Assessment of specific mRNA levels in cartilage regions in a lapine model of osteoarthritis

~~

ELSEVIER

Jouriial of Orthopaedic Research 20 (2002) 535-544

Journal of Orthopaedic Research www.elsevier.com/locate/orthres

Assessment of specific mRNA levels in cartilage regions in a lapine model of osteoarthritis Marie-Pierre Hellio Le Graverand a, Jonna Eggerer a, Eric Vignon Ivan G. Otterness ', Leona Barclay a, David A. Hart '' Fuciilty of

b,

Medicine, McCciig Center for Joint Injirry and Arthritis Reseurcli, Utiiversiiy of Cdgary HSC, 3330 Hospitul Drive N . W., Colgar)), Alberiu, Ccinodu T2N 4 N I Foculiy of Medicine. Cloude Beriiurd University, Ceritre Hospiiolier Lj~on-Sud,69945 Lyon, France Cenrrul Research, P J z r Inc. Groton, CT, USA

Received 22 February 2001; accepted 14 August 2001

Abstract

Osteoarthritis (OA) is the most common form of arthritis and patients with meniscal and ligament injuries of the knee are at high risk to develop the disease. The purpose of this study was to evaluate molecular and structural changes occurring in four articular cartilage (AC) regions from the knees of anterior cruciate ligament (ACL)-transected rabbits at 3 and 8 weeks post-surgery. Rabbit AC from the lateral and medial femoral condyles (LFC and MFC) as well as from the medial and lateral tibia1 plateau (MTP and LTP) were processed for histology and for semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) analysis for a subset of relevant molecules (collagen 11, aggrecan, biglycan, decorin, fibromodulin, MMP-1,-3,-13, and TIMP-1). While the most severe histological changes were observed in the MTP starting as early as 3 weeks post-ACL transection based on Mankin scores, histological examination demonstrated a progression of osteoarthritic changes in the MFC from 3 to 8 weeks post-surgery. In contrast, very few changes were observed within both the LFC and LTP, and these changes did not worsen with increasing time after surgery. The water content increased significantly in the MFC at 8 weeks post-ACL transection and at both 3 and 8 weeks post-ACL transection in the MTP. Significant decreases in DNA content were observed for the MFC, LTP and MTP at 8 weeks post-ACL transection. Total RNA yields from the MFC and MTP were significantly elevated at 8 weeks post-ACL transection, while in the lateral compartment total RNA was unchanged following ACL transection. Analysis of mRNA levels for a subset of matrix molecules, proteinases and proteinase inhibitors, by RT-PCR demonstrated significant region-specific changes at the mRNA level following ACL transection. These results show that following ACL transection, complex molecular, as well as structural changes occur early in cartilage and that the observed changes are both region-specific and time-dependent. 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. Kqworrls: Articular cartilage surfaces; Histology; mRNA levels; Experimental osteoarthritis

Introduction Several experimental animal models have been developed for the study of the pathogenesis of osteoarthritis (OA). One of the most widely used experimental procedures t o induce joint instability a n d subsequent O A is the transection of the anterior cruciate ligament (ACL). A C L transection in the d o g is a recognized model for OA a n d has been extensively studied [2,12,14,19]. However, the natural history of OA in this canine model is fairly slow since full-thickness loss of *Corresponding author. Tel.: + I-403-220-4571/6885; fax: +1-403283-7742. f?-177tii/ address: [email protected] (D.A. Hart).

articular cartilage (AC) develops approximately 54 months after ACL transection. In contrast, recent studies have reported more rapid effects of ACL transection in the rabbit [17,21]. In approximately 6 to 24 weeks, transection of the rabbit A C L leads to gradual a n d progressive development of osteoarthritic changes in AC of the knee a t the gross morphological, histological a n d biochemical levels. Changes in gross appearance include fibrillation a n d erosion of AC, as well as osteophyte development [21]. Histological sections show fissures and chondrocyte cloning [2 11. However, there is currently no data that describe changes in chondrocyte metabolism at different stages during the development of the OA-like cartilage lesions in this experimental model.

