etadfjttami(Mot00gtttfitcsfcssi

a

fi

9

Journal of Structural Biology 136, 46–52 (2001)doi:10.1006/jsbi.2001.4417, available online at http://www.idealibrary.com on

1©A

Regional Structural and Viscoelastic Properties of Fibrocartilageupon Dynamic Nanoindentation of the Articular Condyle

Kai Hu, Priya Radhakrishnan, Rupal V. Patel, and Jeremy J. Mao1

Department of Orthodontics and Department of Bioengineering MC 841, University of Illinois at Chicago,801 South Paulina Street, Chicago, Illinois 60612

Received October 16, 2001

lEbEcc11bttm1FblssmvaisaimimcpyolPmptne1

Fibrocartilage, a tissue with macromaterial prop-rties between dense fibrous tissue and hyaline car-ilage, is not well understood in its ultrastructurend regional viscoelastic properties. Here nanoin-entation with atomic force microscopy was per-ormed on fresh fibrocartilage samples of rabbit jawoint condyles. Each sample was divided into an-eromedial, anterolateral, posteromedial, and pos-erolateral regions for probing and topographic im-ging in 2 3 2 mm and 10 3 10 mm scan sizes. Young’soduli differed significantly among these regions

n a descending gradient from the anteromedial2.34 6 0.26 MPa) to the posterolateral (0.95 6 0.06

Pa). The Poisson ratio, defined as lateral strainver axial strain, had the same gradient distribu-ion: highest for the anteromedial region (0.46 6.05) and lowest for the posterolateral region (0.31 6.05). The same four regions showed a descendingradient of surface roughness: highest for the an-eromedial (321.6 6 13.8 nm) and lowest for the pos-erolateral (155.6 6 12.6 nm). Thus, the regional ul-rastructural and viscoelastic properties ofbrocartilage appear to be coregulated. Based onhese region-specific gradient distributions, fibro-artilage is constructed to withstand tissue-bornehear stresses, which likely propagate across its dif-erent regions. A model of shear gradient and con-entric gradient is proposed to describe the region-pecific capacity of fibrocartilage to sustain sheartresses in tendons, ligaments, joints, and the heal-ng bone across species. © 2001 Elsevier Science (USA)

Key Words: fibrocartilage; shear stress; cartilage;tomic force microscopy; joint; condyle.

INTRODUCTION

Articular cartilage is either hyaline cartilage orbrocartilage, both resilient and virtually friction-

1 To whom correspondence should be addressed. Fax: (312)

96-7854. E-mail: jmao2@uic.edu.

46047-8477/01 $35.002001 Elsevier Science (USA)

ll rights reserved.

ess to allow movement of bones within the joint.ver since fibrocartilage was comprehensivelyrought to the attention of biologists (Benjamin andvans, 1990), its presence and functionality are in-reasingly recognized to be general as a tissue withommon characteristics (Clark and Stechschulte,998; Milz et al., 1999; Petersen and Tillmann,999). The importance of fibrocartilage in skeletaliology is obvious owing to its presence not only inhe knee, jaw, and intervertebral joints, but also inendons, ligaments, and the healing bone in mam-als, reptiles, and birds (Benjamin and Evans,

990; Benjamin and Ralphs, 1998; Carvalho andelisbino, 1999). Due to its presence in these diverseiological environments across species, fibrocarti-age has been thought to play fundamental rolesuch as resistance to tissue-borne mechanicaltresses (Benjamin and Ralphs, 1998). Indeed, largeacromolecular proteoglycans such as aggrecan and

ersican, known for their stress-bearing ability byttracting abundant water molecules, are increas-ngly expressed upon elevated application of sheartresses to fibrocartilage (Koob et al., 1992; Mao etl., 1998; Dowthwaite et al., 1999). Further evidencen support of fibrocartilage’s ability to sustain bio-

echanical stresses has originated from previousnvestigations of its macromaterial properties

ostly from the intervertebral disc and knee menis-us. Typically, tensile or compressive forces are ap-lied to the entire fibrocartilage sample up to theield point so that the modulus of elasticity andther material properties are obtained under steadyoading conditions (e.g., Arnoczky et al., 1988;rocter et al., 1989). This kind of characterization ofacromaterial properties is limited by a lack of ap-

reciation of any regional variation and commonlyo static loading conditions that contrast the dy-amic nature of biological stresses subjected to skel-tal tissues (Mow et al., 1989; Athanasiou et al.,995).

