Topographical variations in articular cartilage and subchondral bone of the normal rat knee are age-related

Page 1

R

To

Na

b

a

ARRA

KRRCMIM

1

iavb1uhathtjj

h0

Annals of Anatomy 196 (2014) 278–285

Contents lists available at ScienceDirect

Annals of Anatomy

j ourna l h omepage: www.elsev ier .de /aanat

esearch article

opographical variations in articular cartilage and subchondral bonef the normal rat knee are age-related

ina Hamanna,∗, Gert-Peter Brüggemanna,b, Anja Niehoffa,b

Institute of Biomechanics and Orthopaedics, German Sport University Cologne, Am Sportpark Müngersdorf 6, 50933 Cologne, GermanyCologne Center for Musculoskeletal Biomechanics, Medical Faculty, University of Cologne, Joseph-Stelzmann-Str. 9, 50931 Cologne, Germany

r t i c l e i n f o

rticle history:eceived 5 February 2014eceived in revised form 24 April 2014ccepted 28 April 2014

eywords:egionalodentartilage biomechanicsicro CT

mmatureature

a b s t r a c t

In osteoarthritis animal models the rat knee is one of the most frequently investigated joint. However, it isunknown whether topographical variations in articular cartilage and subchondral bone of the normal ratknee exist and how they are linked or influenced by growth and maturation. Detailed knowledge is neededin order to allow interpretation and facilitate comparability of published osteoarthritis studies. For thefirst time, the present study maps topographical variations in cartilage thickness, cartilage compressiveproperties and subchondral bone microarchitecture between the medial and lateral tibial compartment ofnormal growing rat knees (7 vs. 13 weeks). Thickness and compressive properties (aggregate modulus) ofcartilage were determined and the subchondral bone was analyzed by micro-computed tomography. Wefound that articular cartilage thickness is initially homogenous in both compartments, but then differen-tiates during growth and maturation resulting in greater cartilage thickness in the medial compartmentin the 13-week-old animals. Cartilage compressive properties did not vary between the two sites inde-pendently of age. In both age-groups, subchondral plate thickness as well as trabecular bone volume ratio

and trabecular thickness were greater in the medial compartment. While a high porosity of subchondralbone plate with a high topographical variation (medial/lateral) could be observed in the 7-week-old ani-mals, the porosity was reduced and was accompanied by a reversion in topographical variation whenreaching maturity. Our findings highlight that there is a considerable topographical variation in articularcartilage and subchondral bone within the normal rat knee in relation to the developmental status.

© 2014 Elsevier GmbH. All rights reserved.

. Introduction

According to Wolff’s law there is a dynamic regulatory systemn bone that adapts its strength to its mechanical environment bylterations in amount and internal architecture (Wolff, 1892). Thus,ariations in bone microarchitecture do not solely exist betweenones such as tibia, femur or lumbar vertebra (Iwamoto et al.,999; Hagihara et al., 2005), but can be also found within individ-al bones (Bourrin et al., 1995). This topographical heterogeneityas also been demonstrated for the subchondral bone with hardernd thicker tissue observed in areas which had undergone addi-ional loading (Oettmeier et al., 1992; Murray et al., 2001). Theypothesis that this concept of ‘functional adaptation’ also applies

o articular cartilage has been strengthened by the observation thatoint cartilage differs depending on the loads to which it is sub-ected. More specifically, previous work has shown that articular

∗ Corresponding author. Tel.: +49 0221 4982 7660; fax: +49 0221 4971 598.E-mail address: [email protected] (N. Hamann).

ttp://dx.doi.org/10.1016/j.aanat.2014.04.006940-9602/© 2014 Elsevier GmbH. All rights reserved.

cartilage from different locations and surfaces varies in thickness,biochemical composition as well as biomechanical properties dueto differences in loading history (Kiviranta et al., 1988; Athanasiouet al., 1991; Rasanen and Messner, 1996; Froimson et al., 1997; Firthand Rogers, 2005). In addition, it has to be considered that especiallyduring growth and maturation both articular cartilage and bonedevelop from a rather homogenous tissue to a topographically spe-cialized and heterogeneous tissue in order to fulfill its mechanicaldemands (Helminen et al., 2000; Khan et al., 2000).

Articular cartilage and subchondral bone act as a functional unit(Lories and Luyten, 2010). Thereby, cartilage behaves as a surfacelayer enabling load transmission and nearly frictionless movement;whereas the subchondral bone forms the structural socket for theoverlying articular cartilage. Both structures are limited in theircapacities to effectively absorb impacts (Malekipour et al., 2013).However, given the close functional relationship of these two struc-

tures, changes in one can be expected to influence the structure andfunction of the other. For example, osteoarthritis (OA) results in aprogressive degradation of articular cartilage as well as in architec-tural alterations of subchondral bone. Thereby, subchondral bone

Page 2

f Anatomy 196 (2014) 278–285 279

lsewl(tca

Iei22aYsisfhntafaaaaswwucd

2

2

Swtitb(dchNC

2

tizpssd

Fig. 1. Cartilage thickness assessment and mechanical testing of tibial articular

N. Hamann et al. / Annals o

oss during early stages of the disease is followed by bone sclero-is in later stages (Dedrick et al., 1993; Botter et al., 2006; Sniekerst al., 2008). Sclerotic changes in subchondral bone are associatedith increased bone stiffness, which in turn enhances local carti-

age stresses and therefore progression of cartilage damage and OARadin and Rose, 1986; Burr, 1998). However, the relationship ofhese structures to each other remains unclear, not only in terms ofartilage degeneration or during growth and development, but alsos to whether they differentiate synchronously or consecutively.

