Keren Keinan-Adamsky, Hadassah Shinar, Gil Navon * Received 26 May 2004

Transcription

1 ELSEVIER Journal of Orthopaedic Research 23 (2005) Journal of Orthopaedic Research The effect of detachment of the articular cartilage from its calcified zone on the cartilage microstructure, assessed by 2H-spectroscopic double quantum filtered MRI Keren Keinan-Adamsky, Hadassah Shinar, Gil Navon * School of Chemistry, Tel Aviv University, Ramut Aviv, Tel-Aviv 69978, Isruel Received 26 May 2004 Abstract Most studies on articular cartilage properties have been conducted after detachment of the cartilage from the bone. In the present work we investigated the effect of detachment on collagen fiber architecture. We used one-dimensional 2H double quantum filtered MRI on cartilage bone plugs equilibrated in deuterated saline. The quadrupolar splittings observed in the different zones were related to the degree of order and the density of the collagen fibers. The method is non-destructive, allowing for measurements on the same plug without the need for fixation, dehydration, sectioning and decalcification. Detachment of the radial from the calcified zone resulted in swelling of the cartilage plug in physiological saline and a concomitant decrease in the quadrupolar splitting. The effect of mechanical pressure on the 'H quadrupolar splittings for the detached cartilage and for the calcified zone-bone plugs were compared with those of the same zones in the intact cartilage-bone plug. The splitting in the radial zone of the detached cartilage collapsed at much smaller loads compared to the intact cartilage-bone plug. The effect of the load on the size of the cartilage was also greater for the detached plug. These results indicate that anchoring of the cartilage to the bone through the calcified zone plays an important role in retaining the order of the collagen fibers. The water 2H quadrupolar splitting in intact and proteoglycan-depleted cartilage was the same, indicating that the proteoglycans do not contribute to the ordering of the collagen fibers Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. Keywords: Articular cartilage; DQF spectroscopic MRI; *H quadrupolar splitting Introduction The exceptional properties of cartilage that allow for load bearing and the ability to withstand recurrent pressure are provided by the composition and structure of the tissue: the integrity of the ordered collagen fibers and the osmotic properties of the proteoglycans (PG). A great amount of work has recently been devoted to studying the influence of load on healthy and impaired * Corresponding author. Tel.: ; fax: address: navon@post.tau.ac.il (G. Navon). cartilage matrix [9,10,16,17,26,34,37]. For many of these studies, cartilage plugs were detached from the bone. Some studies concentrated on the influence of PG depletion on proton magnetic resonance parameters, such as T,, T1 and T1, relaxation times, diffusion constant, and magnetization transfer contrast [6,41]. The relaxation rates were also studied as a function of osmotic pressure [2 11. In other studies, biomechanical properties were determined in unconfined compression [29,33]. Since the cartilage and the subchondral region are one functional unit, cartilage properties might be altered by detachment of the cartilage from the bone. The calcified zone is important in normal cartilage function and for cartilage transplantation [14,3 1,32,49] /$ - see front matter Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. doi: /j.orthres

2 110 K. Keinan-Adumsky et al. I Journul of Orthopaedic Research 23 (2005) Articular cartilage can be divided into zones, which differ in the relative amounts of cartilage constituents and in the orientation of the collagen fibers. The collagen fibers are anchored to the underlying bone, which has been shown to greatly support the collagen fiber matrix [4,14]. They rise vertically from the bone and keep the same orientation throughout the calcified and radial zones [ 15,16,30]. In the intermediate zone, the fibers bend and then align parallel to the surface. The thickness of the collagen fibers is largest near the bone and decreases continuously towards the surface [4]. It has long been known that water molecules, interacting with collagen fibers, tend to align along the fibers and that their motion is anisotropic [1,2,22]. Thus the dipolar interaction between the two water protons has a non-zero average, rendering its TZ orientation dependent. Indeed, the T2 weighted images of articular cartilage [ 12,36,45] exhibit several lamina corresponding to the different histological zones of articular