0736-0266102/$ - see front matter 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 1 ) 0 0 1 2 6 - 7

536

M. -P. Hellio Le Grauerrirzd et (11. / J u i m u I of Orthupuetlic Resrcirch 20 (2002) 535-544

As the quantity of cartilage available from the rabbit knee is limited, and biochemical analysis not readily feasible, semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) has been demonstrated to provide accurate and reproducible assessments of cell activity at the mRNA level [4]. Results obtained in previous studies indicated that the method can be readily used to analyze RNA from rabbit AC [4,6]. It was further demonstrated that differences in RNA levels exist between cartilage regions of the rabbit knee joint, a finding which emphasizes the risk of pooling samples from distinct regions of the knee joint for molecular analysis [6].Moreover, these regional differences in RNA levels could be important in our understanding of cartilage responses in situations leading to OA development, where area-specific alterations in cartilage biology and biochemistry are well documented in the human [20,22]. Therefore, the objective of the present study was to assess mRNA levels for a subset of relevant molecules (collagen 11, aggrecan, biglycan, decorin, fibromodulin, MMP-1,-3,-13, and TIMP-1) within four cartilage regions of the rabbit knee joint, namely the lateral and medial femoral condyles (i.e., LFC and MFC), as well as the lateral and medial tibial plateau (i.e., LTP and MTP) of non-operated rabbits (age-matched controls), and ACL-transected rabbits at 3 and 8 weeks post-surgery. The panel of mRNA species to be assessed was determined based on their known role in maintaining cartilage function (i.e., matrix molecules, proteinases and proteinase inhibitor). While not all inclusive, the molecules assessed are representative of cellular activities known to influence cartilage function.

Materials and methods Esyeririierital OA

Skeletally mature (12 nlonths of age) female New Zealand white rabbits (Reimans Fur Ranch; St. Agatha, ON) were used in the present study (N = 42). All experimental rabbits undergoing ACL transection had surgery under sterile conditions using a general inhalational anaesthetic (1% halothane and oxygen at I I/niin). The surgical procedure has been described in detail previously [lo]. After surgery, animals were allowed unrestricted cage activity until sacrifice at 3 and 8 weeks. Age-matched control animals were also sacrificed. In each group (controls, 3 weeks post-transection, 8 weeks post-transection), N = 14. Of the total of 42 rabbits used in the study, 18 were used for RNA analysis, 18 for water content/DNA determination, and 6 for histology. Animals were housed in the Animal Resource Centre (Faculty of Medicine) in accordance with Canadian Council on Animal Care Guidelines and with the approval of the Animal Care Committee of the Faculty of Medicine. Tissue y reyarn t iori

All animals were sacrificed by Euthanyl overdose (MTC Pharniaceuticals; Cambridge. Ontario) intravenously through the lateral ear vein. For molecular studies, the cartilage from the M F C and LFC. as well as the M T P and LTP was carefully removed aseptically from the knee joint to avoid capsular contamination of the tissue. For DNA

content. the fresh wet weight was immediately determined, followed by lyophilization to constant weight, dry weight determination and subsequent D N A analysis. These values were also used to define the water content of the tissue samples. For RNA extraction, a separate set of fresh tissue samples were immediately weighed before freezing in liquid nitrogen and then stored at -80 "C prior to processing. RNA extruetion

Total R N A was isolated from the four individual cartilage regions of each animal using the TRIspin method as described previously [16]. Total RNA was quantified using the SyBr green reagent (FMC BioProducts; Rockland, ME) and a Perkin Elmer fluorimeter. All samples were stored at -80 "C until analyzed. In each group, N = 6.

Reverse transcr@tion and seriiic~uurititutivrpolyriirrcise cliuin reactiori Reverse transcription (RT) was carried out with 1 pg total RNA using the Stratascript (tm) RT-PCR kit (PDI BioScience; Aurora, ON). Using rabbit specific primer sets described in previous studies [4,5], aliquots of cDNA were amplified by PCR as previously described [5,16]. The sequences for the primers and the conditions for their use are summarized in Table 1. Two independently isolated clones of each amplicon have been sequenced to verify the identity of the cDNA product and their specificity was confirmed by BLAST searches. For all reported experiments, conditions were determined to be in the linear range for both the PCR amplification and the image analysis system as described previously [4]. Briefly, for each group of samples ( i c , RNA from the MFC, LFC, MTP and LTP from operated and non-operated rabbits) all of the samples were subjected to R T at the same time and subsequently, all samples of cDNA amplified by PCR at the same time to avoid any potential experiment to experiment variation in efficiency. Each R T sample was initially assessed for GAPDH (housekeeping gene) cDNA. Levels of GAPDH mRNA did not vary with time after surgery. Aliquots of each R T sample were assessed for GAPDH following 21-22 PCR cycles, the volumes were then normalized to approximate equivalent amplicon densities, and the PCR repeated at 21-22 PCR cycles to ensure very similar GAPDH integrated density values were obtained. Once the GAPDH values were determined to be similar, and in the linear range of detection, the same volumes of each sample were then used to assess the cDNA levels for the remaining molecules of interest. Such an experimental construct allowed for comparisons between groups (i.e., operated vs non-operated). For each primer set, the optimal cycle number was determined and the resulting amplified bands analyzed by densitometry. The no-RT controls were negative for each primer set utilized indicating genomic DNA contamination was undetectable. Integrated density values for the genes in question were normalized to the GAPDH values to yield a semiquantitative assessment. Values for non-operated age-matched controls were set at 100%1and corresponding values for cartilage surfaces at 3 and 8 weeks post-ACL transection are expressed as percentage of control values in order to detect changes. D N A conlenl ciiwlj'sis