OA has been studied in a variety of different mammalian models.n particular, the rat knee joint has been widely used to investigatexperimentally induced OA, either by monosodium iodoacetatenjection (Mohan et al., 2011; Lee et al., 2012; Marker and Pomonis,012; Xie et al., 2012), medial meniscus transection (Janusz et al.,002; Wancket et al., 2005; Xie et al., 2010) or anterior cruci-te ligament transection (McErlain et al., 2012; Tsai et al., 2013;amaguchi et al., 2013). However, due to the fact that differenttudies analyzed different parameters, comparisons across stud-es are sometimes difficult. In addition, the age, developmentaltatus and tissue regions analyzed vary considerably between dif-erent studies. Therefore, detailed knowledge on the topographicaleterogeneity of articular cartilage and subchondral bone in theormal, non-manipulated rat knee, in general, but also in relationo growth and maturation is mandatory. This allows interpretationnd facilitates comparability of published OA research studies. Soar, there is very little information on local variations in cartilagend subchondral bone at different sites in the normal rat knee jointnd how these are linked or influenced by growth of the animals. In

recently published study, we demonstrated a growth-related vari-bility in articular cartilage thickness and biomechanics as well asubchondral bone architecture in the rat knee (Hamann et al., 2013),ithout analyzing local differences. In the present follow-up study,e hypothesized that, under the influence of growth and mat-ration, topographical variations in cartilage thickness, cartilageompressive properties and subchondral bone microarchitectureevelop between the medial and lateral tibial compartments.

. Materials and methods

.1. Samples

Left tibiae from 7- and 13-week-old (each n = 12) femaleprague-Dawley rats were obtained immediately after death,rapped in saline-soaked gauze and stored at −20 ◦C until

esting. All samples were analyzed within 2–4 weeks after freez-ng/collection and underwent only one freeze-thaw cycle. Rightibiae were used for biochemical analyses studying the interactionetween articular cartilage and subchondral bone during growthHamann et al., 2013). The study protocol and all animal proce-ures were in compliance with the principles of laboratory animalare and the German Laws on the Protection of Animals. The studyas been approved by an ethical committee and authorized by theorth Rhine-Westphalian State office for Environment, Health andonsumer Protection (8.87-50.10.45.08.188).

.2. Cartilage thickness assessment

On the testing day, tibiae were thawed for 30 min at roomemperature in 150 mM sodium chloride containing proteasenhibitors (2.3 mM ethylenediaminetetraacetic acid, 5 mM ben-amidine hydrochloride, 10 mM N-ethylmaleimide, and 1 mM

henylmethylsulfonylfluoride; Sigma–Aldrich). After thawing,amples were positioned vertically in a custom made stainlessteel sample chamber with a two component synthetic embed-ing system (Technovit 4004, Heraeus Kulzer GmbH, Werheim,

cartilage (load-bearing area). Overview of the positions of indentation. I (inner),O (outer), P (posterior) position for needle indentation during cartilage thicknessassessment. SR, stress-relaxation test position using a porous plane-ended indenter.

Germany). In order to ensure that the cartilage surface was hori-zontally oriented the sample holder was positioned on a custommade adjustable cross table equipped with millimeter subdivi-sion scaling which enabled a two-plane tilt of the sample. Further,the sample holder was surrounded by a plexiglas ring to allowmeasurements in saline solution including protease inhibitors aspreviously described (Hamann et al., 2013, 2014). Three thicknessmeasurements were performed on each medial and lateral com-partment by slow (0.006 mm/s) insertion of a needle probe (0.1 mmdiameter) (Hoch et al., 1983) attached to a 10 N load cell on ahigh-precision material test machine (traverse resolution 0.04 �m)(Zwick Z2.5/TN1S, Zwick GmbH & Co. KG, Ulm, Germany). The testpoints were arranged in a triangle, measurements were alwaysconducted in the same chronological order (inner (I), outer (O),posterior (P) surface) and positioned slightly posterior to the ante-rior/posterior midline of the tibia (Fig. 1). The load-displacementcurves were recorded at 50 Hz and used for thickness determi-nation: a change in force from zero to non-zero slope indicatedcontact with the cartilage surface and a change to a rapid increasein force indicated contact with the subchondral bone. The differ-ences in displacement between the two characteristics denotedcartilage thickness. A custom written MATLAB (Version 7.8.0.347,MathWorks, Natick, MA, USA) program was used to extract the car-tilage thickness of each measurement. The average value of thethree thickness measurements acted as initial cartilage thicknessfor the following biomechanical test.

2.3. Cartilage biomechanical testing

Biomechanical tests were performed under stress-relaxationimmediately after thickness determination of the respective tib-ial compartment at the load-bearing area as described previously(Hamann et al., 2013, 2014). Briefly, a porous plane-ended inden-ter (0.5 mm diameter) was used and positioned between the threethickness measurements (SR) (Fig. 1). After preloading at 0.002 Nand a subsequent 5 min equilibration period the stress-relaxationtest was performed. Three steps of strain (10, 15 and 20%) wereapplied for 400 s each until cartilage tissue attained ultimate relax-

ation. Force data were recorded with 50 Hz during the wholemeasurement time. Aggregate moduli HA were calculated from thelinear slope of the 10–20% ultimate relaxation stress–strain curvesusing the equation of Hayes et al. (1972) and Korhonen et al. (2002).

Page 3

280 N. Hamann et al. / Annals of Ana

Fig. 2. Reproducibility of mechanical tests with different rubber sheets. (A) Step-wrd

Ia

rbwaedbitcmti

2

3uspuswt

ise relaxation behavior of three measurements at each rubber sheet. (B) Ultimateelaxation stress–strain curves of three measurements at each rubber sheet. Theiameter of the porous plane-ended indenter was 0.5 mm.

n accordance with Wang et al. (2006) the Poisson’s ratio �S was sett 0.3 for rat articular cartilage.