cartilage [5,8, 1 1,19,20,24,4244,47]. Moreover, these images depend on the cartilage orientation relative to the magnetic field. We have recently developed an MRI method that monitors the density and the orientation of the collagen fibers throughout the different cartilage zones [27,39]. Unlike electron microscopy and histology, the method is non-destructive, and the same sample can be monitored following different treatments. The method is based on the NMR properties of the quadrupolar I = 1 nucleus, 2H, and is minimally invasive, only requiring the equilibration of the specimen in deuterated saline. In the present work, we used this method to investigate the effect of the cartilage detachment from the calcified zone, as well as PG depletion on the matrix architecture both at rest and under applied load. Theoretical background In isotropic solutions, fast isotropic tumbling of water molecules averages out the dipolar interaction between their protons and the quadrupolar interaction of the deuterons. Anisotropic molecular tumbling results in a non-zero residual dipolar or quadrupolar interactions. In such cases, for two interacting I= 112 nuclei (e.g., IH) and for the I = 1 nucleus (e.g., 2H), two transitions are observed. This phenomenon was observed for both 'H and 2H of water molecules interacting with the oriented collagen fibers [1,2,22]. 'H is undoubtedly the nucleus of choice for NMR measurements due to its high sensitivity and abundance. Recently the reconstruction of the order of the collagen fibers in regenerating tendon was evaluated by 'H DQF NMR spectroscopy [13]. However, 'H has a major disadvantage for the detection of order in biological tissues. The proton exchange between water molecules leads to spin exchange, modulating the dipolar splitting and causing it to collapse at room and body temperatures [7]. We thus chose to perform our measurements on *H for cartilage plugs equilibrated in deuterated saline. The 2H splitting is unaffected by the exchange of deuterons between water molecules since the residual quadrupolar splitting of the 2H nuclei is about the same for neighboring water molecules. Thus the quadrupolar splitting is clearly resolved. In cartilage the director of the quadrupolar interaction is determined by the direction of the collagen fibers [27]. The frequency separation between the two transitions, tiexp, depends on: the size of the quadrupolar interaction for those water molecules that are in close contact with the collagen fibers and are denoted as "bound" (vh); on the angle between the collagen fibers and the magnetic field (6); and on the ratio between the bound and the free water molecules (Ph), according to the following relation: v,,p = vbpb(3cos28-1)/2 (1) Thus the splitting depends on both the order of the coilagen fibers and the degree of hydration, which is related to the packing density. Measurement of the splitting at various orientations of the sample relative to the magnetic field allows determination of the degree of order of the fibers. For instance, in cases where the fibers are perfectly ordered, the splitting obtained with the fibers parallel to the field is twice that of the splitting obtained in the perpendicular orientation. At 54.7", the magic angle, no splitting should be observed. In biological tissues, when the amount of water molecules interacting with the anisotropic sites is small or when the splitting is small relative to the linewidth, the split lines are masked by the large signal of isotropic water. This problem is overcome by the use of DQF NMR techniques. With these techniques, only those 2H nuclei that exhibit some splitting as a result of anisotropic re-orientation are detected. The large isotropic signal of the bulk water molecules is filtered out allowing a clear observation of the split satellites. 2H DQF technique is based on the following pulse sequence: 9Oo-z/ z/2-900-t~~-900-acq, where z is the creation time of the second rank tensors, and tdq is the DQ evolution time. By proper phase cycling only double quantum coherences are detected, thus allowing for filtering of the isotropic signal. Since the DQ relaxation time in these systems is relatively long [38,40], the phase encoding gradients are employed during tdq. Moreover, the effect of the gradients is doubled during the DQ evolution time, partially compensating for the low y of 2H. This technique, which enables observation of the splitting variation within the cartilage plug from bone to surface, gives the spatial distribution of the deuterium residual quadrupolar interaction and thus maps the order of the collagen fibers.