Fibrocartilage, a tissue with macromaterial prop-

etsutguhpcps1numlgsofwsannaiptjdspati1gsdvds

cwwaapSriof

gqfsTt

pmVaScgtBfwr

wv

ptisrcwrssoTtu

acddlmwftpc

47REGIONAL STRUCTURAL AND VISCOELASTIC PROPERTIES OF FIBROCARTILAGE

rties between dense fibrous tissue and hyaline car-ilage, is not well understood in its surface ultra-tructure, despite its fundamental importance in thenderstanding of skeletal defects such as osteoar-hrosis. Atomic force microscopy (AFM) enables re-ional analysis of cartilaginous samples over its nat-ral substrate, the subchondral bone. Only recentlyas AFM been used to study not only surface mor-hology of biological tissues, but also their microme-hanical properties. An advantage of AFM is to ap-ly dynamic nanoindentation to the substrate’surface of interest (Binning et al., 1986; Radmacher,997). In an elegant study by Jurvelin et al. (1996),anoindentation with atomic force microscopy wassed primarily to map the surface and subsurfaceorphology of the bovine humeral articular carti-

age, which is hyaline cartilage (named after itslassy appearance) rather than fibrocartilage. Theurface irregularities of hyaline cartilage previouslybserved with scanning electron microscopy wereound to be limited to less than 1000 nm observedith AFM (Jurvelin et al., 1996). Fibrocartilage has

tructural components, such as type I collagen, thatre not commonly present in hyaline cartilage. It isot known whether surface characteristics or dy-amic viscoelastic properties of fibrocartilage differmong its regions or from hyaline cartilage. Moremportantly, although fibrocartilage has been pro-osed to resist large shear stresses, little experimen-al evidence is available to sustain this notion (Ben-amin and Evans, 1990). To test a hypothesis thatifferent regions of articular fibrocartilage demon-trate different ultrastructural and viscoelasticroperties, we used dynamic nanoindentation withtomic force microscopy to probe different regions ofhe articular surface of the jaw joint condyle, whichs known to be composed of fibrocartilage (Ten Cate,985). Probing this fibrocartilage for both topo-raphic imaging and force spectroscopy demon-trated that its different regions indeed had a gra-ient distribution of both topographic andiscoelastic properties, providing experimental evi-ence that articular fibrocartilage is constructed toustain tissue-borne shear stresses.

MATERIALS AND METHODS

Sample preparation. A total of 18 left and right jaw jointondyles of nine normal, 6-week-old New Zealand White rabbitsere harvested within 1 h of euthanasia. After the condylar headas dissected transversely at the level of the condylar neck, therticular surface of the condyle was divided into four regions:nteromedial (AM), anterolateral (AL), posteromedial (PM), andosterolateral (PL) (Fig. 1) with a fine electric surgical saw (Hallurgical, Largo, FL). A whole-thickness fibrocartilage sample peregion, approximately 3 3 2 3 3 mm3 (length 3 width 3 height)ncluding the subchondral bone, was harvested. The bony surfacef the sample was rapidly dried and glued onto a glass slide using

ast-drying cyanoacrylate. Using a two-sided adhesive tape, the i

lass slide was fixed to the stainless steel disc, which was subse-uently magnetically mounted onto the piezoscanner of an atomicorce microscope (described below). During these procedures, eachample was constantly irrigated with phosphate-buffered saline.he present work was approved by the Animal Care Committee ofhe University of Illinois at Chicago.

Topographic imaging with atomic force microscope. Both to-ographic and force spectroscopy images were obtained in contactode using an atomic force microscope (AFM) (Nanoscope IIIa,eeco-Digital Instruments, Santa Barbara, CA). Cantilevers withnominal force constant of k 5 0.12 N/m and oxide-sharpened

i3N4 tips were used to apply nanoindentation against the fibro-artilaginous articular surfaces. Scan rates were 1 Hz for topo-raphic imaging and 14 Hz for force spectroscopy. The radius ofhe curvature of the scanning tips was approximately 20 nm.oth topographic and force spectroscopy images were obtained

rom the geometric center of each of the four regions (cf. Fig. 1b)ith 10 3 10 mm and 2 3 2 mm scan sizes. The mean surface

oughness was quantified by using the equation

Ra 5

Oi51

N

uZi 2 Zcpu

N ,

here Zcp is the Z value of the center plane, Zi is the current Zalue, and N is the number of points within a given area.

Dynamic nanoindentation for force spectroscopy. Once the to-ographic image was captured, force spectroscopy data were ob-ained by driving the cantilever tip in the Z plane. Force mappingnvolved data acquisition of nanoindentation load and the corre-ponding displacement in the Z plane during both extension andetraction of the cantilever tip. Upon completion of force spectros-opy for axial indentation on the articular surface, each sampleas bisected perpendicular to the articular surface in its geomet-

ic center and remounted onto another glass slide. The corre-ponding bisected surface of the fibrocartilage, 90° to the articularurface, was exposed for force spectroscopy in 10 condyles tobtain transverse strain in the same fashion as described above.he Poisson ratio (n) was defined as lateral nanoindentation 90°

o the articular surface against axial nanoindentation perpendic-lar to the articular surface and accordingly calculated by divid-

FIG. 1. Sagittal view of the rabbit jaw joint and mandible (a)nd superior view of the articular surface of the condyle that isomposed of fibrocartilage (b). The articular fibrocartilage wasivided and dissected into the following four regions: anterome-ial (AM), anterolateral (AL), posteromedial (PM), and postero-ateral (PL). A whole-thickness fibrocartilage sample, approxi-

ately 3 3 2 3 3 mm3 in size including the subchondral bone,as harvested in each region (b) for further sample preparation

or atomic force microscopy. The 3D view of the AM region illus-rates sample harvest. For each region, the center of the cross-oint as illustrated in the AL region (b) was selected for AFMharacterization.

ng the transverse strain on the bisected surface by its corre-

s(HMcpdpea

wlc

(fisreldtrvwrnonmPtl1w

nfeiewa1tsfstsr

Agf

titgTss(

srtpgMYPicnFbtdttAllartttpst

ttir(

48 HU ET AL.

ponding axial strain on the articular surface. Young’s modulusE) was calculated from force spectroscopy data by following theertz model (A-Hassan et al., 1998; Heinz and Hoh, 1999;athur et al., 2000), which defines a relationship between the

ontact radius, the nanoindentation load, and the central dis-lacement. For each sample, the average Young’s modulus waserived from individual calculations of three randomly selectedoints on the articular surface, each of 10 nm2 in size, using thequation (A-Hassan et al., 1998; Heinz and Hoh, 1999; Mathur etl., 2000)

E 53F~1 2 n!

4ÎRd3/ 2,

here E is the Young’s modulus, F is the applied nanomechanicaload, n is the Poisson ratio for a given region, R is the radius of theurvature of the AFM tip, and d is the amount of indentation.

Data analysis and statistics. The raw data of the Poisson ration) and Young modulus (E) for the left and right condyles wererst subjected to Student T tests to determine whether the Pois-on ratios and Young moduli differed between the left and theight sides. Once a lack of statistical significance in the differ-nces of both the Poisson ratios and Young moduli between theeft and the right condyles was respectively established, all theata between the left and the right sides were pooled for each ofhe four corresponding regions. All surface roughness, Poissonatio, and Young’s modulus data were subjected to the analysis ofariance (ANOVA) with Bonferroni adjustment to determinehether they differed significantly among and between the four

egions. Linear regression analysis was applied between theanoindentation load and deflection of the cantilever tip for eachf the AM, AL, PM, and PL regions. Linear regression and poly-omial equations up to the fourth order were applied to deter-ine the best fit for the relationship between Young’s moduli andoisson ratios of all four regions in order to determine whetherhe distribution of Young’s moduli among the four regions isinear and isotropic (Snedecor and Cochran, 1980; Mao et al.,996). For all the statistical analyses, P values of less than 0.05ere considered to be of statistical significance.

RESULTS

The microscopic appearance of the fibrocartilagi-ous articular surfaces was smooth, pale, and uni-orm with no indication of surface defect or degen-ration under the stereomicroscope. Topographicmaging with the AFM, however, revealed a gradi-nt distribution of the mean surface roughness thatas significantly different among the AM, AL, PM,nd PL regions with horizontal resolutions at 10 30 mm and 2 3 2 mm (P , 0.05) (Fig. 2). Represen-ative topographic images of the AM region (highesturface roughness) and the PL region (lowest sur-ace roughness) for both 10 3 10 mm and 2 3 2 mmcan sizes are illustrated in Fig. 3. Corresponding tohe mean surface roughness (Fig. 2), the AM regionhowed more robust surface topography than the PLegion for both 10 3 10 mm and 2 3 2 mm scans.

Poisson ratios differed significantly among theM, AL, PM, and PL regions (P , 0.01) (Fig. 4). Aradient distribution of Poisson’s ratios among the

our regions was noted: the highest (0.46 6 0.05) for e

he AM and lowest for the PL (0.31 6 0.05), indicat-ng different intrinsic capacities of different regionso withstand biomechanical stresses at various an-les to the articular surface, i.e., shear stresses.hese Poisson ratio data were verified by a lack ofignificant differences (P . 0.05) of correspondingamples between the left and the right condylesdata not shown).

Young’s moduli of all samples showed statisticallyignificant differences among the AM, AL, PM, and PLegions (P , 0.05) (Fig. 5). Notably the AM possessedhe highest Young’s modulus (2.34 6 0.26 MPa), ap-roximately onefold higher than the two posterior re-ions (PM, 1.11 6 0.07 MPa; and PL, 0.95 6 0.06Pa). There was a descending gradient distribution ofoung’s moduli among the four regions: AM . AL .M . PL (Fig. 5), suggesting that the intrinsic capac-

ty of fibrocartilage to sustain biomechanical loads de-reases from the anterior and medial regions diago-ally toward the posterior and lateral regions (also cf.ig. 9A). These data on Young’s modulus were verifiedy a lack of statistical significance in the differences ofhe average Young’s moduli of the left and right con-yles (Fig. 6), further validating the present AFMechnique. There was a quasi-linear relationship be-ween nanoindentation loads and displacement of theFM scanning tip as shown in Fig. 7. The range of

inear regression coefficients (r) was 0.99 for all fourinear regression equations, suggesting that themount of displacement of the AFM scanning tips iseliable for predicting nanoindentation forces appliedo the present samples. Despite the quasi-linear rela-ionship between force and displacement for each ofhe AM, AL, PM, and PL samples, Young’s modulilotted against Poisson’s ratios for the four corre-ponding regions showed an apparent nonlinear dis-ribution (Fig. 8). Linear regression and polynomial

FIG. 2. Surface roughness showed decreasing gradients fromhe anteromedial (AM) region toward the anterolateral (AL), pos-eromedial (PM), and posterolateral (PL) regions of fibrocartilagen both 10 3 10 mm and 2 3 2 mm scans. Differences in surfaceoughness among various regions were of statistical significanceP , 0.05).

quations up to the fourth order revealed that a third-

FIG

.3.

Rep

rese

nta

tive

topo

grap

hic

imag

esof

the

ante

rom

edia

l(A

M)

and

post

erol

ater

al(P

L)

regi

ons

ofar

ticu

lar

fibr

ocar

tila

gefo

rbo

th10

310

mm

(aan

db)

and

23

2m

msc

ans

(can

dd)

.T

he

AM

regi

ons

(aan

dc)

show

edm

ore

robu

stto

pogr

aph

yth

anth

eP

Lre

gion

s(b

and

d).

49REGIONAL STRUCTURAL AND VISCOELASTIC PROPERTIES OF FIBROCARTILAGE

o2rttl

fiamAageno(H

lfisehcnsihtos

lptt

((dt

((dt

sp(bA

aa(LA

50 HU ET AL.

rder polynomial equation, y 5 2355x3 1 481x2 202x 1 28, provided the best fit (P , 0.05) for theelationship between Young’s moduli and Poisson ra-ios, suggesting region-specific, anisotropic distribu-ion of viscoelastic properties of articular fibrocarti-age.

DISCUSSION

The gradient distribution in surface topography ofbrocartilage reflects microstructural differencesmong regions. This was evidenced by an approxi-ately 50% reduction in surface roughness from theM to PL regions at 10 3 10 mm resolution andpproximately 1/3 reduction from the AM to PL re-ions at 2 3 2 mm resolution. These mean differ-nces in surface roughness up to a few hundredanometers, however, are more minute than previ-usly observed with scanning electron microscopye.g., Bloebaum and Wilson, 1980; Ghadially, 1983;elminen et al., 1985; Kirk et al., 1994). Neverthe-

FIG. 4. The average Poisson ratios for the anteromedialAM), anterolateral (AL), posteromedial (PM), and posterolateralPL) regions of articular fibrocartilage showed a descending gra-ient distribution with statistically significant differences be-ween them (P , 0.01).

FIG. 5. The average Young’s moduli among the anteromedialAM), anterolateral (AL), posteromedial (PM), and posterolateralPL) regions of articular fibrocartilage showed a descending gra-ient distribution with statistically significant differences be-

ween them (P , 0.01). L

ess, the upper Z range of surface topography in ourndings was approximately 2000 nm (data nothown), as opposed to less than 1000 nm in Jurvelint al. (1996). This perhaps can be attributed toigher type I collagen content in the present fibro-artilaginous specimens as opposed to the predomi-ance of type II collagen in the hyaline cartilagepecimens in Jurvelin et al. (1996). The present find-ngs of a gradient distribution of surface roughness,ighest in the AM region and gradually decliningoward the PL region, suggest that different regionsf articular surfaces of fibrocartilage have differenturface ultrastructures.Nanomechanical analysis of articular fibrocarti-

age divided into four arbitrary regions in theresent work indicates the likelihood that fibrocar-ilage is capable of withstanding different magni-udes of shear stresses among its various regions.

FIG. 6. The Young’s modulus data were verified by a lack oftatistically significant differences (P . 0.05) of equivalent sam-les of the anteromedial (AM), anterolateral (AL), posteromedialPM), and posterolateral (PL) regions of articular fibrocartilageetween the left and the right condyles, validating the presentFM technique.

FIG. 7. Linear regression analysis of nanoindentation loadnd displacement of AFM scanning tips on the articular surfacesmong the anteromedial (AM), anterolateral (AL), posteromedialPM), and posterolateral (PL) regions of articular fibrocartilage.inear regression equations were as follows. AM: y 5 0.4x 2 9.2;L: y 5 0.3x 2 11.2; PM: y 5 0.2x 2 12.6; PL: y 5 0.2x 2 10.9.

inear correlation coefficients (r) were 0.99 for all four equations.

Fsfrststvtecso2tdrssdtwTftij1rcbidcip

tagvs(sfinlotaclsts

SWiapfN

ccloscfhAujbagcou

appba

51REGIONAL STRUCTURAL AND VISCOELASTIC PROPERTIES OF FIBROCARTILAGE

irst, Poisson ratio is defined by dividing transversetrain by axial strain. A greater Poisson ratio there-ore indicates a greater capacity of the structure toesist tissue-borne stresses at an angle to its axialurface (Mow et al., 1993; Zysset et al., 1999). Herehe axial surface is the articular surface that with-tands mostly compressive stresses at a right angleo the articular surface. The present region-specificariation in the distribution of Poisson ratios fromhe highest AM region (mean 0.46) toward the low-st PL region (mean 0.31) suggests a gradient in theapacity of the entire articular fibrocartilage to re-ist shear stresses. Second, the gradient distributionf Young’s moduli from the highest AM region (mean.34 MPa) to the lowest PL region (mean 0.95) fur-her differentiates stress-bearing capacities amongifferent regions. It is known that highly stressedegions of articular cartilages are stiffer in compres-ion than regions that experience less compressivetresses (Ahmed and Burke, 1983). These regionalifferences are statistically significant, indicatinghat different regions may have been constructed toithstand a gradient propagation of shear stresses.hat the distribution of Young’s moduli among dif-

erent regions is both nonlinear and anisotropic fur-her indicates both intrinsic control and exogenousnfluence on the properties of fibrocartilage (Ben-amin and Ralphs, 1998; Petersen and Tillmann,999). Fourth, the decreasing gradient of surfaceoughness from the AM to the PL regions suggestsoregulation of the ultrastructural properties of fi-rocartilage with its viscoelastic properties. Accord-ngly, a model that could potentially explain theistribution of regional ultrastructural and vis-oelastic properties of fibrocartilage with regard tots ability to withstand tissue-borne shear stresses is

FIG. 8. Young’s modulus plotted against Poisson ratio for thenteromedial (AM), anterolateral (AL), posteromedial (PM), andosterolateral (PL) regions of articular fibrocartilage. A third-orderolynomial equation, y 5 2355x3 1 481x2 2 202x 1 28, provided theest fit for the distribution of Young’s moduli of the four regionsmong linear regression and polynomial fits up to the fourth order.

roposed in Fig. 9. For instance, fibrocartilage in a

endons, ligaments, the healing bone, and certainrticulations such as the jaw joint likely has a “shearradient” distribution of both ultrastructural andiscoelastic properties so as to withstand sheartresses that radiate from the darkly shaded originof shear stresses) toward the opposite pole (lightlyhaded areas) (Fig. 9A). On the other hand, thebrocartilaginous intervertebral disc and knee me-iscus, which are stressed from omni-directions,

ikely will have a “concentric gradient” distributionf ultrastructural and viscoelastic properties, hencehe capacity to withstand shear stresses that prop-gate in all directions from the periphery toward theenter (Fig. 9B). In either case, articular fibrocarti-age possesses a gradient distribution of regionaltructural and viscoelastic properties in a designhat is apparently optimized for withstanding tis-ue-borne shear stresses.

We are grateful to Drs. C. S. Greene, J. W. Osborn, and R. P.capino for critically reviewing early versions of the manuscript.e thank three anonymous reviewers for their constructive crit-

cisms. Eric Rufe from Veeco-Digital Instruments is gratefullycknowledged for his technical support. This research was sup-orted by USPHS Research Grants DE13964 and DE13088, bothrom the National Institute of Dental and Craniofacial Research,ational Institutes of Health, Bethesda, Maryland.

FIG. 9. A model with two variations of shear gradient (A) andoncentric gradient (B) is proposed to describe the region-specificapacity of fibrocartilage to sustain shear stresses in tendons,igaments, joints, and the healing bone. Heavily shaded areas aref high shear stresses, whereas lightly shaded areas are of lowhear stresses. Variation A describes the shear gradient of fibro-artilage’s capacity to withstand shear stresses that propagaterom the anterior and medial toward the opposite pole (fromeavily shaded areas to lightly shaded areas). The two curves inare arbitrary divisions. This variation accounts for the present

ltrastructural and viscoelastic properties of fibrocartilage in theaw joint and perhaps also tendons, ligaments, and the healingone. In all these tissues, tissue-borne shear stresses likely prop-gate from one pole to another. Variation B describes a concentricradient that perhaps applies to fibrocartilage in the knee menis-us and intervertebral disc. Shear stresses likely originate frommni-directions and therefore influence the distribution of theltrastructural and viscoelastic properties from heavily shaded

reas to lightly shaded areas.

A

A

A

A

B

B

B

B

C

C

D

G

H

H

H

H

J

J

K

K

K

K

M

M

M

M

M

M

P

P

P

R

R

S

S

T

Z

52 HU ET AL.

REFERENCES

-Hassan E., Heinz, W. F., Antonik, M. D., D’Costa, N. P., Nag-eswaran, S., Schoenenberger, C. A., and Hoh, J. H. (1998)Relative microelastic mapping of living cells by atomic forcemicroscopy. Biophys. J. 74, 1564–1578.hmed, A. M., and Burke, D. L. (1983) In-vitro measurement ofstatic pressure distribution in synovial joints. Part I: Tibialsurface of the knee. J. Biomech. Eng. 105, 216–225.

rnoczky, S. P., McDevitt, C. A., Schmidt, M. B., Mow, V. C., andWarren, R. F. (1988) The effect of cryopreservation on caninemenisci: A biochemical, morphologic, and biomechanical eval-uation. J. Orthop. Res. 6, 1–12.thanasiou, K. A., Agarwal, A., Muffoletto, A., Dzida, F. J., Con-stantinides, G., and Clem, M. (1995) Biomechanical propertiesof hip cartilage in experimental animal models. Clin. Orthop.Rel. Res. 316, 254–266.enjamin, M., and Evans, E. J. (1990) Fibrocartilage. J. Anat.171, 1–15.enjamin, M., and Ralphs, J. R. (1998) Fibrocartilage in tendonsand ligaments—An adaptation to compressive load. J. Anat.193, 481–494.inning, G., Quate, C. F., and Gerber, C. (1986) Atomic forcemicroscopy. Phys. Rev. Lett. 56, 930–933.loebaum, R. D., and Wilson, A. S. (1980) The morphology of thesurface of articular cartilage in adult rats. J. Anat. 131, 333–346.arvalho, H. F., and Felisbino, S. L. (1999) The development ofthe pressure-bearing tendon of the bullfrog, Rana catesbeiana.Anat. Embryol. 200, 55–64.lark, J., and Stechschulte, D. J., Jr. (1998) The interface be-tween bone and tendon at an insertion site: A study of thequadriceps tendon insertion. J. Anat. 192, 605–616.owthwaite, G. P., Ward, A. C., Flannely, J., Suswillo, R. F.,Flannery, C. R., Archer, C. W., and Pitsillides, A. A. (1999) Theeffect of mechanical strain on hyaluronan metabolism in em-bryonic fibrocartilage cells. Matrix Biol. 18, 523–532.hadially, F. N. (1983) “Fine Structure of the Synovial Joints. AText and Atlas of the Ultrastructure of Normal and Patholog-ical Articular Tissues,” Butterworth, London.einz, W. F., and Hoh, J. H. (1999) Spatially resolved forcespectroscopy of biological surfaces using the atomic force mi-croscope. Trends Biotechnol. 17, 143–150.elminen, H. J., Jurvelin, J. S., Tammi, M., Pelttari, A., Svart-back, C. M., Kiviranta, I., Saamanen, A. M., and Paukkonen, K.(1985) Prolonged ethanol replacement by CO2 increases splitson articular cartilage surface after critical point drying. J.Microsc. 137, 305–12.och, D. H., Grodzinsky, A. J., Koob, T. J., Albert, M. L., andEyre, D. R. (1983) Early changes in material properties ofrabbit articular cartilage after meniscectomy. J. Orthop. Res. 1,4–12.ori, R. Y., and Mockros, L. F. (1976) Indentation tests of humanarticular cartilage. J. Biomech. 9, 259–268.

urvelin, J. S., Arokoski, J. P. A., Hunziker, E. B., and Helminen,H. J. (2000) Topographical variation of the elastic properties ofarticular cartilage in the canine knee. J. Biomech. 33, 669–675.

urvelin, J. S., Muller, D. J., Wong, M., Studer, D., Engel, A., andHunziker, E. B. (1996) Surface and subsurface morphology ofbovine humeral articular cartilage as assessed by atomic forceand transmission electron microscopy, J. Struct. Biol. 117,45–54.

empson, G. E., Freeman, M. A. R., and Swanson, S. A. V. (1971)

The determination of a creep modulus for articular cartilagefrom indentation tests on the human femoral head. J. Biomech.4, 239–250.irk, T. B., Stachowiak, G. W., and Wilson, A. S. (1994) Themorphology of the surface of articular cartilage. In “Proceed-ings of the Second World Congress of Biomechanics, Amster-dam, July 10–15, 1994,” (L. Blankevoort, Ed.), Vol. II, p. 208a.Elsevier Science, Amsterdam, The Netherlands.oob, T. J., Clark, P. E., Hernandez, D. J., Thurmond, F. A., andVogel, K. G. (1992) Compressive loading in vitro regulatesproteoglycan synthesis by tendon fibrocartilage. Arch. Biochem.Biophys. 298, 303–312.oob, T. J., and Vogel, K. G. (1987) Proteoglycan synthesis inorgan cultures from regions of bovine tendon subjected to dif-ferent mechanical forces. Biochem. J. 246, 589–598.ao, J. J., Major, P. W., and Osborn, J. W. (1996) Couplingelectrical and mechanical outputs of jaw muscles undertakingdirectional bite-force tasks. Arch. Oral. Biol. 41, 1141–1147.ao, J. J., Rahemtulla, F., and Scott, P. G. (1998) Proteoglycanexpression in the rat temporomandibular joint in response tounilateral bite raise. J. Dent. Res. 77, 1520–1528.athur, A. B., Truskey, G. A., and Reichert, W. M. (2000) Atomicforce and total internal reflection fluorescence microscopy forthe study of force transmission in endothelial cells. Biophys. J.78, 1725–1735.ilz, S., Putz, R., Ralphs, J. R., and Benjamin, M. (1999) Fibro-cartilage in the extensor tendons of the human metacarpopha-langeal joints. Anat. Rec. 256, 139–45.ow, V. C., Ateshian, G. A., and Spilker, R. L. (1993) Biomechan-ics of diarthrodial joints: A review of twenty years of progress.J. Biomech. Eng. 115, 460–467.ow, V. C., Gibbs, M. C., Lai, W. M., Zhu, W. B., and Athanasiou,K. A. (1989) Biphasic indentation of articular cartilage, Part II.A numerical algorithm and an experimental study. J. Biomech.22, 853–861.

erez-Castro, A. V., and Vogel, K. G. (1999) In situ expression ofcollagen and proteoglycan genes during development of fibro-cartilage in bovine deep flexor tendon. J. Orthop. Res. 17,139–148.

etersen, W., and Tillmann, B. (1999) Structure and vasculariza-tion of the cruciate ligaments of the human knee joint. Anat.Embryol. 200, 325–334.

roctor, C. S., Schmidt, M. B., Whipple, R. R., Kelly, M. A., andMow, V. C. (1989) Material properties of the normal medialbovine meniscus. J. Orthop. Res. 7, 771–82.

admacher, M. (1997) Measuring the elastic properties of biolog-ical samples with the AFM. IEEE Eng. Med. Biol. Mag. 16,47–57.admacher, M., Fritz, M., and Hansma, P. K. (1995) Imaging softsamples with the atomic force microscope: Gelatin in water andpropanol. Biophys. J. 69, 264–270.

etton, L. A., Mow, V. C., Muller, F. J., Pita, J. C., and Howell,D. S. (1994) Mechanical properties of canine articular cartilageare significantly altered following transection of the anteriorcruciate ligament. J. Orthop. Res. 12, 451–463.

nedecor, G. W., and Cochran, W. G. (1980) “Statistical Methods,”pp. 291–292, Iowa State Univ. Press, Ames.

en Cate, A. R. (1985). “Oral Histology: Development, Structure,and Function,” 2nd ed., pp. 382–383, Mosby, St. Louis.

ysset, P. K., Guo, X. E., Hoffler, C. E., Moore, K. E., and Gold-stein, S. A. (1999) Elastic modulus and hardness of cortical andtrabecular bone lamella measured by nanoindentation in the

human femur. J. Biomech. 32, 1005–1012.