In order to evaluate the reproducibility of the mechanical tests,epeated measurements were performed on three different rub-er sheets of varying thickness. The thickness of each rubber sheetas determined by three needle indentation tests. Subsequently,

ggregate moduli of rubber sheets were calculated from the lin-ar slope of the 10–20% ultimate relaxation stress–strain curves asescribed above. A Poisson’s ratio of 0.45 was defined for the rub-er sheets and the identical porous plane-ended indenter of 0.5 mm

n diameter as used for the cartilage samples was utilized. For thehickness measurements, the calculated root mean squared coeffi-ient of variation (RMS CV (%)) was 1.58%, while for the aggregateodulus a RMS CV of 1.43% was determined. The reproducibility of

he relaxation curves as well as the stress–strain curves is shownn Fig. 2.

.4. Micro-computed tomography (�CT) of subchondral bone

After biomechanical testing, high-resolution �CT scanning (�CT5, Scanco Medical AG, Bassersdorf, Switzerland) was used to eval-ate morphological indices of bone volume and architecture of theubchondral bone. Samples were separated from sample holder andlaced vertically in a specimen tube, with the tibia plateau facing

pwards. During the entire measurement time the samples wereubmerged in 70% ethanol. Whole tibial epiphyses were scannedith an isotropic voxel size of 6 �m, 55 kVp tube voltage, 145 mA

ube current, and 400 ms integration time. A 0.5 mm aluminum

tomy 196 (2014) 278–285

filter was used. Gray-scale CT images were filtered using a con-strained Gaussian filter (support = 2, sigma = 1.2) to remove noisein the original volume data. After reconstruction, 2D images of thetibial subchondral bone (sagittal plane perpendicular to the car-tilage surface) were segmented into cortical (subchondral plate)and trabecular (subchondral trabecular bone) bone using a fully-automatic image analysis algorithm (Scanco Medical) based on thedual threshold technique by Buie et al. (2007). Segmented sagi-ttal subchondral bone images were evaluated on a slice by slicebasis until subchondral bone was fully expanded (central region)in each compartment so that 100 slices for medial and 125 slices forthe lateral site were defined. Then, individual volumes of interest(VOI) were oriented posterior to the anterior/posterior midline ofthe tibia (analogs regions of stress-relaxation tests) in each medialand lateral compartment by cutting the pre-segmented corticaland trabecular bone to a width of 1.0 mm. Medial and lateral com-partments could be delimited due to loss of typical appearance ofsubchondral trabecular and cortical bone in the mid-compartmentarea where the crucial ligaments insert. In order to extract the bonetissue (cortical and trabecular) data were globally thresholded (25%for cortical bone of the 7-week-old and 34% for cortical and tra-becular bone of the 13-week-old animals of the maximal possiblegray value). Thresholds were adjusted due to obvious variationsin the degree of bone mineralization between the two age-groupsto allow for accurate reconstructions of the scanned subchondralbone (Bouxsein et al., 2010).

For the subchondral plate analyses, plate thickness (Pl.Th, �m),plate porosity (Pl.Por, %) describing the ratio of the volume of thepores over the total volume as well as the plate bone mineral den-sity (Pl.BMD, mg HA/ccm BV) of bone tissue were calculated. For thesubchondral trabecular bone analyses, bone volume ratio (BV/TV,%) was calculated by summing the bone volume divided by thetotal volume, and expressed as a percentage. Trabecular thickness(Tb.Th, �m), trabecular number (Tb.N, 1/mm), trabecular separa-tion (Tb.Sp, �m) and trabecular bone mineral density (Tb.BMD, mgHA/ccm BV) of bone tissue were also determined. The �CT data ofthe medial compartment were published earlier (Hamann et al.,2013).

In order to validate the image analysis/bone segmentation ofthe �CT measurements a reproducibility study was performed. Weused five �CT datasets of 7- and 13-week-old animals, respectively.For intraobserver reproducibility, one operator analyzed the tendatasets three times. In order to determine interobserver repro-ducibility, two operators analyzed the ten datasets once. Intra- andinterobserver reproducibility was estimated by the RMS CV (%). Theanalyses were performed separately for the subchondral plate andsubchondral trabecular bone. Intraobserver analysis reproducibil-ity values of the 7- and 13-week-old animals for subchondral plateparameters (Pl.Th, Pl.Por) ranged from 0.42 to 2.39% and from 0.54to 1.29% for the subchondral trabecular bone parameters (BV/TV,Tb.Th), respectively. Interobserver analysis reproducibility valuesof the 7- and 13-week-old animals for subchondral plate parame-ters (Pl.Th, Pl.Por) ranged from 0.96 to 3.14% and from 0.94 to 1.61%for the subchondral trabecular bone parameters (BV/TV, Tb.Th),respectively.

2.5. Statistics

All data were checked for normal distributions and homogeneityof variances by Kolmogorov–Smirnov and Levene’s test, respec-tively. Results are presented as mean values ± standard deviation.A two-way (age × compartment) ANOVA with repeated measure-

ments was used to examine the effect of age and compartment ontibial articular cartilage and subchondral bone. Duncan’s post hoccomparison was performed when a significant main effect or inter-action was detected. Interrelations between cartilage parameters

Page 4

N. Hamann et al. / Annals of Anatomy 196 (2014) 278–285 281

Fa

(bbwawG

3

3

de(tn(iirtmtoom

3

salsa(tm7

3

(cpl

Fig. 4. Tibial cartilage biomechanics of 7- and 13-week-old animals. (A) Represen-tative stepwise relaxation behavior of articular cartilage of the medial and lateral

a significant (P < 0.05) interaction (age × compartment) for thePl.Por and Pl.BMD. The post hoc analysis revealed a significantly(P < 0.05) lower Pl.Por and Pl.BMD at the medial compared to the

ig. 3. Mean tibial cartilage thickness of the medial and lateral compartments of 7-nd 13-week-old animals. Values are means ± SD.

cartilage thickness and cartilage biomechanics) and subchondralone parameters (subchondral plate and subchondral trabecularone) were analyzed using Pearson’s bivariate correlation analysisith Bonferroni correction. A P-value less than 0.05 was regarded

s statistically significant. All statistical analyses were performedith Statistica for Windows (Version 7.1, StatSoft GmbH, Hamburg,ermany).