3 Materials and methods Cartilage-bone plugs, 7-8 mm in diameter, were obtained from bovine femoral lateral and medial condyles. The samples were equilibrated in deuterated saline for 12 h. For NMR measurements the samples were wiped dry and immersed in fluorinated oil (Fluorinert, FC-77, 3M), which has low water solubility and the same susceptibility as water. The plugs were positioned in a 10 nim NMR tube so that the surface ofthe cartilage was perpendicular to the magnetic field. Each cartilage-bone plug was placed between two pieces of Teflon, which exerted no pressure and provided stabilization. Imaging and spectroscopic measurements were conducted on a Bruker DMX-360WB NMR spectrometer using a 10 mm imaging probe (200 Gtcm) tuned for deuterium (55.3 MHz). The MRI parameters that were used for the one-dimensional 'H DQF spectroscopic imaging were: field of view 0.7 cm; data matrix 64 x 4096; 7 = 100 ps; and image resolution = 109 pm. The 2D FT was performed after zero filling of data to I28 x A homebuilt pressure device was used for applying mechanical pressure (Fig. I). The device was composed of a heavy wall NMR tube (513-7TRM-7, Wilmad Buena, NJ), held in a special Teflon holder. and a long glass rod with a brass disk on top which supported a series of weights. The samples were placed in the NMR tube with the cartilage surface on top of a sintered glass filter. Pressure was applied from the bone side by the weights. After the application of each load, 30 min were allowed to reach a new equilibrium. This device is an improvement of the former non-magnetic calibrated springs that were used to apply pressure [39]. To evaluate the stabilizing effect of the bone on the collagen fiber orientation, cartilage was separated from the bone by cutting the plug above the calcified zone, where the plug was soft and cutting was feasible. To induce in vitro PG depletion as a model for OA, plugs were treated with 4.0 mg/ml trypsin in saline (ph = 7.4) at room temperature for 12 h with constant agitation. The samples were then re-equilibrated in deuterated saline before measurements were repeated. '3Na quadrupolar-echo spectroscopic imaging measurements before and after degradation with trypsin have shown an increase of the splitting in cartilage plugs, an indication for proteoglycans extraction [ 181. Results We found a large quadrupolar splitting near the bone and a smaller one in the radial zone of all cartilage-bone plugs (Fig. 2a), where the spectral and the spatial dimensions are represented by the X and Y axes respectively. No 'H DQF signal could be detected from the bone [39]. After detachment of the cartilage from the bone, Fig. 1. A homebuilt device. which was used for mechanical compression, inside the NMR spectrometer. Fig. 2. One-dimensional 'H DQF spectroscopic images of: (a) bovine articular cartilage-bone plug, equilibrated in deuterated saline; (b) the detached cartilage after separation of the radial zone from the bone above the calcified zone; (c) the bone with the calcified zone-bone plug; and (d) the detached cartilage after re-equilibration in deuterated saline.

4 I12 K. Keinun-Adamsky et ul. I Journal of Orthopedic Reseurcli 23 (2005) above the calcified zone, the splitting in the cartilage plug (Fig. 2b) was reduced while that of the calcified zone-bone plug did not change (Fig. 2c). The cartilage plug and the calcified zone-bone plug were both re-equilibrated for a few hours in deuterated saline after which the 2H spectroscopic images were measured again. The splitting in the radial zone of the cartilage plug was further reduced (Fig. 2d), while in the calcified zone-bone plug no change in splitting was observed. The splittings (Fig. 2a-d) were plotted as a function of the distance from the surface (Fig. 3). Comparison of the linewidth at half height with the splitting throughout the plug yielded a good linear relationship. Thus we used this relation to estimate the splittings when they were not clearly resolved. The splitting in the intact plug was largest in the calcified zone and gradually decreased moving through the radial zone. The splitting of the calcified zone-bone plug exactly coincided with the splitting of the same area on the intact plug. The splitting of the cartilage plug, in all locations was somewhat smaller than in the intact plug and was further reduced after re-equilibration. Swelling of the plug as a result of detachment (h + c > a) and re-equilibration (d > 6) was also seen. The experiment (described in Figs. 2 and 3) was repeated on 13 different plugs taken from different cows (Table 1). The average values obtained were _ 240 and 420 f 80 Hz for the splitting in the calcified and in the radial zones of the intact cartilage-bone plug, respectively. After detachment, the splitting in the calcified zone, which was attached to the bone, did not change (1 350 f 240 Hz). The average value obtained for the splitting in the detached cartilage plug was 390 k 40 Hz and decreased to 230 k 70 Hz after re-equilibration of the plug in deuterated saline. On average, N g g C 800- Table 1 Average splittings and widths at the calcified and radial zones of intact cartilage-bone plugs before and after detachment and after reequilibration (n = 13) Samples vo (Hz) f sd Width (%) If. sd Intact cartilage-bone plug Calcified zone ( 100) Radial zone 420 -t 80 Detached cartilage f 8 Detached and re-equilibrated t 16 cartilage Calcified zone _ _ 16 The splitting are averages of slices located at the center of each zone. The widths are relative to that of the intact cartilage-bone plugs (taken as 100%) for each of the 13 samples. the width of the calcified zone was 34% that of the intact plug, The detached cartilage swelled by approximately 24%, and after the re-equilibration' the swelling was 51%. The size of the calcified zone-bone plug was not changed upon re-equilibration. To examine whether collagen fiber orientation changed after the cartilage was removed from the bone, the quadrupolar splitting of the detached and re-equilibrated cartilage was measured with the cartilage surface parallel and perpendicular to the magnetic field (Fig. 4). For the intact cartilage-bone plug, the splitting in the calcified and radial zones was approximately half when the surface of the plug was parallel relative to when the surface of the plug was perpendicular to the field, indicating that the fibers in these zones were perpendicular to the surface. In the intermediate zone the splitting almost vanished as expected since in the transition from perpendicular to tangential to the surface the fibers ori E - E 800- rn 5.- C : '0 0'5 1.o 1: Distance from the surface (mm) I, I. I ~ I. I ' I. I Distance from the surface (mm) Fig. 3. The 'H quadrupolar splittings of the sample shown in Fig. 2 as a function of the distance from the surface: (a) intact cartilage-bone plug (W); (b) the cartilage after detachment (0);(c) the calcified zone after detachment (0); and (d) detached cartilage after re-equilibration with deuterated saline (A). Near the surface (up to 0.2 mm for intact cartilage-bone plug and 0.8 mm for detached samples), the splitting was estimated from the line width. Fig. 4. 'H quadrupolar splittings of bovine articular cartilage-bone plug equilibrated in deuterated saline and of the same cartilage after detachment and re-equilibration. The quadrupolar splitting is given as a function of the distance from the cartilage surface, at two orientations of the plug relative to the magnetic field. The images were obtained with the cartilage surface parallel (W, A) and perpendicular (0, A) for the intact (W, 0) and re-equilibrated detached cartilage (A, A). Near the surface (up to 0.2 mm for intact cartilagebone plug and 0.6 mm for detached samples), the splitting was estimated from the line width.

5 K. Keinan-Adurnsky et a1. I Journal of Orthopuedic Research 23 (2005) entation goes through the magic angle. After detachment and re-equilibration, the splitting was greatly reduced, but the factor of 2 difference between the splittings of the radial zone at the two orientations was preserved, indicating that in the radial zone of the detached cartilage, the fibers remained perpendicular to the surface. Thus, collagen fiber orientation did not change after detachment, and the changes in the a Fig. 5. (a) 2H DQF spectroscopic images of bovine articular cartilage-bone plug, equilibrated in deuterated saline under stepwise application of 0, 0.13, 0.4, and 0.97 MPa. At the end of the experiment, the plug was equilibrated and measured again. The cartilage was then detached, reequilibrated in deuterated saline, and measured under stepwise application of the same pressures. After measurements the detached cartilage was reequilibrated and measured again. The bone with the calcified zone was also measured at 0 and 0.97 MPa. Images and extracted spectra of the cartilage-bone plug, re-equilibrated detached cartilage and calcified zone at 0 (b-d) and (e-g) at 0.97 MPa.