Using the method reported by Lipman [9],the DNA content was measured using a fluorophotometric assay as previously described. In each group, N = 6. Histolog),

After sacrifice, the whole femoral condyles and tibial plateaus were resected and immediately fixed in 10% formalin bufier (pH 7.4) containing 0.5% of cetylpyridinium chloride, decalcified with formic acid in formalin, dehydrated through graded alcohols, cleared with xylene. and embedded in paraffin. Consecutive coronal sections of the whole femoral condyle or tibial plateau were cut at 6 pm and mounted on aminoalkylsilane-treated slides. The sections were then stained with Safranin-O/fast green for proteoglycan staining and graded according to the Mankin scale [lo] by two experienced and blinded observers. For quantification of the surface area of the histopathological lesions (mm'), the number of consecutive sections showing histopathological lesions was multiplied by both the thickness of the sections and the width of the lesion. I n each group, N = 2.

M.-P. Hellio Le Graverand et al. I Journal of Orthopaedic Research 20 (2002) 535-544

537

Table 1 Details of the PCR primers used in this study Gene

Temperature OC

Cycle number

bp

Primer sequences

Primer source

Collagen I1

65

24

366

Metsaranta, M., Vuorio, E."

Aggrecan

65

26

313

Biglycan

60

28

406

Decorin

60

28

419

Fibromodulin

60

30

442

MMP-13

65

32

527

MMP-1

55

28

322

MMP-3

65

26

363

TIMP-1

60

26

326

GAPDH

55

22

293

GCACCCATGGACATTGGAGGG GACACGGAGTAGCACCATCG GAGGAGATGGAGGGTGAGGTCTTT CTTCGCCTGTGTAGCAGATG GATGGCCTGAAGCTCA A GGTTGTTGAAGAGGCTG TGTGGACA ATGGTTCTCTGG CCACATTGCAGTTAGGTTCC CTGGACCACAACAACCTGAC GGATCTTCTGCAGCTGGTG TTCGCTTAGAGGTGACAGG ACTCTTGCCGGTGTAGGTGT TCAGTTCGTCCTCACTCCAG TTGGTCCACCTGTCATCTTC GCCAAGAGATGCTGTTGATG AGGTCTGTGAAGGCGTTGTA GCAACTCCGACCTTGTCATC AGCGTAGGTCTTGGTGAAGC TCACCATCTTCCAGGAGCGA CACAATGCCGAAGTGGTCGT

Bayne, E.K. et Human, rat, mouse, cowc Zhan, Q.et Human, rat, mouse, cowc Vicenti et al.' Fini, M.E. et al.' Fini, M.E. et a1.g Horowitz, S. et aLh Applequist, S.E. et al.'

Personal communication. Genbank accession number L38480. Consensus sequences for these species were found from Genbank sequences. Genbank accession number 576584. Genbank accession number AF059201 fGenbank accession number M17821. Genbank accession number M25644. Genbank accession number 504712. Genbank accession number L23961.

a

Safranin-0 (Table 2 and Fig. 1). By 3 weeks post-ACL transection, the most severe changes were observed in the MTP (Fig. 1, Panel K and Table 1). While the Mankin score for the MTP did not vary significantly over time (9-10 at 3 weeks post-ACL transection versus 11-10 at 8 weeks post-ACL transection), the extent of the histopathological lesions doubled in surface area (0.59 mm2 at 3 weeks post-ACL transection to 1.24 mm2 at 8 weeks post-ACL transection). The second most altered articular region was the MFC, demonstrating significant changes from 3 (Panel E) to 8 (Panel F) weeks post-ACL transection [Mankin score: 6-9 and tripling of the area of the altered surface (0.36 mm2 at 3 weeks post-ACL transection to 1.1 mm2 at 8 weeks post-ACL transection)]. In contrast, very few changes were observed within both the LFC and LTP and these changes did not

Statistical analysis

Water content, DNA and total RNA values are given as mean fstandard deviation for six rabbitslgroup. Comparisons were made using one-way ANOVA and the BonFerroni test was subsequently applied as a post-test when comparing pairs of group means. Differences with p values of less than 0.05 were considered significant.