. Results

.1. Cartilage thickness

In the 7-week-old animals, the mean tibial cartilage thicknessid not differ significantly (P > 0.05) between the medial and lat-ral compartments (medial: 232 ± 27 �m, lateral: 232 ± 37 �m)Fig. 3). In the 13-week-old animals, the mean tibial cartilagehickness of the medial compartment was slightly, but not sig-ificantly (P = 0.07) thicker compared to the lateral compartmentmedial: 150 ± 20 �m, lateral: 137 ± 13 �m). There was a signif-cant (P < 0.05) age effect on the mean tibial cartilage thicknessndependent of the analyzed compartment. The post hoc analysisevealed that the 13-week-old animals had significantly (P < 0.05)hinner tibial articular cartilage compared to the 7-week-old ani-

als (Fig. 3). Structural defects could be observed in one medial andwo lateral tibial compartments of the 7-week-old animals and inne lateral tibial compartment of the 13-week-old animals. Basedn this, these samples were excluded from the thickness measure-ents and the subsequent biomechanical tests.

.2. Cartilage biomechanics

Representative relaxation curves as well as ultimate relaxationtress–strain curves from medial and lateral tibial cartilage of bothge-groups are shown in Fig. 4. In both groups the aggregate modu-us HA obtained from the equilibrium stress–strain curve at 10–20%train was not significantly (P > 0.05) different between the medialnd lateral compartments (Fig. 5). However, we found a significantP < 0.05) age effect on the aggregate modulus HA independent ofhe analyzed compartment. With post hoc testing, lower aggregate

oduli could be observed for the 13-week-old compared to the-week-old animals (P < 0.05) (Fig. 5).

.3. Subchondral plate microarchitecture

For the Pl.Th there was a significant compartment and age effect

both P < 0.05). Independent of the age of the animals the medialompartment showed higher values compared to the lateral com-artment. Furthermore, the 7-week-old animals demonstrated

ower values compared to the 13-week-old animals in both

compartments. (B) Representative ultimate relaxation stress–strain curves of artic-ular cartilage of the medial and lateral compartments. The diameter of the porousplane-ended indenter was 0.5 mm.

compartments (Fig. 6B). For the Pl.Por and Pl.BMD a significant(P < 0.05) age effect could be detected. In both compartments,higher Pl.Por values could be observed for the 7-week-old ani-mals, whereas for the Pl.BMD lower values were found for the7-week-old animals (Fig. 6B and Table 1). In addition, there was

Fig. 5. Aggregate moduli HA of articular cartilage of the medial and lateral tibialcompartments of 7- and 13-week-old animals. Values are means ± SD.

Page 5

282 N. Hamann et al. / Annals of Anatomy 196 (2014) 278–285

Fig. 6. �CT analyses of tibial subchondral bone. (A) Structural differences in subchondral plate and subchondral trabecular bone between the medial and lateral compartmentsof the two age-groups. Analyzed regions with a width of 1.0 mm are indicated by dashed lines. (B) Plate thickness (Pl.Th) and plate porosity (Pl.Por) of the subchondral platein the medial and lateral compartments of 7- and 13-week-old animals. (C) Bone volume fraction (BV/TV) and trabecular thickness (Tb.Th) of the subchondral trabecularbone in the medial and lateral compartments of 7- and 13-week-old animals. Values are means ± SD.

Table 1Subchondral bone parameters at the medial and lateral tibial compartment of the 7- and 13-weeks-old animals determined by �CT analyses.

7 weeks (n = 12) 13 weeks (n = 12)

Medial Lateral Medial Lateral

Pl.Th (�m) 58.2 ± 11.5a,b 39.1 ± 6.0b 105.4 ± 10.1a 93.2 ± 6.3Pl.Por (%) 16.94 ± 6.12a,b 36.37 ± 7.52b 4.73 ± 1.21 7.21 ± 1.73Pl.BMD (mg HA/ccm BV) 754 ± 40a,b 738 ± 29b 930 ± 17 922 ± 18BV/TV (%) 41.2 ± 3.2a,b 38.1 ± 3.1b 50.9 ± 3.0a 45.9 ± 3.5Tb.Th (�m) 72.0 ± 3.6a,b 65.7 ± 3.2b 92.2 ± 4.3a 84.3 ± 5.0Tb.N (1/mm) 6.19 ± 0.45 6.01 ± 0.39 6.90 ± 0.59a 5.94 ± 0.53Tb.Sp (�m) 142 ± 14 150 ± 13 133 ± 14a 156 ± 20Tb.BMD (mg HA/ccm BV) 919 ± 19a,b 906 ± 17b 984 ± 14a 966 ± 13

Plate thickness (Pl.Th), plate porosity (Pl.Por), plate bone mineral density (Pl.BMD) of bone tissue, bone volume ratio (BV/TV), trabecular thickness (Tb.Th), trabecular number(Tb.N), trabecular separation (Tb.Sp), trabecular bone mineral density (Tb.BMD) of bone tissue. Values are means ± SD.

a Statistical significant (P < 0.05) compartment effect medial vs. lateral (two-way ANOVA).b Statistical significant (P < 0.05) age effect 7 weeks vs. 13 weeks (two-way ANOVA).

Table 2Linear correlations between cartilage parameters and subchondral bone parameters of the 7- and 13-weeks-old animals.