6 114 K. Keinun-Acluinsky et ul. I J~IL~)LI).IIU~ of Orthopuerlic Rrseurcli 23 (2005) quadrupolar splittings were caused by changes in the density of the fibers. The effects of pressure on the 'H quadrupolar splitting were studied for cartilage-bone plugs as well as for the re-equilibrated detached cartilage and for the calcified zone-bone plugs. One-dimensional 'H DQF spectroscopic images were obtained under stepwise application of pressure up to 0.97 MPa (Fig. 5). Before the application of load, large splitting in the calcified zone, small splitting in the radial zone, and much smaller splitting in the surface were observed (Fig. 5a and b). Upon applying 0.13 MPa, a large decrease of the splitting in the radial zone and an increase of the splitting in the surface were observed. After increasing the pressure to 0.97 MPa, the plug became considerably thinner, and the splitting in the calcified zone almost vanished (from 1650 Hz to approximately I00 Hz), while the splitting in the surface considerably increased from 90 to 900 Hz (Fig. 5a and e). At the same time a thickening of the surface zone was observed. The changes in the magnitude of the quadrupolar splitting indicate the disruption of the order in the calcified zone and the buildup of order at the surface. The loss of water may also contribute to the large increase of the splitting at the surface. After releasing the pressure and re-equilibrating, the DQF spectroscopic image was the same as the one obtained before the application of the pressure, indicating that the original collagen fiber orientation was fully restored [44]. Compared to the cartilage-bone plug, pressure had a greater effect on the detached cartilage: at 0.13MPa, the size of the cartilage in the cartilage-bone plugs was 86 k 9% (n = 4) of its original size, while for the detached cartilage it was 72 k 5%) (for the same 4 plugs) after re-equilibration. The effect of pressure on the quadrupolar splitting was also larger for the detached cartilage. The quadrupolar splitting of the radial zone of the detached cartilage at 0.13 MPa was only 11 k 19% (n = 3) compared to the unloaded detached cartilage. For the cartilage-bone plug, the splitting at 0.13 MPa was reduced to 82 f 13% (n = 4). Again, at 0.97 MPa the splitting at the surface, which was not resolved in the absence of load, increased to approximately 350 Hz, and the surface zone comprised a larger fraction of the cartilage depth. After release of the pressure and re-equilibration, the detached cartilage returned to its original size. The image of the calcified zone-bone plug displayed a 1600 Hz splitting, the same splitting that was observed near the bone in the cartilage-bone plug. This splitting did not change after the application of the 0.97 MPa pressure (Fig. 5a, d, and g). To test whether the order parameter of the collagen fibers changed following PG depletion, 'H DQF spectroscopic images were obtained for the same plug before and after degradation with trypsin (Fig. 6). Extracted spectra from the calcified and radial zones demonstrated that the splitting was unaffected by the degradation (Fig. 6a and b). Moreover, the results obtained at two orientations of the plug relative to the field were the same as for the intact and depleted plugs. These results indicate that the collagen matrix was intact after this procedure and that PG did not contribute to the ordering of the water molecules. By applying pressure to the depleted Fig. 6. 'H DQF spectroscopic images and extracted spectra of bovine articular cartilageebone plug before (a) and after (b) degradation with trypsin. The depleted cartilage was compressed by 0.13 (c) and 0.5 (d) MPa.

7 K. Keinan-Adamsky er al. I Journal of Orthopaedic Research ) cartilage, significant modifications occurred: after application 0.13 MPa, the plug thickness decreased by 500/0 and a very large splitting was observed at the surface (Fig. 6c). The thickness decrease of the depleted plug at 0.13 MPa was larger than the decrease at 0.97 MPa for the intact plug (Figs. 5a, e, and 6c). The depleted plug became thinner at 0.5 MPa, and the large splitting observed at the surface increased further. The image of the plug at this pressure was almost the same as observed for intact cartilage under 0.97 MPa (Figs. 5a, e, and 6d), an indication of a larger pressure effect on depleted cartilage, which reflects the loss of PG from the matrix. Discussion We studied the importance of the anchoring of the collagen fibers to the calcified zone in maintaining the integrity of the articular cartilage matrix. We found that the quadrupolar