Results Histopathology

Following ACL transection, all of the samples of AC from the four regions revealed histologic changes characteristic of OA, including surface disruption, hypocellularity, and areas exhibiting diminished staining with

Table 2 Histopathological scores of the various cartilage surfaces using the Mankin classification Specimen number

Surgery

LFC

MFC

LTP

1 2 3 4 5 6

No No 3 weeks post-ACL transection 3 weeks post-ACL transection 8 weeks post-ACL transection 8 weeks post-ACL transection

1 1

1

1

1

6 6 6 6

6 6 9 9

1 5 5

5 5

LTP 1 1

9 10 11 10

538

M.-P. Hellio Lr Gravercmd et al. I Journul of Orthopaedic Researell 20 (2002) 535-544

Fig. 1. Histology of the four cartilage regions from control, and 3 and 8 weeks post-ACL transected rabbit knees. Representative histology observed in the LFC, MFC, LTP, and MTP from control animals (Panels A, D, G, J), and in animals at 3 weeks (Panels B, E, H, K) and 8 weeks (Panels C, F, I, L) post-ACL transection. Original magnification, l o x .

worsen with time after surgery (0.5 mm2 at both 3 and 8 weeks post-ACL transection). Water content

The percentage of water content of the different cartilage regions are presented in Table 3. The water content increased significantly in the MFC at 8 weeks post-ACL transection (p < 0.01) and at both 3 and 8 weeks post-ACL transection in the MTP (p < 0.01).

Similarly, there was a trend toward an increase in water content in both the LFC and LTP after ACL transection, however the differences between the operated and non-operated animals were not significant. Concentrations of DNA

The mean concentration of DNA per mg dry weight of tissue for the four cartilage regions are given in Table 3. The values obtained for the control tissues is similar

S.D. p values (BonFerroni post-test) DNA p g h g dry 5.39 5.57 weight S.D. 1.13 1.02 p values (BonFerroni NS post-test) RNA pg/rng wet 0.27 0.25 weight S.D. 0.13 0.06 p values (BonFerroni NS post-test ) a 3 weeks post-ACL-T vs Controls. 8 weeks post-ACL-T vs Controls. ' 3 weeks post-ACL-T vs 8 weeks post-ACL-T. * p < 0.05: ** p < 0.01; *** p < 0.001. 1.03 0.025'.* 0.29

0.49

0.2 0.05

0.82

0.35 0.12

0.51

0.74

3.86

70.75 1.76

8 weeks

0.05 0.17 0~0004a.*~*,c.*~

5.16

4.96

4.27

68.48 1.42 0.0014".'*

3 weeks

65.4 2.69

70.17 3.37

Control

65.61 1.34 NS

8 weeks

3 weeks

Control

67.13 3.16

'%I Water content

Post-ACL transection

Post-ACL transection

Groups

MFC

LFC

Articular cartilage regions

Table 3 Water Content. DNA per mg dry weight and total RNA concentration per mg wet weight of cartilage regions

0.04

0.21

1.42

5.35

68.78 1.96

Control

0.2 NS

0.3

0.58 0.05".'

4.88

67.90 1.68 NS

3 weeks

Post-ACL transection

LTP

0.15

0.38

0.30

3.96

69.41 3.20

8 weeks

0.04

0.14

0.42

4.18

69.15 1.17

Control

o,001ya.*.*.c.*-

0.03

0.14

0.34

0.16

0.2 0.078".' 0.15

3.74

71.76 0.47

8 weeks

4.16

71. 61 1.19 0.0006b.-*.a.*=

3 weeks

Post-ACL transection

MTP

c ru

fibromodulin), matrix iiietalloproteinases: MMP- 1 (collagenase-1), MMP-3 (stroinelysin-I) and MMP-13 (collagenase-3), as well as the proteinase inhibitor TIMP- 1 from cartilage surfaces of ACL-transected and nonoperated control rabbits were analyzed by seniiquantitative RT-PCR. Figs. 2 and 3 illustrate changes in mRNA levels of ACL-transected rabbits compared to mRNA values of control animals for each of the molecules studied within each of tlie four AC regions.

to those reported by Lipman [9]. While there was a nonsignificant trend toward a decrease in DNA content for the LFC after ACL transection, significant decreases in DNA content were observed for the MFC, LTP and MTP at 8 weeks post-ACL transection (u < 0.05).

