J. Comp. Path. 2011, Vol. -, 1e13 Available online at www.sciencedirect.com

Cor

002

doi

www.elsevier.com/locate/jcpa

Pl

A

Expression of Genes Encoding ExtracellularMatrixMacromolecules and Metalloproteinases in Avian

Tibial Dyschondroplasia

I. Velada*,†, F. Capela-Silva‡, F. Reisx, E. Pires{, C. Egask,P. Rodrigues-Santos† and M. T. Barros#

*Department of Biology, University of Aveiro, 3810-193 Aveiro, †Laboratory of Functional Genomics, Center for

Histocompatibility, 3001-301 Coimbra, ‡Institute of Mediterranean Agricultural Sciences (ICAM), University of �Evora,7002-554 �Evora, xInstitute of Pharmacology and Experimental Therapeutics, IBILI, Faculty of Medicine, University of

Coimbra, 3000-354 Coimbra, {Centre for Neuroscience and Cell Biology, University of Coimbra, 3004-517 Coimbra,kBIOCANT Park, 3060-197 Cantanhede and #Department of Health Sciences, Portuguese Catholic University, 3504-505

Viseu, Portugal

resp

1-99

:10.1

ease

vian

Summary

Avian tibial dyschondroplasia (TD) is a skeletal disease characterized by disruption of endochondral bone forma-tion. The aim of this study was to determine the expression of extracellular matrix (ECM) macromolecules andECM-degrading enzymes [matrix metalloproteinases (MMPs)] in the growth plates of normal and TD-affected3-week-old broiler chicks (Cobb strain). Protein levels were analyzed by immunoblotting and gelatin zymogra-phy and gene expression by polymerase chain reaction. Expression of genes encoding the ECMmacromolecules(collagen types II, IX, X andXI; and aggrecan) was not altered in dyschondroplasia; however, there was down-regulation of genes encoding MMP-9, MMP-13, MMP-10 and MMP-11 in addition to reduced amounts ofMMP-2 andMMP-13 proteins. In contrast, there was up-regulation of genes encodingMMP-7 and the vascularendothelial growth factor. These findings suggest that the accumulation of cartilage associated with the diseasemay be the result of decreased proteolysis due to the down-regulation of MMPs and not to an increased produc-tion of ECM macromolecules.

� 2010 Elsevier Ltd. All rights reserved.

Keywords: extracellular matrix; gene expression; growth plate; matrix metalloproteinases; tibial dyschondroplasia

Introduction

Long bones develop by the process of endochondralossification. Further elongation of long bones occursat the level of the growth plate cartilage until thetime that they can no longer grow in length. Thegrowth plate shows a highly complex and synchronousmechanism of continuous cell proliferation, cell en-largement or hypertrophy and cell removal. Thus,the growth plate can be divided into different regions,including the reserve, the proliferative, the hypertro-phic and the vascularization zones (Howell andDean, 1992). The avian growth plate contains longercolumns of cells than that of mammals and these be-

ondence to: I. Velada (e-mail: ivelada@gmail.com).

75/$ - see front matter

016/j.jcpa.2010.12.008

cite this article in press as: Velada I et al., Expression of Genes En

Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

come randomly oriented or absent from the hypertro-phic and calcifying zones. More cells are found ineach region of the avian growth plate, where metaphy-seal blood vessels penetrate more deeply (Leach andGay, 1987). Nevertheless, despite some structural dif-ferences between the avian and the mammaliangrowth plate, electron microscopical similarities sug-gest identical physiological control mechanisms. Infact, the process of endochondral bone growth hasbeen reported to be functionally similar in birds andmammals (Leach and Lilburn, 1992).

The chondrocytes of the growth plate are embeddedin an extracellular matrix (ECM), which supportsthem. The matrix is composed of ECMmacromolecules, ECM-remodelling enzymes and var-ious growth factors (Horton, 1993). Type II collagen

� 2010 Elsevier Ltd. All rights reserved.

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

2 I. Velada et al.

(collagen II) is the primary collagen species in thegrowth plate, which also expresses collagen types IX,X and XI (Ballock and O’Keefe, 2003). Growth platecartilage must maintain a tightly controlled balancebetween cartilage synthesis and degradation (Orth,1999). Matrix metalloproteinases (MMPs), a groupof ECM-remodelling enzymes, play a crucial role inECM remodelling and degradation and are involvedin the preservation of its integrity (Ortega et al.,2004). The ECMalso functions as a reservoir of variousgrowth factors, which are released to adjust chondro-cyte function when the ECM is degraded (van derEerden et al., 2003).