splitting in the radial zone decreased after disconnecting the radial from the calcified zone. The decrease was small right after detachment, but was very significant after re-equilibration in deuterated saline. This result is consistent with the 30% increase of the proton T2 observed by Xia [46] for excised cartilage relative to cartilage-bone, as the proton T2 depends on the residual proton-proton dipolar interaction [7], which has the same dependence on the fraction of bound water molecules as the 2H residual quadrupolar interaction. We also studied the orientation of the collagen fibers by performing the 2H DQF spectroscopic imaging at two orientations of the plug relative to the magnetic field, before and after detachment. Our results clearly indicate that the orientation and the degree of order of the collagen fibers did not change in response to the detachment and that the fibers remained perpendicular to the bone in the radial zone. Since after detachment and re-equilibration the splitting was greatly reduced and no change occurred in the degree of order, we concluded that the reduction resulted from the decrease of the fraction of bound water. This is in agreement with the observed tissue swelling seen by the increase of about 30% in the thickness upon equilibrium with saline (Table 1). Our results are supported by electron microscopy studies [3], showing that the collagen framework was considerably less dense than in the intact plug when articular cartilage is separated from its junction with the calcified cartilage. We found that the 2H quadrupolar splitting in the calcified zone, which remained attached to the bone, was changed by neither the detachment procedure nor the re-equilibration after detachment. This is in contrast with the effect of the detachment on the splitting of the cartilage plug. We can thus conclude that the attach- ment of the radial zone to the calcified zone has a stabilizing effect on the collagen fibers in the radial zone. In PG depleted cartilageebone plugs, we found that collagen fiber orientation remained the same as in the intact plugs. Moreover, the depletion did not affect the quadrupolar splitting. These results agree with the scanning electron microscopy results of Clark et al. [4], who found that the appearance of individual collagen fibrils did not change upon enzymatic digestion. Also reports that T2 in articular cartilage is insensitive to the amount of PG lead to the same conclusions [23,25,28]. Thus the hypothesis that the orientation dependence of T2 is caused by the anisotropic orientation of PG [lo] seems less likely. While proteoglycans have a major role in the ability of the matrix to resist compression, they do not contribute to the anisotropic motion of water molecules in their vicinity. When cartilage was subjected to an applied pressure, fluid was lost from the matrix, which lead to a decrease in tissue volume. Our 2H DQF spectroscopic images indicated a massive rearrangement of the polymer network [39]. The stabilizing effect of the attachment to the calcified zone on the ability of cartilage to withstand pressure was demonstrated by two of our results. First, the 0.13 MPa pressure had a greater effect on the thickness and splitting of the detached cartilage as compared to the cartilageebone plug at the same load. This is an indication that without the calcified zone, which supports the collagen fibers, the pressure effect on collagen fiber orientation was large. Second, the splitting in the calcified zone in the intact cartilage-bone plug gradually decreased and eventually vanished with increasing pressure, whereas the splitting in the calcified zone in the calcified zone-bone plug did not change even after the application of 0.97 MPa. This point is very intriguing. The marked difference in the response to pressure of the collagen fibers in the calcified zone before and after the disconnection from the radial zone leads to the conclusion that in intact cartilage the effect of pressure is transmitted through the radial zone to the calcified zone. A possible hypothesis is that in the cartilage-bone plug, in addition to the crimping of the collagen fibers in the radial zone in response to pressure, the collagen fibers penetrate into the calcified zone, destroying the order. The two gold standards for deducing collagen fiber architecture are scanning electron [4,15,16,30] and polarized light microscopy [48]. However, these methods involve fixation, decalcification, dehydration, and sectioning, and the measurement of the effects of pressure has to be performed on a series of plugs, each at a different pressure. Scanning transmission ion microscopy is another method that is used to investigate the collagen network [35]. Although fixation is not required, the technique still involves cryosection and dehydration. NMR, on the other hand, is a non-destructive method, and thus the whole experiment is conducted on one plug. At the