The mean concentration of total RNA per nig wet weight of tissue are summarized in Table 3. Total RNA levels i n both the MFC and MTP were sigliificantly elevated to 255% and 243'%1of control values at 8 weeks post-ACL transection ( p .= 0.001 and y < 0.01, respectively). Total RNA Ievels also showed a trend toward an increase in both the LFC and LTP at 8 weeks post-ACL transection, however the differences from control tissue total RNA values were not statistically significant.

Transcripts for type I1 collagen were significantly elevated in the LFC at 8 weeks post-ACL transection ( p < 0.05) and at both 3 and 8 weeks post-ACL transection in both the LTP ( p < 0.01 and p < 0.001, respectively) and MTP O-, < 0.01 and p < 0.05, respectively). mRNA levels for type I1 collagen exhibited a trend toward an increase in the MFC following ACL tl-ansection, however the differences from control mRNA levels were not significant. 111 contrast, aggrecan mRNA levels were significantly increased only in tlie MFC of

R T-PCR nnolysis Trailscript levels for structural niatrix macroniolecules (type I1 collagen, aggrecan, biglycan, decorin and

LFC

MFC b"'

250

b' 2 200

-f 200 -9e 150

T

;100 ;50 CI

(c

0

COL2

Aggrecan Biglycan Decorin Fibromodulin

COL2 Aggrecan Biglycan Decorin Fibromodulin

MTP

LTP 500

b***

$400

-a -300

u)

500

a**

400

T

-a9 300 -e

2

CI

b'

c

g 200

i200

u

(c

0

(c

$100

s0 100

0

0

COL2 Aggrecan Biglycan Decorin Fibromodulin

COL2 Aggrecan Biglycan Decorin Fibromodulin

0 Controls; 61 3 weeks post-ACL-T;

8 weeks post-ACL-T

(a: 3 weeks post-ACL-T vs Controls; b: 8 weeks post-ACL-T vs Controls; c: 3 weeks post-ACL-T vs 8 weeks post-ACL-T) Fig. 2. Inliuence of' ACL transection on inRNA levels for matrix macromolecules in the our cartilage regions. mRNA levels were deteriiiined by semiquantitative RT-PCR ;IS described in Materials aiid Methods. The mean v;ilues & S.D. values for each molecule in every cnrtilage region are presented as percent of non-operated control vnlues ( N = (,/group). All values indicated by an * are signific;lntly dilTerent between the two indicated cartilage icgions (*: p < 0.05; **: p < 0.01: * 4: *: 11 < 0.001).

54 I

LFC 60C

T

-

:.IAIL g 2000

r

a**

c

$100

0

MMP-I3

MMP-1

MMP-3

TIMP-1

MMP-13

500

MMP-3

TIMP-1

MTP

LTP 40001

MMP-1

1

a"

r"'

40001

-

3

400

1

2 3000

-z 100

2

2000

8 b

1000

s 0

MMP-13

MMP-1

MMP-3

TIMP-1

MMP-13

0 Controls; 0 3 weeks post-ACL-T;

MMP-1

MMP-3

TIMP-I

8 weeks post-ACL-T

(a: 3 weeks post-ACL-T vs Controls; b: 8 weeks post-ACL-T vs Controls; c: 3 weeks post-ACL-T vs 8 weeks post-ACL-T) Fig. 3. Effect of ACL transection on mRNA levels for MMPs and TIMP-I in the four cartilage regions. iiiRNA levels were determined by semiquantitative RT-PCR as described in Materials and Methods. The mean values fS.D. values for each molecule in every cartilage region are presented as percent of non-operated control values (N = 6Igroup). All values indicated by an * are significantly different between the two indicated cartilage region (*: p < 0.05;**: p < 0.01; * * *: p < 0.001).

the operated rabbits at 8 weeks post-surgery (p < 0.001). Transcripts for biglycan were only significantly elevated at 8 weeks post-ACL transection in the MTP (p < 0.05). In contrast, niRNA levels for decorin in the MFC were dramatically depressed to 10% and 9Yn of control values at 3 and 8 weeks post-ACL transection (p < 0.001). In addition, mRNA levels for decorin in the LTP were depressed to 43% of control values at 3 weeks post ACL transection (u < 0.01). Decorin niRNA levels did not vary in the LFC or the MTP following ACL transection. At 3 and 8 weeks post-ACL transection, fibroinodulin mRNA levels were significantly depressed in the M F C (p < 0.001) and MTP ( p < 0.01 and p < 0.05, respectively), while niRNA levels for this molecule did not vary in the LFC and LTP following ACL transection. Proteirinses mid Iwotriiitrse irihihitor

Dramatic increases in niRNA levels for MMP-13 were observed only at 3 weeks post-ACL transection in all four cartilage regions. At 3 weeks post-ACL tran-

section, mRNA levels for MMP- 13 were significantly switched on to 1439% of control values in the LFC (p < 0.01), to 3130% of control values in the MFC 0, < 0.01), to l668Y0 of control values in the LTP ( p < 0.05) and to 2863% of control values in the MTP (p < 0.05). These large percentage increases are due to the very low levels of MMP-13 mRNA observed in AC in the control animals at this age. At 8 weeks post-ACL transection, MMP-13 mRNA levels decreased to values that were not significantly different from control values, primarily due to considerable animal to aniinal variation at this time point. I n contrast, MMP-1 (collagenase-I ) mRNA levels were significantly elevated only in the LFC of operated rabbits at 8 weeks post-ACL transection (u < 0.05), whereas mRNA levels for MMP-1 were i i n cliaiiged in the other cartilage surfaces after ACL transection. Transcript levels for MMP-3 (stromelysin-1) were not modified after ACL trailsection in either MFC nor the MTP. In contrast, a t 3 weeks post-ACL transection. transcript levels for MMP-3 were significantly elevated in both the LFC and LTP ( p < 0.01 and

p < 0.05, respectively), but by 8 weeks post-ACL transection niRNA levels for MMP-3 returned to control values in both of these cartilage surfaces. At 3 weeks postACL transection, mRNA levels for TIMP-I were significantly elevated only in the LTP ( p < 0.05) and MTP ( p < 0.001), while by 8 weeks post-ACL transection mRNA levels for TIMP-1 had returned to control values.

Discussion The results presented in this report demonstrate that transection of the rabbit knee ACL leads to molecular and structural changes in different cartilage regions of the knee joint, and that the observed changes are both region-specific and time-dependent. Analysis of the histopathological changes denionstrated that the progression of tlie osteoarthritic lesions within each of the cartilage regions was not constant with time after ACL transection among different cartilage regions from the same joint. Indeed, at 3 weeks post-ACL transection the MTP already exhibited advanced alterations, while the femoral condyles and the LTP demonstrated fewer changes. By 8 weeks post-ACL transection, while the disease did not progress in the lateral compartment, the MFC and the MTP presented similar significant levels of osteoarthritic changes at both the Mankin score and with respect to increases in the surface area altered. The results presented in this report also indicate that following ACL transection, significant changes occur with regard to water content, DNA content and total RNA content within AC from different regions of the rabbit knee joint. The observed increases in water and RNA content, as well as decreases in DNA content with time after ACL transection confirm and extend similar observations previously reported in various experiniental models of OA [2,19]. The dramatic increases in water content observed at 8 weeks post-ACL transection in both tlie MFC and MTP parallel the alteration of the extracellular matrix as indicated by the highest histo-

pathological scores recorded at that period of time (MFC: 9-10 and MTP: 11-10). The normal integrity of AC is maintained by resident chondrocyte metabolism. Therefore, in this rabbit model of OA, a characteristic decrease in cell number with an apparent cell “liypermetabolisni” occurred in both the MFC and MTP early after ACL transection. From the present study, it appears that most of the changes occur in the medial compartment (i.e., MFC and MTP) compared to the lateral compartment (i.e,, LFC and LTP), which is in agreement with previous findings demonstrating dramatic early alterations of the medial meniscus compared to the lateral meniscus in this rabbit model of OA [7]. The complex changes in mRNA levels observed following ACL transection within each of the cartilage regions studied (summarized in Table 4), likely reflect variations in the progression of the osteoarthritic changes, as indicated by the variations in histopathological scores. Considering the evident progression of the lesions observed in the M F C from 3 to 8 weeks post-ACL transection, analysis of the mRNA data from this AC region suggests that two of the molecules studied could potentially play a role in development of the osteoartliritic lesions following ACL transection. Providing that mRNA is translated into protein, the dramatic decreases in mRNA levels for both decorin and fibromodulin within the MFC, may be linked to cartilage tissue damage following ACL transection. Decorin and fibroniodulin are small proteoglycans which can interact with collagen fibrils and regulate fibrillogenesis [18]. A lack of decorin from the superficial zone of AC has been previously reported in experimental [l I], as well as in human OA [15]. This was contrasted by an increase in decorin content in the mid and deeper zones of osteoarthritic cartilage [ 151. While differences in reports for decorin expression may be related to variations in species and/or variations in stages of development of the disease, the potential lack of decorin may likely play a role in the enlargement of the collagen bundles detected in osteoarthritic cartilage [ 1 ll. However, further

Table 4 Siimmary of IiiRNA level changes following ACL transection Cartilage regions time after ACL-T

3 weeks Type I 1 collagen Aggrecan Biglycan Decorin Fibromodulin MMP-13 MMP-I MMP-3 TIMP-I

M FC

LFC

8 weeks

3 weeks

LTP 8 weeks

3 weeks

MTP

8 weeks

3 weeks

A -

A -

A A

8 weeks

comparison of changes in mRNA levels within cartilage rcgions between the present and other published molecular studies in experimental [I] and liuinan OA [3] is difficult due to both tlie techniques used (i.e., Northern analysis versus RT-PCR), as well as the origin of the cartilage sampled for molecular analysis (i.e., sampling of variable regions of AC within the knee joint and/or pooling of tissue from several cartilage regions). Another interesting finding of the present study was the observation of dramatic increases in inRNA levels for MMP-13 at 3 weeks post-ACL transection in all cartilage regions. This rapid and significant switching on of MMP- 13 mRNA was followed by a decline at 8 weeks post-ACL transection. While the apparent up-regulation of MMP-I 3 niRNA expression within AC seeins to be an early event following ACL transection, the absence of coordination between the observed increases in MMP- 13 mRNA levels in all of the cartilage surfaces and the histopathological score suggests that MMP-13 may be involved in the early remodeling of the matrix in response to ACL transection. Interestingly, it has recently been demonstrated that excessive MMP-13 activity results in osteoarthritic degradation of AC in transgenic mice in which tetracycline-regulated transcription in conjunction with a type I1 collagen promoter targets a constitutively active human MMP-13 to hyaline cartilage [13]. In the present study, ACL transection led to a transient increase in MMP-13 mRNA levels in all four cartilage regions, but the majority of the histopathologic lesioiis were restricted to the medial compartment of the joint. Thus, perhaps following ACL transection the release of an unknown soluble factor(s) leads to the rapid induction of MMP-13 in all of the cartilage regions, but it is only in areas where the bioniechanical environment is most altered do the histopathologic alterations develop. Therefore, the development of cartilage degeneration in this model may be the result of proteinase modification of the cartilage matrix plus an abnormal loading environment. This concept would be consistent with our previous studies demonstrating that following ACL transection, the medial meniscus rather than the lateral meniscus, is affected first [7]. I n conclusion, the present study demonstrates that complex inRNA changes occur in different AC regions of the same joint during the early phases of the knee joint responses to ACL transection. Each of the articular regions demonstrated specific alterations i n mRNA levels in responses to tlie altered biomeclianical environment and joint instability. Few changes were observed i n tlie lateral compartment following ACL transection. The most severe histopathological changes were observed in the medial compartment, which is consistent with observations from patients with OA which indicate more involvement of the medial compartment than tlie lateral compartment in the disease [ 8 ] . Further molecular analysis. and confirmation that the inRNA changes are

reflected by protein changes, may lead to new insights into the disease process and identification of additional therapeutic targets to. interfere with the development and progression of OA. Acknowledgements The authors acknowledge the support of the Canadian Institute for Health Research, The Arthritis Society, and The Canadian Arthritis Network for these studies. MPHLG was supported by a Fellowship from The Arthritis Society, JE was supported by a Summer Studentship from tlie Alberta Heritage Foundation for Medical Research, and DAH is the Calgary Foundation-Grace Glaum Professor in Arthritis Research.