Tibial dyschondroplasia (TD) is a skeletal disease ofbirds in which there is impairment of normal endo-chondral bone formation. TD is characterized by theformation of a lesion composed of non-vascularizedand non-mineralized cartilage that can extend fromthe epiphyseal growth plate into the metaphysis(Leach and Nesheim, 1965). The mechanisms respon-sible for the development of TD are still unclear. Nev-ertheless, several molecules have been studied in TD,including the macromolecules of the ECM, theMMPs and several growth factors. Several MMPsare expressed during endochondral ossification, includ-ing MMP-2, -3, -9, -10, -13 and -14 (Ortega et al.,2004). Among these, only MMP-2, -3, -13 (D’Angeloet al., 2000), -9 (Tong et al., 2003) and -10 (Simsaet al., 2007a) have been cloned and studied in avianspecies. MMP-11 and -14 have not yet been identified.MMPs are a family of zinc-dependent proteases capa-ble of cleaving a variety of substrates, including ECM,non-ECM and cell surface proteins (Ortega et al.,2004). Because members of the MMP family havebeen implicated in physiological and pathological pro-cesses underlying ECMproteolytic cleavage or remod-elling (Vu and Werb, 2000), these enzymes have beenstudied recently in TD (Simsa et al., 2007b; Hasky-Negev et al., 2008; Dan et al., 2009), but themechanisms underlying their involvement in TDremain to be elucidated. The aim of the present studywas to contribute to elucidation of the mechanismsunderlying the development of TD by assessing theexpression of ECM macromolecules and ECM-remodelling enzymes.

Materials and Methods

Induction of Tibial Dyschondroplasia

One-day-old male broiler chickens (Cobb strain) wereobtained from a commercial hatchery (Santa Zitahatchery, Tomar, Portugal) and raised in two separategroups (n¼ 6 each) for 3 weeks under the recommen-ded temperature regime. Chickens from the control

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

group were fed ad libitum a normal diet, while chickensfrom the experimental group were fed ad libitum a dietcontaining 25 mg/kg of the fungicide tetramethylth-iuram disulfide (thiram) (Sigma, St Louis, Missouri),which is known to induce a high incidence of TD. At3 weeks of age the chickens were killed by cervical dis-location under ether anaesthesia. All experimentswere conducted in accordance with the principles andspecific guidelines presented in the Guidelines for theCare and Use of Agricultural Animals in AgriculturalResearch andTeaching. All procedures were approvedby a scientific committee, supervised by a Federation ofEuropean Laboratory Animal Science Associations(FELASA)-trained scientist and conforming to theregulations of the Portuguese law (Portaria 1005/92),following European Union Laboratory Animal Exper-imentation Regulations.

Tissue Extraction

Growth plates from birds in both groups were carefullyremoved from the upper articular cartilage and the un-derlying bone. Tissues were cut into 1 mm cubes andplaced immediately in liquid nitrogen and then storedat�80�C until used for zymography and immunoblot-ting, or were stored in RNAlater (Ambion, Austin,Texas) and kept at �20�C until used for real-timequantitative reverse transcriptase polymerase chainreaction (RT-PCR).

Total RNA Extraction and cDNA Synthesis

The samples were removed from RNAlater solutionand 30 mg of tissue was disrupted and homogenizedin 600 ml of RNeasy Lysis Buffer (RLT buffer e Qia-gen, Hilden, Germany) using a rotorestator YstralX10/25 (Ystral, BallrechteneDottingen, Germany).The tissue lysate was centrifuged for 1 min at 16,200g and the resultant supernatant was processed usingthe RNeasy Plant Mini kit (Qiagen). Spin columns ofthese kits eliminate the interfering glycosaminoglycansin cartilage (Huang et al., 2002). TotalRNAwas elutedin a 50 ml volume of RNase-free water. In order toquantify the amount of total RNA extracted and verifyRNA integrity, samples were analyzed using a 6000NanoChip kit, in anAgilent 2100 bioanalyzer (AgilentTechnologies, Walbronn, Germany) and 2100 expertsoftware, according to the manufacturer’s instructions.To eliminate the presence of contaminating genomicDNA in the RNA samples, DNase treatment was per-formed with Deoxyribonuclease I, AmplificationGrade kit (Invitrogen, Carlsbad, California) in accor-dance with the manufacturer’s instructions. RNA wasreverse transcribed with SuperScript III First StrandSynthesis System for RT-PCR (Invitrogen). One mgof DNase-treated RNA was added to 0.8 mM

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Avian Tibial Dyschondroplasia 3

deoxyribonucleotide triphosphate (dNTP) mix anda mixture of 4 ng/ml random hexamers/4 mM Oligo(dT)20, and cooled on ice for 1 min. A reactionmixtureof 1x buffer RT, 5 mM MgCl2, 10 mM dithiothreitol(DTT), 2 U/ml RNAseOUT and 10 U/ml of Super-Script III Reverse Transcriptase was added to the pre-vious RNA/primers mixture solution. The reactionswere conducted in a Gene Amp PCR System 9600thermocycler (Perkin Elmer, Norwalk, California) at25�C for 10 min, 50�C for 50 min and 85�C for5 min, and chilled on ice for 10 min.The reaction prod-ucts were then digested with 1 ml RNase H at 37�C for20 min and finally eluted with RNase-free water andstored at �20�C

Real-Time Quantitative PCR

Relative quantification of gene expression by real-timePCR was performed in the 7900 HT Sequence Detec-tion System (Applied Biosystems, Foster City, Califor-nia). Real-time PCR reactions were carried out using1x QuantiTect SYBR Green PCR Master Mix(Qiagen), forward and reverse primers at a previouslydetermined concentration (Table 1) and 20 ng of

Table

Oligonucleotide sequences for real-ti

Gene Accession number Oligonucleotide seq

MMP-2 NM_204420 for: 50-GCTCTGCAAGCArev: 50-CTCGAGGGTTTAC

MMP-7 NM_001006278 for: 50-CGCTGCGCTTCA

rev: 50-GCCACCTCTTCCMMP-9 NM_204667 for: 50-GCCATCACTGAGA

rev: 50-GATAGAGAAGGC

MMP-10 XM_417175 for: 50-TGCTTCTGGATT

rev: 50-AGTGGGCATCCMMP-11 XM_001232776 for: 50-CAAGGTACTGGC

rev: 50-AAAATGGACATC

MMP-13 AF070478 for: 50-GGAACACTCCAG

rev: 50-CTTGGATCCCTTCollagen II NM_204426 for: 50-GATGCCACCCTC

rev: 50-CAGAGTTTGAT

Collagen IX NM_205305 for: 50-TCAATCACCTCArev: 50-AAAAAGCTGCGCT

Collagen X M13496 for: 50-CTGGCCAATCC

rev: 50-TCCTGTGAGAGCT

Collagen XI XM_422303 for: 50-GGTCCCCCAGGTrev: 50-GAAGAATAGCTTG

Aggrecan NM_204955 for: 50-CCAAGGGAAGAG

rev: 50-CATTCCGGTGG

VEGF NM_205042 for: 50-CGATGAGGGCCTArev: 50-AGCTCATGTGCG

GAPDH NM_204305 for: 50-TCAGCAATGCA

rev: 50-GGCATGGACAGTG

HPRT1 NM_204848 for: 50-TGTAACCACCCAGrev: 50-CGGAGCTCACAA

TBP NM_205103 for: 50-TGAAAAGGCATT

rev: 50-CAGGGAAATAGGC

*Primers correspond to cDNA sequences deposited in GenBank; for, forwa

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

cDNA sample, in a total volume of 25 ml. The oligonu-cleotides used for PCR studies were designed using thePrimerExpress v2.0 software (Applied Biosystems) andare described in Table 1. Basic local alignment searchtool (BLAST) searches for all of the primer sequenceswere conducted to ensure gene specificity. The reac-tions were performed using the following thermal pro-file: 15 min at 95�C (PCR initial activation step) and40 cycles of 15 sec at 94�C, 30 sec at 55�C and 30 secat 72�C. All samples were run in duplicate. The result-ing PCRproduct sizes are given in Table 1. In additionto themelting point analysis done to ensure specific am-plification, PCR amplified products were purified foridentification by sequence analysis. Real-time PCRresults were analyzed with SDS 2.1 software (AppliedBiosystems). The housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (GAPDH), hypoxanthinephosphoribosyl-transferase (HPRT1) and TATA boxbinding protein (TBP) were selected by the softwaregeNorm (for underlying principles and calculationssee Vandesompele et al., 2002) as being the most stablegenes from a set of tested genes, and were used to nor-malize the results. The Ct values of each sample for

1

me quantitative RT-PCR analysis

uences* Concentration (nM) Product size (bp)

CGACATTG-30 300 101AGTCCTCCAC-30 300

AAAGAGTT-30 900 101

ATCAAAAGG-30 100TCAATGGAG-30 300 102

GCCCTGAGT-30 300

TCACGGTG-30 300 101

CCTCCTATC-30 300ATGGTGACA-30 300 101

TCCTTCCCG-30 300

AGACCCTGG-30 100 101

GCACATCAT-30 100AAATCCCT-30 100 101

GTCGCGGC-30 100

CCTTCCCTG-30 100 103

AGTACACCC-30 300ACAATCCC-30 100 101

TGATTTGCTC-30 100

ACCATGTT-30 100 101TGCTTGAGCC-30 100

AACGTGACC-30 100 102

TGAAAGCA-30 300

GAATGTGTC-30 300 101CTATGTGC-30 300

TCGTGCAC-30 300 101

GTCATAAGAC-30 300

TGCAATTCT-30 900 101ACAGCACA-30 300

GCATATGGC-30 300 101

ACTAACTGGG-30 300

rd primer; rev, reverse primer.

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

4 I. Velada et al.

each gene of interest were transformed into relativequantities using the formula 2(min Ct�sample Ct) (delta-Ct method), where min Ct¼ lowest Ct value¼Ctvalue of sample with highest expression. The normal-ized gene of interest expression levels can be calculatedby dividing these relative quantities by the appropriatenormalization factor (geometric mean of the threehousekeeping genes), provided by the geNorm software(Vandesompele et al., 2002).

Extraction of Proteins

Frozen tibial growth plates were homogenized in ice-cold buffer, 20 mM Tris, pH 8.0, 0.5 mM CaCl2,0.5% NP-40 (Ohkubo et al., 2003). The homogenatewas incubated at 4�C for 2 h with gentle shaking. Aftercentrifugation at 10,000 g for 10 min, the supernatantwas collected and protein concentration was deter-mined using the bicinchoninic acid (BCA) protein as-say (Pierce, Rockford, Illinois) according to themanufacturer’s instructions.

Gelatin Zymography

Aliquots (10 mg) of protein extracts of growth plateswere mixed with equal volumes of sample loadingbuffer (100 mM Tris, pH 6.8, 5% sodium dodecylsulphate [SDS], 20% glycerol, 0.1% bromophenolblue) and subjected to SDS-polyacrylamide gel elec-trophoresis (SDS-PAGE) (Hoefer, San Francisco,California) through 4% stacking and 10% separat-ing (containing 0.1% gelatin) gels, under non-re-ducing conditions. The gels were washed for30 min at room temperature in 2.5% TritonX-100, rinsed in distilled water and developed for16 h at 37�C in 50 mM Tris pH 7.4, 5 mM NaCl,10 mM CaCl2, 1 mM ZnCl2, 0.02% (v/v) TritonX-100. Finally, gels were stained with Coomassiebrilliant blue R250 (Pierce, Steinheim, Germany)to visualize protease activity and scanned witha Gel Image Analyzer (FX-710 scanner, QuantityONE software, Bio-Rad, Hercules, California) in or-der to estimate the molecular weights and opticaldensities of the protein bands.

SDS-PAGE and Immunoblotting

Aliquots (24 mg) of protein extracts from growthplates were separated by SDS-PAGE through a 4%stacking and 10% separating gel, under reducingconditions. Proteins were electrotransferred at90 mA to polyvinylidene difluoride (PVDF) mem-branes (Bio-Rad) overnight at 4�C. The membraneswere blocked for 1 h at room temperature with 5%dried milk (w/v) in tris-buffered saline and tween 20(TBS-T). Blocked membranes were incubated for

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

1 h at room temperature with rabbit anti-chickenMMP-2 antibody, produced against the active formof the protein (kindly provided by Dr. JP Quigleyand Dr. EI Deryugina, Scripps Research Institute,La Jolla, California), diluted 1 in 1000 in TBS-T.The membranes were washed three times with TBS-T and then incubated for 1 h at room temperaturewith the anti-rabbit horseradish peroxidase (HRP)-labelled secondary antibody (Amersham PharmaciaBiotech, Freiburg, Germany), diluted 1 in 1000 inTBS-T. After washing as described above, mem-branes were developed with enhanced chemilumines-cence (ECL) reagents (Amersham PharmaciaBiotech, Freiburg, Germany) following the manufac-turer’s instructions. The rabbit monoclonal antibodyspecific for b-actin (Santa Cruz Biotechnology, SantaCruz, California) was used as loading control. Detec-tion of protein bands was performed by short expo-sure to blue light-sensitive autoradiography film.The films were scanned on a Gel Image Analyzer(FX-710 scanner, Quantity ONE software, Bio-Rad) in order to estimate the molecular weights andoptical densities (ODS) of the protein bands and theresults were expressed relative to the b-actin signals.

Statistical Analysis

Differences between the means of the control and ex-perimental groups for the normalized expressionlevels of the genes of interest, or for the optical density(OD) values, were evaluated with the Student’s t-test.Differences were considered significant at P# 0.05and for gene expression levels when, in addition to be-ing statistically significant (P# 0.05), there wasa 2-fold or greater difference in expression betweenthe control and experimental groups. Each real-timequantitative RT-PCR experiment was performedtwice and each zymography and immunoblottingexperiment at least three times. The graphical dataand the images show a representative experiment.

Results

Morphological Analysis of Growth Plates from Thiram-Fed

Chickens

When the proximal tibia and tarsus were dissected bysectioning the bones lengthwise, a mass of opaque car-tilage was observed below the epiphyseal plateextending into the metaphysis (Fig. 1b). There wasa uniform band across the width of the bone that af-fected the entire growth plate, resulting in an epiphy-sis comprised solely of cartilage and periosteum. Thiscartilage was unmineralized and devoid of blood ves-sels, giving it a soft and pliable texture. The articularcartilage was not affected by TD.

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Fig. 1. Morphological analysis of the avian growth plate. (a)Normal (N) growth plate and (b) growthplate affected byTD.TheTD lesionis an opaque and avascular cartilaginous mass, which occupies a large area of the metaphysis under the growth plate.

Avian Tibial Dyschondroplasia 5

Expression of Genes Encoding Matrix Macromolecules

Expression of genes encoding collagens II, IX and XIwas not significantly different between the TD andthe control birds (Fig. 2b, c and d). There was a signif-icant increase in expression of genes encoding colla-gen X and aggrecan in TD cartilage (P< 0.05;Fig. 2a and e), but these increments were very low(1.7-fold for collagen X and 1.6-fold for aggrecan),so they were not considered relevant.

Gene Expression and Protein Levels of Matrix

Metalloproteinases

There was no difference in expression of the gene en-coding MMP-2 in samples from control and TDgroups (Fig. 3a). Gelatin zymography revealed highactivity of the 64 and 60 kDa proteins of MMP-2 innormal growth plates (Fig. 3b) and a significantreduction (P< 0.05) in the activity of both proteinswas detected in TD samples (Fig. 3b and c). In orderto confirm that both bands wereMMP-2 proteins, im-munoblotting analysis was performed by using an anti-chickenMMP-2 antibody specific for the active form ofthe enzyme. Only the band with an apparent molecu-lar weight of 64 kDa was recognized by the antibody(Fig. 3d), suggesting that the 60 kDa band may bean active degradation product of MMP-2. The immu-noblotting analysis confirmed a significant (P< 0.05)reduction in the amount of 64 kDa MMP-2 proteinin TD lesions, (Fig. 3d and e).

There was significant (P< 0.01) down-regulation(of approximately 3-fold) of expression of the gene en-

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

coding MMP-9 (Fig. 4) and marked down-regulation(of approximately 6.7-fold) of expression of the gene en-coding MMP-13 (P< 0.05) in TD lesions (Fig. 5a).Gelatin zymography revealed a significant decrease(P< 0.05) in the activity of the 50 kDa MMP-13enzyme in TD lesions (Fig. 5b and c). The 46 kDagelatinolytic band of MMP-13 is probably an activedegradation product of the protein. There was a signif-icant reduction in the gene expression levels of about2.6-fold (P< 0.01) for MMP-10 (Fig. 6a) and of about5-fold (P< 0.05) for MMP-11 (Fig. 6b) in TD lesionsversus control tissues. There was marked (P< 0.05)up-regulation (7.3-fold) of MMP-7 gene expression inTD lesions (Fig. 6c).

Expression of the VEGF Gene

In order to explain the presence of avascular cartilagein TD lesions, the expression of the gene encoding vas-cular endothelial growth factor (VEGF) was investi-gated. There was significant (P< 0.01) up-regulation(of about 2.8-fold) of VEGF gene expression in TDlesions (Fig. 7).

Discussion

TD is characterized by disruption of normal endo-chondral ossification, but the precise mechanismsinvolved in development of the lesions remain to beelucidated. TD involves an accumulation of cartilageextending from the epiphyseal growth plate into themetaphysis (Leach and Nesheim, 1965). One possible

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Fig. 2. RelativemRNA expression for (a) collagenX, (b) collagen II, (c) collagen IX, (d) collagenXI and (e) aggrecan. Expression levelsare normalized to the geometric mean of the levels of GAPDH, HPRT1 and TBP gene expression. Values are shown asmean � standard deviation (SD) (n¼ 6 for each group). *P< 0.05. Results are shown relative to mRNA expression levels fromthe control group (normal growth plates) set to one, thus corresponding to the N-fold difference in relation to the control group.Normal, normal growth plates; TD, TD-affected growth plates.

6 I. Velada et al.

explanation for the accumulation of cartilage in TDlesions could be an increased synthesis of ECM mac-romolecules; however, the results of the present studyshow no significant changes in the gene expression ofmacromolecules such as collagens II, IX, X and XI,and aggrecan (the main proteoglycan of the cartilagematrix). Collagen II interacts with collagens IX andXI to form heterotypic fibrils that are distributed

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

throughout the cartilage matrix (van der Rest andMayne, 1988; Vaughan et al., 1988) and thereforecollagens IX and XI play an important role in thegrowth plate. The finding of unaltered collagengene expression in TD contrasts with the reductionin protein levels described by Wardale and Duance(1996), suggesting a defect in the secretion and incor-poration of collagens IX and XI into the matrix, as

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Fig. 3. Relative mRNA expression for MMP-2 and down-regulation of MMP-2 protein expression and activity. (a) Real-time quantita-tive RT-PCR analysis of MMP-2. Expression levels are normalized to the levels of the geometric mean of GAPDH, HPRT1 andTBPgene expression.Values are shown asmean� SD (n¼ 6 for each group).Results are shown relative tomRNAexpression levelsfrom the control group (normal growth plates) set to one, thus corresponding to the N-fold difference in relation to the controlgroup. (b, c) Gelatin zymographic analysis of normal and TD-affected growth plate extracts. (b) The zymogram shows proteinbands corresponding to the 60 and 64 kDa MMP-2. (c) The ODs of the bands corresponding to the 60 and 64 kDa MMP-2were calculated for normal and TD-affected growth plate extracts by scanning densitometry and are given as arbitrary units.Values are shown as mean� SD (n¼ 6 for each group). *P< 0.05. (d, e) Immunoblotting analysis of normal and TD-affectedgrowth plate extracts. Proteins were probed with an anti-chicken MMP-2 antibody that recognized the band corresponding tothe 64 kDa MMP-2. Membranes were also probed with an antibody against the 43 kDa b-actin to verify that equal amounts ofproteins were loaded. (e) The OD of the band corresponding to the 64 kDa MMP-2 was calculated for normal and TD-affectedgrowth plate extracts by scanning densitometry, and was normalized relative to the amount of b-actin protein, and is given asarbitrary units.Values are shownasmean� SD (n¼ 6 for eachgroup). *P< 0.05.Normal, normal growthplates;TD,TD-affectedgrowth plates; MMP, matrix metalloproteinase.

Avian Tibial Dyschondroplasia 7

suggested for collagen X by others (Thorp et al., 1993;Reginato et al., 1998).

The importance of MMPs in endochondral bonegrowth has been investigated in studies of mice lacking

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

these enzymes (Vu et al., 1998; Inada et al., 2004;Stickens et al., 2004). The involvement of MMPs inTD has been also studied (Gay et al., 2007; Hasky-Negev et al., 2008; Dan et al., 2009) however, the role

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Fig. 4. Relative mRNA expression for MMP-9. Expression levelswere normalized to the levels of the geometric mean ofGAPDH, HPRT1 and TBP gene expression. Values areshown as mean� SD (n¼ 6 for each group). **P< 0.01.Results are shown relative to mRNA expression in the con-trol group (normal growth plates), set to one, thus corre-sponding to the N-fold difference in relation to the controlgroup. Normal, normal growth plates; TD, TD-affectedgrowth plates.

Fig. 5. Down-regulation of MMP-13 mRNA expression andactivity. (a) Real-time quantitative RT-PCR analysis ofMMP-13 gene expression. Expression levels were normal-ized to the levels of the geometric mean of GAPDH,HPRT1 and TBP gene expression. Values are shown asmean� SD (n¼ 6 for each group). **P< 0.01. Resultsare shown relative to mRNA expression levels from thecontrol group (normal growth plates), set to one, thus cor-responding to the N-fold difference in relation to the con-trol group. (b, c) Gelatin zymographic analysis of normaland TD-affected growth plate extracts. (b) Thezymogram shows protein bands corresponding to the 46and 50 kDa MMP-13. (c) The OD of the band corre-sponding to the 50 kDa MMP-13 was calculated for nor-mal and TD-affected growth plate extracts by scanningdensitometry and is given as arbitrary units. Values areshown as mean� SD (n¼ 6 for each group). *P< 0.05.Normal, normal growth plates; TD, TD-affected growthplates; MMP, matrix metalloproteinase.

8 I. Velada et al.

of these enzymes in the development of the pathologyremains to be elucidated. In the present study therewas a general down-regulation in MMP geneexpression in TD, except for genes encoding MMP-2(gelatinase A) and MMP-7 (matrilysin-1). MMP-9(gelatinase B), MMP-13 (collagenase-3), MMP-10(stromelysin-2) and MMP-11 (stromelysin-3) mRNAlevels were markedly reduced in TD, and there wasalso a decrease in MMP-2 and MMP-13 proteinlevels. MMP-2 mRNA levels did not change in TD,suggesting that this protein may play the samehousekeeping function in the avian growth plate,during endochondral bone formation, as itshomologue in mammals (Matrisian, 1994). To ourknowledge, this is the first study of TD in which the ex-pression of MMP-2 protein and mRNA has been eval-uated in parallel. The findings suggest that the reducedamount of MMP-2 active protein in TD is not a conse-quence of decreased expression of its gene. Further-more, it is possible that the reduction of MMP-2protein levels without a concomitant decrease ofmRNA may be the result of decreased translation, in-creased degradation of the protein or decreased proteinactivation. The antibody employed in the presentstudy was specific for the active form of the moleculerather than the pro-form ofMMP-2, so it is not possibleto speculate about the possibility of decreased transla-tion of MMP-2. We hypothesize that decreasedMMP-2 activation may occur, and that this may bedue to reduced expression of MMP-13. Indeed, previ-ous reports have shown that MMP-13 can activateMMP-2 in mammals (Chakraborti et al., 2003).

Please cite this article in press as: Velada I et al., Expression of Genes Encoding Extracellular Matrix Macromolecules and Metalloproteinases in

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011), doi:10.1016/j.jcpa.2010.12.008

Fig. 6. Relative mRNA expression for (a)MMP-10, (b)MMP-11and (c) MMP-7. Expression levels were normalized to thelevels of the geometric mean of GAPDH, HPRT1 andTBP gene expression. Values are shown as mean� SD(n¼ 6 for each group). *P< 0.05 and **P< 0.01. Resultsare shown relative tomRNAexpression levels from the con-trol group (normal growth plates), set to one, thus corre-sponding to the N-fold difference in relation to thecontrol group. Normal, normal growth plates; TD,TD-affected growth plates.

Fig. 7. Relative mRNA expression for VEGF. Expression levelswere normalized to the levels of the geometric mean ofGAPDH, HPRT1 and TBP gene expression. Values areshown as mean� SD (n¼ 6 for each group). **P< 0.01.Results are shown relative to mRNA expression levelsfrom the control group (normal growth plates), set toone, thus corresponding to the N-fold difference in relationto the control group. Normal, normal growth plates; TD,TD-affected growth plates.

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

Avian Tibial Dyschondroplasia 9

Previous reports have demonstrated the involvement ofMMP-13 in chondrocyte hypertrophy. For example,D’Angelo et al. (2000) observed, using chondrocytesfrom prehypertrophic cartilage of chick embryos, thatwhen hypertrophy was induced by serum-free culturewith inducers of hypertrophy, MMP-13 mRNA levelswere increased. The results of our study strengthenthese previous findings. Indeed, we have shown de-creased expression of MMP-13, both at the proteinand mRNA levels, associated with a pathology wherethere is a failure in terminal differentiation from prehy-pertrophic to hypertrophic chondrocytes (Poulos et al.,1978; Hargest et al., 1985).

Similarly, there was decreased expression of the geneencoding MMP-9 in the lesions of TD. This observa-tion, in a disease where vascularization is interrupted,strengthens the previous findings of Tong et al.

(2003), who demonstrated a positive correlation be-tween substances that promote vascularization of thechicken growth plate and the expression of MMP-9.The only stromelysin cloned and studied so far in theavian growth plate is stromelysin-2 (MMP-10; Simsaet al., 2007a). The present study has shown down-regu-lation of MMP-10 associated with TD, a disease wherevascularization is inhibited, supporting the findings ofSimsa et al. (2007a), who demonstrated expression ofMMP-10 in the turkey growth plate in cells surround-ing the blood vessels penetrating the growth plate. The

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

10 I. Velada et al.

present study is the first to show the expression of theMMP-11 gene in the normal chicken growth plate.Like MMP-10, MMP-11 was also down-regulated inTD lesions.

This is also the first report of the expression ofMMP-7 in the normal chicken growth plate. As men-tioned above, the MMP-2 and MMP-7 genes werenot down-regulated in TD, as was the case with the re-maining MMPs genes examined. The substrate speci-ficity of MMP-7 suggests that this molecule may playa role in ECM turnover under normal conditions(Wilson and Matrisian, 1996). Thus, given thatMMP-7 is an ECM-degrading enzyme, our results

Fig. 8. A proposedmodel for the involvement ofMMPs in the developmproductionofMMP-13,whichmay lead to reduced activation ofmay also contribute to reduced activation ofMMP-9 andMMP-lack ofECMproteolysis and consequently to the accumulationofto a defect in the release of regulatory molecules sequestered intoduction and/ indicates activation. The activation cascade of

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

showing its up-regulation associatedwithTD, a diseaseinvolving an accumulation of cartilage ECM, are in-triguing. One possible explanation is that MMP-7 isbeing induced by the hypoxic environment in TD car-tilages. In this respect, Burke et al. (2003) demonstratedthat up-regulation of both VEGF and MMP-7 geneexpression occurs in human monocyte-derived macro-phages during hypoxia.

In the present study we also demonstrated a pro-nounced increase in VEGF mRNA in TD lesions,which is in accordance with the results of Rath et al.