8 116 K Keinan-Adumsky et ul. I Journal of' Orthopaedic Research 23 (2005) end of the pressure experiment, the plug can be re-equilibrated, and the experiment can be repeated. Acknowledgement Support from the German Federal Ministry of Education and Research (BMBF) in the framework of German-Israeli Project Cooperation (DIP) is gratefully acknowledged. References [I] Berendson HJC. Nuclear magnetic resonance study of collagen hydration. J Chem Phys 1962;36: [2] Chapman GE, McLauchlan KA. The hydration structure of' collagen. Proc R Soc B 1971;178: [3] Chen MH, Broom ND. Concerning the ultrastructural origin of large-scale swelling in articular cartilage. J Anat 1999;194: [4] Clark JM. The organization of collagen in cryofractured rabbit articular cartilage: A scanning electron microscopy study. J Orthop Res 1985;3: [5] Cova M, Toffanin R, Frezza F, Pozzi-Mucelli M, Mlynarik V, Pozzi-Mucelli RS, et al. Magnetic resonance imaging of articular cartilage: ex-vivo study on normal cartilage correlated with magnetic resonance microscopy. Eur Radio1 1998;8: [6] Duvvuri U, Goldberg AD, Kranz JK, Hoang L, Reddy R, Wehrli FW, et al. Water magnetic relaxation dispersion in biological system: the contribution of proton exchange and implications for the noninvasive detection of cartilage degradation. Pnas 2001; 98(22): [7] Eliav U, Navon G. A study of dipolar interactions and dynamics processes of water molecules in tendon by 'H and 2H honionuclear and heteronuclear multiple-quantum-filtered NMR spectroscopy. J Magn Reson 1999;137: [8] Freeman DM, Bergman G, Glover G. Short TE MR microscopy: accurate measurement and zonal differentiation on normal hyelin cartilage. Magn Reson Med 1997;38: [9] Glaser C, Putz R. Functional anatomy of articular cartilage under compressive loading quantitative aspects of global, local and zonal reactions of the collagenous network with respect to he surface integrity. Osteoarthritis Cart 2002; 10: [lo] Griinder W, Kanowski M. Wagner M, Werner A. Visualization of pressure distribution within loaded joint cartilage by application of angle-sensitive NMR microscopy. Magn Reson Med 2000:43: [l I] Griinder W, Wagner M, Werner A. MR-microscopic visualization of anisotropic internal cartilage structures using magic angel techniques. Magn Reson Med 1998;39: [I21 Henkelman MR, Stanisz GJ, Kim JK, Bronskill MJ. Anisotropy of NMR properties of tissues. Magn Reson Med 1994;32: [I31 Ikoma K, Kusaka Y, Takamiya H, Eliav U, Navon G, Seo Y. Evaluation of collagen fiber maturation and ordering in regenerating tendons employing H-l double quantum filtered NMR spectroscopy. J Orthop Res 2003;21: [I41 Imhof H, Breitenseher M, Kainberger F, Rand T. Importance of subchondral bone to articular cartilage in health and disease. Top Magn Reson hag 1999;10: [I51 Jeffery AK, Blunn GW, Archer CW, Bently G. Three-dimensional collagen architecture in bovine articular cartilage. J Bone Joint Surg 1991;73-B: [I61 Kaab MJ, Ito K, Clark JM, Notzli HP. Deformation of articular cartilage collagen structure under static and cyclic loading. J Orthop Res 1998;16: [17] Kaufman JH, Regatte RR, Bolinger L, Kneeland JB, Reddy R, Leigh JS. A novel approach of observing articular cartilage deformation in vitro via magnetic resonance imaging. J Magn Reson hag 1999;9: [I81 Keinan-Adamsky K, Shinar H, Eliav U, Seo Y, Navon G. The effect of trypsin degradation on the 23Na spectroscopic MRI spectra of articular cartilage. In: Proc 10th Annual Meeting ISMRM, Honolulu, Hawaii, p. 66. [I91 Kim DJ, Suh JS, Jeong EK, Shin KH, Yang WI. Correlation of laminated MR appearance of articular cartilage with histology, ascertained by artificial landmarks on the cartilage. J Magn Reson hag 1999;10: [20] Lehner KB, Rechl HP, Gmeinwieser JK, Heuck AF, Lukas HP, Kohl HP. Structure, function, and degeneration of bovine hyaline cartilage: assessment with MR imaging in vitro. Radiology 1989;170: [21] Liisse S, Knauss R, Werner A, Griinder W, Arnold K. Action of compression and cations on the proton and deuterium relaxation in cartilage. Magn Reson Med 1995;33: [22] Migchelsen C, Berendson HJC. Proton exchange and molecular orientation of water in hydrated collagen fibers. An NMR study of H20 and D20. J Chem Phys 1973;59: [23] Mlynarik V, Tratting S, Huber M, Zembsch A, Imhof H. The role of relaxation times in monitoring proteoglycan depletion in articular cartilage. J Magn Reson hag 1999;10: [24] Mod1 JM, Sether LA, Huaghton VM, Kneeland JB. Articular cartilage: correlation of histologic zones with signal intensity at MR imaging. Radiology 1991;81: [25] Mosher TJ, Chen Q, Smith MB. 