References [ I ] A d a m ME, Matyas JR, Huaiig D, Dourado GS. Expression of proteoglycans and collagen in the hypertrophic phase of experimental osteoarthritis. J Rheuinatol Suppl 1995:43:94-7. [2] Brandt KD. Myers SL, Burr D , Albrecht M. Osteoartliriticclianges in canine articular ciirtilage subchondral bone and synovium fiftyfour months after transection of the anterior cruciate ligainent. Arthritis Rheum 1991:34:1560-70. [3] Cs-Szabo G. Melcliing LI. Roghley PJ, Glant TT. Changes in messenger RNA and protein levels of proteoglycitns and link protein in Iiuiiian osteoartliritic cartilage samples. Arthritis Rheum 1997;40:103745. [4] Hellio Le Graverand M. Reno C, Hart D. Influence of pregiiancy on gene expression in rabbit articular cartilage. Ostearthritits Cart 1998:6:341-50. [S] Hellio Le Graverand M, Reno C, Hart DA. Gene expression in menisci froiii the knees of skeletally immature and mature female rabbits. J Orthop Res 1999;17:73844. I61 Hellio Le Gi-aver:iid M, Reno C. Hart D. Heterogenous response of knee cartilage to pregnancy in tlie rabbit: assessment of specific inRNA levels. Osteoarthritis Cart 2000;S:53-62. [7] Hellio Le Graverand M, Vigiioii E. Otterness IG. Hart DA. Early changes in lapine menisci during osteoarthritis developmeiit Part I: cellular and matrix alternations. Osteoartliritis Cart 2001:9: 56-64. [S] Lewandrowski KU, Muller J. Schollmeier G. Colicomitant meniscal and articular cartilage lesions in the feniorotibial joint. Am J Sports Med 1997:25:486-94. [9] Lipinan JM. Fluoropliotometric qtraiititatioii of DNA in articular cartilage utilizing Hoechst 33258. J Anal Biochem 1989: 176: 128-31. [lo] Mankin HJ. Dorfinan H. Lippiello L, Zarins A. Bioclieniical and inetabolic abnor~i~ilities in articular cartilage from osteoarthritic human hips. II. Correlation of iiiorphology with biochemical and nietabolic data. J Bone Joint S ~ i r g[AM] 197 I S A : 523-37. [ I I] Mass6 PG, Carrino DA. Morris N , Wenger L. Mahuren JD, Howell DS. Loss of decorin from the surl'ace zone ofiirticular cartilage in a chick model of osteoartliritis. Acta Histocliein 1997;99: 43 1-14. [I21 McDevitt CA. Muir H. Biochemical changes in tlie cartilage ofthe knee in experimental and natural osteoarthritis in tlie dog. J Bone Joint Surg [Br] 1976:58:94-101. [I31 Neuhold LA. K i l l x L. Zhno W, Sung ML. Warner L. Kulik J , el ill. Postnatal expression in hyaline cartilage of constitntively

544 active human collagenase-3 (MMP-13) induces osteoarthritis i n mice. J Clin Invest 2001;107:3544. [I41 Pond MJ, Nuki G . Experimentally-induced osteoarthritis in the dog. Ann Rheum Dis 1973;32:387-8. [I51 Poole AR, Rosenberg LC, Reiner A, lonescu M, Bogoch E, Roughly PJ. Contents and distributions of the proteoglycans decorin and biglycan in norinal and osteoarthritic human articular cartilage. J Orthop Res 1996;14:681-9. [I61 Reno C, Marchuk L, Sciore P, FI-ank CB, Hart DA. Rapid isolation of total RNA from small samples of hypocellular dense connective tissues. Biotechniques 1997;22:1082-6. [I71 S a h RL. Yang AS, Hant JJ, Halili RB, Yoshioka M, Amiel D, et al. Physical properties of rabbit articular cartilage after transection of the anterior cruciate ligament. J Orthop Res 1997;15: 197-703.

[I 81 Scott JE, Orford CR, Hughes EW. Proteoglycan~ollagen arrangements in developing rat tail tendon. An electron microscopial and biochemical investigation. Biochem J 1981;195:57381. [I91 Vignon E, Arlot M.Hartmann D. Moyen B, Ville G. Hypertrophic repair of articular cartilage in experimental osteoarthrosis. Ann Rheum Dis 1983;42:82-8. 1201 Vigiion E, Arlot M, Patricot LM, Vignon G . The cell density of human femoral head cartilage. C h i Orthop Re1 Res 1976:303-8. [21] Vignon E, Bejui J, Mathieu P, Hartmann JD. Ville G, Evreux JC. et al. Histological cartilage changes in a rabbit model of osteoarthritis. J Rheum 1987;14 Spec:104-6. [22] Yoshida K, Azutna H. Contents and compositions of glycosaniinoglycans i n different sites of the hunian hip joint cartilage. Ann Rheum Dis 1982;41:512-9.