Pl.Th (�m) Pl.Por (%) BV/TV (%) Tb.Th (�m)

r P-value r P-value r P-value r P-value

Medial (n = 23)Cartilage thickness (�m) −0.826 <0.001 0.773 <0.001 −0.841 <0.001 −0.904 <0.001Aggregate modulus (MPa) −0.442 0.140 0.316 0.568 −0.463 0.104 −0.581 0.016

Lateral (n = 21)Cartilage thickness (�m) −0.871 <0.001 0.854 <0.001 −0.783 <0.001 −0.852 <0.001

P lar th

la

3

aeBTclao

Aggregate modulus (MPa) −0.517 0.064 0.437

late thickness (Pl.Th), plate porosity (Pl.Por, %), bone volume ratio (BV/TV), trabecu

ateral compartment in the 7-week-old, but not in the 13-week-oldnimals (Fig. 6B and Table 1).

.4. Subchondral trabecular bone microarchitecture

For the BV/TV, Tb.Th and Tb.BMD a significant compartment andge effect (both P < 0.05) was observed. In comparison to the lat-ral compartment, the medial compartment demonstrated a higherV/TV, Tb.Th and Tb.BMD independent of the age of the animals.he 7-week-old animals showed a lower BV/TV, Tb.Th and Tb.BMD

ompared to the 13-week-old animals independent of the ana-yzed compartment (Fig. 6C and Table 1). For the Tb.N and Tb.Sp

significant (P < 0.05) interaction (age × compartment) could bebserved. Only in the 13-week-old animals was the Tb.N of the

0.188 −0.449 0.164 −0.548 0.040

ickness (Tb.Th). Significant correlations are indicated bold lettering.

medial compartment significantly (P < 0.05) higher compared to thelateral compartment, whereas the Tb.Sp was lower in the medialcompared to the lateral compartment (Table 1).

3.5. Correlations between cartilage and subchondral bone

The mean cartilage thickness and aggregate modulus HA werecorrelated with the Pl.Th, Pl.Por, BV/TV and Tb.Th. Linear correlationcoefficients as well as P-values are summarized in Table 2.

4. Discussion

The rat knee is widely used to investigate experimentallyinduced OA (Xie et al., 2010, 2012; Yamaguchi et al., 2013). There

Page 6

f Anat

iltpaal

awmwIbtw(dsinup

acnsoiteiltticircdhehth

pagltposbnoha(fmfm

N. Hamann et al. / Annals o

s a lack of knowledge of topographical variations in articular carti-age and subchondral bone within the normal rat knee and howhey are linked or influenced by growth and maturation. In theresent study, we therefore evaluated whether topographical vari-tions exist in cartilage thickness, cartilage compressive propertiesnd subchondral bone microarchitecture between the medial andateral tibial compartments of the normal growing rat knee.

According to Ellender et al. (1988) 7-week-old rats are still in rapid growth phase and reach skeletal maturity at an age of 13eeks. Recently, we demonstrated that cartilage thickness in theedial tibial compartment (load-bearing area) decreases by 35%hen comparing 7 with 13 week old animals (Hamann et al., 2013).

n the present study, a thinner tibial articular cartilage could alsoe observed at the lateral compartment in 13-week-old comparedo 7-week-old animals. Our observations are in good agreementith earlier findings by Brommer et al. (2005), van Turnhout et al.

2010) and Xie et al. (2009) demonstrating that cartilage thicknessecreases during growth due to the subsequent development of theubchondral bone (Marks and Hermey, 1996). Further, this results supported by our analyses showing that cartilage thickness isegatively correlated with subchondral plate thickness, bone vol-me ratio, trabecular thickness and positively correlated with plateorosity.

It was expected that growth would lead to topographical vari-tions in cartilage thickness between the medial and lateral tibialompartments. Our local comparisons revealed that cartilage thick-ess was tendentially higher in the medial compared to the lateralite in 13-week-old animals, whereas no differences could bebserved in 7-week-old animals. On the one hand, mechanical load-ng has been considered to be the source of initiating OA when it isoo excessive (Griffin and Guilak, 2005), but, on the other hand, it isssential to the formation and preservation of cartilage structure,ts biomechanical composition and mechanical resistance againstoads (Helminen et al., 2000). In addition, it is widely acceptedhat, at birth, articular cartilage is a quite uniformly designed tissuehat is remodeled into a topographical heterogeneous tissue dur-ng growth and under the influence of mechanical loading. Thesehanges are needed to fulfill the required biochemical and mechan-cal demands (Brama et al., 2000). Gardner-Morse et al. (2013)ecently published that peak contact stresses in the rat medial tibialompartment are higher compared to the lateral site. In accor-ance with earlier findings that articular cartilage is thicker ineavily loaded regions of joints (Jurvelin et al., 1986; Kivirantat al., 1988), we speculate that, in terms of a functional adaptation,igher mechanical loads acting on the medial site during locomo-ion and normal activity over time contribute to this topographicaleterogeneity in cartilage thickness in mature rats.

Subsequent development and condensation of subchondrallate as well as subchondral trabecular bone from 7 to 13 weeks ofge are shown in Fig. 6A. Growth was expected to generate topo-raphical variations in subchondral bone between the medial andateral tibial compartment. In accordance with our hypothesis, thehickness of the subchondral plate was higher at the medial com-ared to the lateral compartment in both age-groups, while thepposite was seen in the porosity of the subchondral plate. Theame pattern could be observed for the subchondral trabecularone showing a higher bone volume ratio and trabecular thick-ess at medial sites compared to lateral sites regardless of the agef the animals. Additional loading resulting from physical activityas been demonstrated to increase subchondral bone thickness in

site-specific manner in dogs (Oettmeier et al., 1992) and horsesMurray et al., 2001). Bone develops its structure according the

orces acting on it (Wolff, 1892). We therefore speculate that higher

echanical loads acting on the medial tibial compartment led to aunctional, developmental adaptation similar to that described in

edial cartilage thickness. This assumption is further strengthened

omy 196 (2014) 278–285 283

by the study of Gardner-Morse et al. (2013) that demonstrated thatthe medial tibial site in the rat knee joint is under higher com-pressive contact stress than the lateral site. Moreover, it appearsthat condensation of the subchondral plate as reflected by the plateporosity is limited. While a high porosity with a high topographicalvariation (medial/lateral) could be observed in the 7-week-old ani-mals, the porosity was much more reduced and was paralleled bya reversion in its topographical variation when reaching maturity.This seems to be a functional limitation because the subchondraltrabecular bone is suggested to provide nutrient delivery to theadjacent cartilage by its capillaries which need to pass the sub-chondral plate (Malinin and Ouellette, 2000).

Cartilage compressive properties were expected to develop dif-ferentially between the medial and lateral tibial compartmentduring growth. Surprisingly, the compressive properties of articu-lar cartilage did not vary between the medial and lateral site and didnot depend on the age of the animals. Based on the mentioned the-ory of loading-induced functional adaptation during growth, thiswas unexpected. One could speculate that in parallel to the alter-ations in cartilage thickness and subchondral bone during growthan adaptation in mechanical properties should also take place.However, our findings are in accordance with Athanasiou et al.(1991), Rasanen and Messner (1996) and our previously publishedobservation that cartilage thickness and compressive properties oftibial cartilage do not correlate positively (Hamann et al., 2013).This suggests that cartilage mechanical quality is determined ratherby its organization and composition than its thickness. Alterna-tively, differences in loading between the two compartments mightnot be large enough to stimulate a substantial change in carti-lage composition resulting in alterations in cartilage biomechanicalproperties. However, the strong coupling between the subchon-dral bone and articular cartilage should be taken into account. Itseems that with increasing age and the subsequent developmentof the subchondral bone, the cartilage compressive properties arereduced as demonstrated by the negative correlations between theaggregate modulus and subchondral bone parameters.

There are a few limitations which should be mentioned here. Inour biomechanical measurements we analyzed rat tibial cartilageregions slightly posterior to the respective compartment assumedto be the load-bearing area. Because of the well-known varietyof articular cartilage thickness and compressive properties overregions and joint surfaces, even small variations in testing locationscould significantly influence the outcome of the biomechanicalmeasurements, especially for smaller joint surfaces. Therefore,quite large variances are commonly seen in other studies analyzingcartilage biomechanical properties (aggregate modulus) in animalssuch as rats (Athanasiou et al., 2000; Wang et al., 2006) or rabbits(Athanasiou et al., 1991; Roemhildt et al., 2006). Our variances arein the same range as those reported earlier. However, this couldbe a plausible explanation for the missing differences in the aggre-gate modulus between the two compartments. Furthermore, themethod may not be sensitive enough and may have underesti-mated minor differences in biomechanics of very thin rat cartilageat different locations. In addition, in our study design histological orbiochemical analyses of the articular cartilage from the medial andlateral compartments were not included, which could potentiallyhelp to explain the lacking topographical differences in cartilagebiomechanics. Nevertheless, this study is the first one investigatingtopographical variations in cartilage and subchondral bone withinthe normal growing rat knee joint. By reviewing the existing liter-ature the results provided are very unique and give novel insightinto the age-related topographical heterogeneity and relationship

of articular cartilage and subchondral bone in the rat tibia.

In summary, our study shows that there is considerableage-dependent topographical variation in articular cartilage andsubchondral bone within the normal rat knee. Cartilage thickness

Page 7

2 f Ana

igccdwstavtOkmdji

A

vO

A

f2

R

A

A

B

B

B

B

B

B

B

D

E

F

F

G

G

84 N. Hamann et al. / Annals o

s initially homogenously designed, but then differentiates duringrowth and maturation resulting in a higher thickness at the medialompared to the lateral tibial compartment. In contrast, cartilageompressive properties did not vary between both sites indepen-ently of age. Topographical heterogeneity of subchondral boneas already prominent in the 7-week-old animals and also per-

isted in the mature animals with higher values (subchondral platehickness, trabecular bone volume ratio and trabecular thickness)t the medial compared to the lateral site. In contrast, medial/lateralariations in porosity of subchondral plate were only present inhe immature animals and disappeared when reaching maturity.ur findings highlight that topographical differences within the ratnee joint are strongly related to the development status of the ani-als. Hence, our study demonstrated that local and site-specific

ifferences have to be considered when analyzing rat models ofoint diseases. Further, the age of the animals might be of particularmportance when different studies are to be compared.

cknowledgments

The study was supported by a grant of the German Sport Uni-ersity Cologne. The authors gratefully acknowledge Dr. Kai Danielberländer for his support in MATLAB.

ppendix A. Supplementary data

Supplementary data associated with this article can beound, in the online version, at http://dx.doi.org/10.1016/j.aanat.014.04.006.

eferences

thanasiou, K.A., Rosenwasser, M.P., Buckwalter, J.A., Malinin, T.I., Mow, V.C., 1991.Interspecies comparisons of in situ intrinsic mechanical properties of distalfemoral cartilage. J. Orthop. Res. 9, 330–340.

thanasiou, K.A., Zhu, C.F., Wang, X., Agrawal, C.M., 2000. Effects of aging and dietaryrestriction on the structural integrity of rat articular cartilage. Ann. Biomed. Eng.28, 143–149.

otter, S.M., van Osch, G.J., Waarsing, J.H., Day, J.S., Verhaar, J.A., Pols, H.A., vanLeeuwen, J.P., Weinans, H., 2006. Quantification of subchondral bone changesin a murine osteoarthritis model using micro-CT. Biorheology 43, 379–388.

ourrin, S., Palle, S., Pupier, R., Vico, L., Alexandre, C., 1995. Effect of physical train-ing on bone adaptation in three zones of the rat tibia. J. Bone Miner. Res. 10,1745–1752.

ouxsein, M.L., Boyd, S.K., Christiansen, B.A., Guldberg, R.E., Jepsen, K.J., Muller, R.,2010. Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. J. Bone Miner. Res. 25, 1468–1486.

rama, P.A., Tekoppele, J.M., Bank, R.A., Barneveld, A., van Weeren, P.R., 2000.Functional adaptation of equine articular cartilage: the formation of regionalbiochemical characteristics up to age one year. Equine Vet. J. 32, 217–221.

rommer, H., Brama, P.A., Laasanen, M.S., Helminen, H.J., van Weeren, P.R., Jurvelin,J.S., 2005. Functional adaptation of articular cartilage from birth to maturityunder the influence of loading: a biomechanical analysis. Equine Vet. J. 37,148–154.

uie, H.R., Campbell, G.M., Klinck, R.J., MacNeil, J.A., Boyd, S.K., 2007. Automatic seg-mentation of cortical and trabecular compartments based on a dual thresholdtechnique for in vivo micro-CT bone analysis. Bone 41, 505–515.

urr, D.B., 1998. The importance of subchondral bone in osteoarthrosis. Curr. Opin.Rheumatol. 10, 256–262.

edrick, D.K., Goldstein, S.A., Brandt, K.D., O‘Connor, B.L., Goulet, R.W., Albrecht, M.,1993. A longitudinal study of subchondral plate and trabecular bone in cruciate-deficient dogs with osteoarthritis followed up for 54 months. Arthritis Rheum.36, 1460–1467.

llender, G., Feik, S.A., Carach, B.J., 1988. Periosteal structure and development in arat caudal vertebra. J. Anat. 158, 173–187.

irth, E.C., Rogers, C.W., 2005. Musculoskeletal responses of 2-year-old Thorough-bred horses to early training. 7. Bone and articular cartilage response in thecarpus. N. Z. Vet. J. 53, 113–122.

roimson, M.I., Ratcliffe, A., Gardner, T.R., Mow, V.C., 1997. Differences inpatellofemoral joint cartilage material properties and their significance to theetiology of cartilage surface fibrillation. Osteoarthritis Cartilage 5, 377–386.

ardner-Morse, M., Badger, G., Beynnon, B., Roemhildt, M., 2013. Changes in in vitrocompressive contact stress in the rat tibiofemoral joint with varus loading. J.Biomech. 46, 1216–1220.

riffin, T.M., Guilak, F., 2005. The role of mechanical loading in the onset and pro-gression of osteoarthritis. Exerc. Sport Sci. Rev. 33, 195–200.

tomy 196 (2014) 278–285

Hagihara, Y., Fukuda, S., Goto, S., Iida, H., Yamazaki, M., Moriya, H., 2005. How manydays per week should rats undergo running exercise to increase BMD. J. BoneMiner. Metab. 23, 289–294.

Hamann, N., Zaucke, F., Dayakli, M., Bruggemann, G.P., Niehoff, A., 2013. Growth-related structural, biochemical, and mechanical properties of the functionalbone-cartilage unit. J. Anat. 222, 248–259.

Hamann, N., Zaucke, F., Heilig, J., Oberlander, K.D., Bruggemann, G.P., Niehoff, A.,2014. Effect of different running modes on the morphological, biochemical,and mechanical properties of articular cartilage. Scand. J. Med. Sci. Sports 24,179–188.

Hayes, W.C., Keer, L.M., Herrmann, G., Mockros, L.F., 1972. A mathematical analysisfor indentation tests of articular cartilage. J. Biomech. 5, 541–551.

Helminen, H.J., Hyttinen, M.M., Lammi, M.J., Arokoski, J.P., Lapvetelainen, T., Jurvelin,J., Kiviranta, I., Tammi, M.I., 2000. Regular joint loading in youth assists in theestablishment and strengthening of the collagen network of articular cartilageand contributes to the prevention of osteoarthrosis later in life: a hypothesis. J.Bone Miner. Metab. 18, 245–257.

Hoch, D.H., Grodzinsky, A.J., Koob, T.J., Albert, M.L., Eyre, D.R., 1983. Early changes inmaterial properties of rabbit articular cartilage after meniscectomy. J. Orthop.Res. 1, 4–12.

Iwamoto, J., Yeh, J.K., Aloia, J.F., 1999. Differential effect of treadmill exercise on threecancellous bone sites in the young growing rat. Bone 24, 163–169.

Janusz, M.J., Bendele, A.M., Brown, K.K., Taiwo, Y.O., Hsieh, L., Heitmeyer, S.A., 2002.Induction of osteoarthritis in the rat by surgical tear of the meniscus: inhibitionof joint damage by a matrix metalloproteinase inhibitor. Osteoarthritis Cartilage10, 785–791.

Jurvelin, J., Kiviranta, I., Tammi, M., Helminen, H.J., 1986. Effect of physical exerciseon indentation stiffness of articular cartilage in the canine knee. Int. J. SportsMed. 7, 106–110.

Khan, K., McKay, H.A., Haapasalo, H., Bennell, K.L., Forwood, M.R., Kannus, P., Wark,J.D., 2000. Does childhood and adolescence provide a unique opportunity forexercise to strengthen the skeleton? J. Sci. Med. Sport 3, 150–164.

Kiviranta, I., Tammi, M., Jurvelin, J., Saamanen, A.M., Helminen, H.J., 1988. Moder-ate running exercise augments glycosaminoglycans and thickness of articularcartilage in the knee joint of young beagle dogs. J. Orthop. Res. 6, 188–195.

Korhonen, R.K., Laasanen, M.S., Toyras, J., Rieppo, J., Hirvonen, J., Helminen, H.J.,Jurvelin, J.S., 2002. Comparison of the equilibrium response of articular cartilagein unconfined compression, confined compression and indentation. J. Biomech.35, 903–909.

Lee, J.H., Chun, K.J., Kim, H.S., Kim, S.H., Han, P., Jun, Y., Lim, D., 2012. Alter-ation patterns of trabecular bone microarchitectural characteristics induced byosteoarthritis over time. Clin. Interv. Aging 7, 303–312.