'H magnetic resonance spectroscopy of nanomelic chicken cartilage: effect of aggrecan depletion on cartilage T2. Osteoarthritis Cart 2003;11: and references cited therein. [26] Mow CM, Guo XE. Mechano-electrochemical properties of articular cartilage: their inhomogeneities and anisotropies. Ann Rev Biomed Eng 2002;4: and references cited therein. [27] Navon G, Shinar H, Eliav U, Seo Y. Multiple quantum filters and order in tissues. NMR Biomed 2001;14: [28] Nieminen MT, Toyras J, Rieppo J, Hakumaki JM, Silvennoinen J, Helminen HJ, et al. Quantitative MR microscopy of enzymatically degraded articular cartilage. Magn Reson Med 2000;43: [29] Nieminen MT, Toyras J, Laasanen MS, Silvennoinen J, Helminen HJ, Jurvelin JS. Prediction of biomechanical properties of articular cartilage with quantitative magnetic resonance imaging. J Biomech 2004;37: [30] Notzli H, Clark JM. Deformation of loaded articular cartilage prepared for scanning electron microscopy with rapid freezing and freeze-substitution fixation. J Orthop Res 1997;15: [3 I] Ogema TR, Thompson RC. Cartilage-bone interface (tidemark). In: Brandt K, editor. Cartilage changes on osteoarthritis. Ciba- Geiby, Basel: Indiana School of Medicine; p [32] Ogema TR, Thompson RC. Cartilage-bone interface (tidemark). The zone of calcified cartilage. Its role on osteoarthritis. Articular cartilage and osteoarthritis. New York: Raven Press; p [33] Park S, Hung CT, Ateshian GA. Mechanical response of bovine articular cartilage under dynamic unconfined compression loading at physiological stress levels. Osteoarthritis Cart 2004; 12: [34] Regatte RR, Kaufman JH, Noyszewski EA, Reddy R. Sodium and proton MR properties of cartilage during compression. J Magn Reson hag 1999;10: [35] Reinert T, Rebetanz U, Vogt J, Werner A, Grunder W. Visualization of collagen fibrils in joint cartilage using STIM. Nucl Instr and Meth B 2001;181: [36] Rubenstein JD, Kim JK, Morova-Protzner 1, Stanchez PL, Henkelman RM. Effect of collagen orientation on MR imaging

9 K. Krinan-Adurnsky et al. I Journal of Orthopardic Research 23 (2005) I7 117 characteristics of bovine articular cartilage. Radiology 1993; 188: [37] Rubenstein JD, Kim JK, Morova-Protzner I, Stanchez PL, Henkelman RM. Effects of compression and recovery on bovine articular cartilage: appearance on MR images. Radiology 1996; 21 0: [38] Sharf Y, Seo Y, Eliav U, Akselrod S, Navon G. Mapping of the strain exerted on blood vessel walls using deuterium filtered MRI. Proc Natl Acad Sci USA 1998;95: [39] Shinar H, Seo Y, Ikoma K, Kusaka Y, Eliav U, Navon G. Mapping the fiber orientation in articular cartilage at rest and under pressure studied by 2H double quantum filtered MRI. Magn Reson Med 2002;48: [40] Shinar H, Seo Y, Navon G. Discrimination between the different compartments in sciatic nerve by H double-quantum filtered NMR. J Magn Reson 1997;129: [41] Toffanin R, Mlynarik V, Russo S, Szomolanyi P, Piras A, Vittur F. Proteoglycan depletion and magnetic resonance parameters articular cartilage. Arch Biochem Biophys 2001;390(2): [42] Uhl M, Ihling C, Allmann KH, Lauberberger J, Tauer U, Adler CP, et al. Human articular cartilage: in vitro correlation of MRI and histologic findings. Eur Radio1 1998;8: [43] Xia Y. Heterogenity of cartilage laminae in MR imaging. J Magn Reson Imag 2000; I1: [44] Xia Y. Relaxation anisotropy in cartilage by NMR microscopy (pmri) at 14-pm resolution. Magn Reson Med 1998;39: [45] Xia Y, Farquhar T, Burton-Wurster N, Lust G. Origin of cartilage laminae in MRI. J Magn Reson Imag 1997;7: [46] Xia Y, Farquhar T, Burton-Wurster N, Ray E, Jelinski LW. Diffusion and relaxation mapping of cartilageebone plugs and excised disks microscopic magnetic resonance imaging. Magn Reson Med 1994;31: [47] Xia Y, Moody JB, Alhadlaq H. Orientational dependence of T2 relaxation in articular cartilage: a microscopic MRI (pmri) study. Magn Reson Med 2002;48:46&9. [48] Xia Y, Moody JB, Burton-Wurstert N, Lust G. Quantitative in situ correlation between microscopic MRI and polarized light microscopy studies of articular cartilage. Osteoarthritis Cart ;9: [49] Yu H, Grynpas M, Kandel RA. Composition of cartilagenous tissue with mineralized and non-mineralized zones formed in vitro. Biomaterials 1997; 18: