Under the aegis of the Academia Malpighiana Studiorum Anatomicae Microscopicae Within the 7th Centennial of the Foundation of the University of Rome ‘La Sapienza’
13th Annual Meeting
Accessible online at: www.karger.com/cto
Information: www.ao-asif.ch/index.shtml
99
23.9.–25.9.2003 Chesapeake Bay, Md. USA
3rd Annual Somatic Cell Therapy Symposium: Regulatory Issues for Scientists & Clinicians
Information: Martha Davis or Lee Buckler, International Society for Cellular Therapy, 777 West Broadway, Suite 401, Vancouver, B.C. V5Z 4J7, Canada Tel. +1 604 874 4366, Fax +1 604 874 4378 E-Mail
[email protected], www.celltherapy.org
24.9.–27.9.2003 Mainz Germany
10th Conference of the European Placenta Group (EPG)
Information: General Organisation: Dr. Rudy Leiser, Dr. Christiane Pfarrer and Carola Lang, DVM: Dept of Veterinary Anatomy (EPG 2003), Frankfurter Strasse 98 D–35392 Giessen, Germany Tel: +49 641 99 38101, Fax: +49 641 99 38109 E-Mail
[email protected]
1.10.–4.10.2003 Tübingen Germany
18th European Conference on Biomaterials
Information: www.esb2003.org/
5.10.–8.10. 2003 Port Douglas Australia
International Meeting on Epithelio-Mesenchymal Transitions Cancer – Development – Pathology
Information: www.magicdatabases.com/emt.html
12.10.–15.10.2003 Berlin Germany
OARSI’s 2003 World Congress on Osteoarthritis
Information: www.oarsi.org/
12.10.–17.10.2003 Shenzhen P.R. China
China Hi-Tech Fair / Biotech 2003
Information: Mr. Deric Liu, Project Manager, Coastal International Exhibition Co. Ltd., Rm. 3808, China Resources Bldg., 26 Harbour Rd., Wanchai, Hong Kong Tel: (852) 2827 6766, Fax: (852) 2827 6870 E-Mail
[email protected] Web site: www.biotech-exhibition.com/ www.coastal.com.hk
17.10.–22.10.2003 Obernai France
Molecular Biology of Cellular Interactions
Information: Scientific programme and on-line application at: www.esf.org/euresco/03/lc03066 For printed copies contact: Dr. J. Hendekovic, European Science Foundation, EURESCO Unit, 1 quai Lezay-Marnésia, 67080 Strasbourg Cedex, France Tel. +33 388 76 71 35, Fax +33 388 36 69 87 E-Mail
[email protected]
19.10.–21.10.2003 New Orleans, La. USA
3rd Annual Nonhematopoietic & Mesenchymal Stem Cells Conference
Information: Martha Davis or Lee Buckler, International Society for Cellular Therapy, 777 West Broadway, Suite 401, Vancouver, B.C. V5Z 4J7, Canada Tel. +1 604 874 4366, Fax +1 604 874 4378 E-Mail
[email protected], www.celltherapy.org
22.8.–27.8.2004 Kyoto Japan
16th International Congress of the International Federation of Associations of Anatomists (IFAA) Anatomical Science 2004 From Gene to Body
Information: 16th International Congress of the IFAA, c/o JTB Communications, Inc., Sankei Bldg., 7F, Umeda 2-4-9, Kita-ku, Osaka 530-0001, Japan Tel. +81 6 6348 1391, Fax +81 6 6456 4105 E-Mail
[email protected], www.ifaa2004.org
100
Commercialisation of Tissue Engineering & Regenerative Medicine
EuroConference on Adhesion Receptors during Development and Pathologies
Conference Calendar