Lories, R.J., Luyten, F.P., 2010. The bone-cartilage unit in osteoarthritis. Nat. Rev.Rheumatol. 7, 43–49.

Malekipour, F., Whitton, C., Oetomo, D., Lee, P.V., 2013. Shock absorbing ability ofarticular cartilage and subchondral bone under impact compression. J. Mech.Behav. Biomed. Mater. 26, 127–135.

Malinin, T., Ouellette, E.A., 2000. Articular cartilage nutrition is mediated by sub-chondral bone: a long-term autograft study in baboons. Osteoarthritis Cartilage8, 483–491.

Marker, C.L., Pomonis, J.D., 2012. The monosodium iodoacetate model of osteoarthri-tis pain in the rat. Methods Mol. Biol. 851, 239–248.

Marks, S.C., Hermey, D.C., 1996. The structure and development of bone. In:Bilezikian, J.P., Raisz, L.G., Rodan, G.A. (Eds.), Principles of Bone Biology. AcademicPress, New York (NY), pp. 3–14.

McErlain, D.D., Ulici, V., Darling, M., Gati, J.S., Pitelka, V., Beier, F., Holdsworth, D.W.,2012. An in vivo investigation of the initiation and progression of subchondralcysts in a rodent model of secondary osteoarthritis. Arthritis Res. Ther. 14, R26.

Mohan, G., Perilli, E., Kuliwaba, J.S., Humphries, J.M., Parkinson, I.H., Fazzalari, N.L.,2011. Application of in vivo micro-computed tomography in the temporalcharacterisation of subchondral bone architecture in a rat model of low-dosemonosodium iodoacetate-induced osteoarthritis. Arthritis Res. Ther. 13, R210.

Murray, R.C., Vedi, S., Birch, H.L., Lakhani, K.H., Goodship, A.E., 2001. Subchondralbone thickness, hardness and remodelling are influenced by short-term exercisein a site-specific manner. J. Orthop. Res. 19, 1035–1042.

Oettmeier, R., Arokoski, J., Roth, A.J., Helminen, H.J., Tammi, M., Abendroth, K., 1992.Quantitative study of articular cartilage and subchondral bone remodeling inthe knee joint of dogs after strenuous running training. J. Bone Miner. Res. 7(Suppl. 2), S419L 424.

Radin, E.L., Rose, R.M., 1986. Role of subchondral bone in the initiation and progres-sion of cartilage damage. Clin. Orthop. Relat. Res. 213, 34–40.

Rasanen, T., Messner, K., 1996. Regional variations of indentation stiffness and thick-ness of normal rabbit knee articular cartilage. J. Biomed. Mater. Res. 31, 519–524.

Roemhildt, M.L., Coughlin, K.M., Peura, G.D., Fleming, B.C., Beynnon, B.D., 2006. Mate-rial properties of articular cartilage in the rabbit tibial plateau. J. Biomech. 39,2331–2337.

Sniekers, Y.H., Intema, F., Lafeber, F.P., van Osch, G.J., van Leeuwen, J.P., Weinans,H., Mastbergen, S.C., 2008. A role for subchondral bone changes in the processof osteoarthritis; a micro-CT study of two canine models. BMC Musculoskelet.Disord. 9, 20.

Tsai, W.Y., Wu, J.L., Liu, C.C., Cherng, C.H., Tsai, R.Y., Jean, Y.H., Wong, C.S., 2013. Early

intraarticular injection of hyaluronic acid attenuates osteoarthritis progressionin anterior cruciate ligament-transected rats. Connect. Tissue Res. 54, 49–54.

van Turnhout, M.C., Schipper, H., Engel, B., Buist, W., Kranenbarg, S., van Leeuwen, J.L.,2010. Postnatal development of collagen structure in ovine articular cartilage.BMC Dev. Biol. 10, 62.

Page 8

f Anat

W

W

W

X

N. Hamann et al. / Annals o

ancket, L.M., Baragi, V., Bove, S., Kilgore, K., Korytko, P.J., Guzman, R.E., 2005.Anatomical localization of cartilage degradation markers in a surgically inducedrat osteoarthritis model. Toxicol. Pathol. 33, 484–489.

ang, L., Kalu, D.N., Banu, J., Thomas, J.B., Gabriel, N., Athanasiou, K., 2006. Effectsof ageing on the biomechanical properties of rat articular cartilage. Proc. Inst.Mech. Eng. H 220, 573–578.

olff, J., 1892. Das Gesetz der Transformation der Knochen. Reprint der Aus-gabe Berlin, Hirschwald 1892. Wessinghage DJ (Hrsg.). Verlag Schattauer,Stuttgart.

ie, L., Lin, A.S., Kundu, K., Levenston, M.E., Murthy, N., Guldberg, R.E., 2012. Quan-titative imaging of cartilage and bone morphology, reactive oxygen species,

omy 196 (2014) 278–285 285

and vascularization in a rodent model of osteoarthritis. Arthritis Rheum. 64,1899–1908.

Xie, L., Lin, A.S., Levenston, M.E., Guldberg, R.E., 2009. Quantitative assessment ofarticular cartilage morphology via EPIC-microCT. Osteoarthritis Cartilage 17,313–320.

Xie, Z., Liachenko, S., Chiao, P.C., Carvajal-Gonzalez, S., Bove, S., Bocan, T., 2010.

In vivo MRI assessment of knee cartilage in the medial meniscal tear modelof osteoarthritis in rats. Med. Image. Comput. Comput. Assist. Interv. 13, 57–64.

Yamaguchi, S., Aoyama, T., Ito, A., Nagai, M., Iijima, H., Zhang, X., Tajino, J., Kuroki, H.,2013. Effects of exercise level on biomarkers in a rat knee model of osteoarthritis.J. Orthop. Res. 31, 1026–1031.