(2005, 2007). It is well established that hypoxia is oneof the principal stimuli for the expression of VEGF.

ent of TD. A delay in chondrocyte hypertrophy results in decreasedMMP-2 andMMP-9.Additionally, reduced expression ofMMP-1013.The general down-regulation ofMMPactivity contributes to thecartilage, aswell as to the failure in growthplate vascularization duethe matrix, such as VEGF and FGF. The symbol0 indicates pro-

MMPs is adapted from Chakraborti et al. (2003).

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Avian Tibial Dyschondroplasia 11

Chondrocytes exist in a hypoxic environment in thegrowth plate because the blood supply to these cells islimited and therefore these cells are a potential sourceof VEGF (Gerber et al., 1999). However, Shapiroet al. (1997) have demonstrated that chondrocytes inthe chicken growth plate are not hypoxic. This is re-lated to structural differences inmammalian and aviangrowth plates whereby the metaphyseal blood vesselspenetrate more deeply into the avian growth plate(Leach and Gay, 1987). We speculate that, despitethe fact that the normal avian growth plate is not hyp-oxic, dyschondroplastic growth plates may becomehypoxic owing to the increased accumulation of carti-lage matrix and, therefore, there is induction ofVEGF and MMP-7.

The observations made in the present study allowthe development of a model for the development ofTD (Fig. 8). Since MMP-13 is exclusively synthesizedby hypertrophic chondrocytes, the reduced MMP-13mRNA and protein levels in TD lesions may be a con-sequence of the delay in chondrocyte hypertrophy,a feature of dyschondroplasia. In other words, thereduction of MMP-13 expression may be simply theresult of decreased MMP-13-producing cells. BecauseMMPs are involved in activation cascades, any alter-ation in the protein levels of one MMP will lead tochanges in the levels of MMPs that are in turn acti-vated by that MMP. Because MMP-13 can activateMMP-2 and MMP-9 (Chakraborti et al., 2003),reduced expression of MMP-13 may lead to decreasedactivation of MMP-2 and MMP-9. In the same man-ner, because MMP-10 can activate MMP-9 andMMP-13 (Chakraborti et al., 2003), reduced expres-sion of MMP-10 may lead to decreased activation ofMMP-9 and MMP-13.

Given that the main function of MMPs is the deg-radation of ECM, the decreased expression of theseECM-degrading enzymes may lead to a reductionin ECM proteolysis. This may lead to the failureof growth plate vascularization, as well as to im-paired chondrocyte apoptosis, owing to a defect inthe release of regulatory molecules sequesteredinto the matrix. These include VEGF and the fibro-blast growth factor (FGF), which are thus pre-vented from exerting their normal function in thevascularization of the growth plate. Indeed, FGFhas also been shown to act as an angiogenic factorin mammals, in addition to its functions in chondro-cyte proliferation and differentiation (Folkman andKlagsbrun, 1987). The results of the present studyhave shown that failure in the vascularization ofthe affected growth plates is not due to down-regu-lation of VEGF and FGF gene expression. As dis-cussed above, VEGF was increased in TD lesionsand FGF was unchanged (data not shown). The

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

existence of abnormal ECM degradation is sup-ported by the fact that increased VEGF mRNAlevels in affected cartilages do not stimulate theirvascularization. Thus, even when VEGF geneexpression is up-regulated, the corresponding pro-tein cannot exert its function because it is arrestedin ECM storage sites.

In conclusion, the present study has identified, forthe first time, MMP-7 and MMP-11 gene expressionin the avian tibial growth plate. The MMP genesexpressed in the mammalian growth plate are alsoexpressed in the avian growth plate, suggesting thatthe MMPs expressed in the avian growth plate mayplay the same role as their counterparts in mammals.This occurred for MMP-2, which plays the samehousekeeping function, for MMP-13, which plays thesame role in chondrocyte hypertrophy, and forMMP-9, which plays the same role in the vasculariza-tion of the growth plate.The results of the present studysuggest that dyschondroplastic cartilage is an hypoxicenvironment and that the cartilage that accumulatesin TD is the result of decreased matrix proteolysis,due to the down-regulation of the ECM-remodellingenzymes (MMPs), and not due to an increased produc-tion of the ECM macromolecules. The failure in ex-pression or lack of activation of MMPs might beassociated with the development of avian TD. Thesedata strengthen the link between the lack of MMP ex-pression and abnormal endochondral bone formation,which might be pivotal findings for further research onpathologies related to abnormal bone growth.

Acknowledgments

The authors would like to thank the Fundac~ao paraa Ciencia e Tecnologia (FCT), Portugal, for support-ing I. Velada (SFRH/BD/6420/2001). The authorsare grateful to Dr. J. P. Quigley and Dr. E. I. Deryu-gina (both from the Scripps Research Institute,La Jolla, California) for kindly providing the anti-chicken MMP-2 antibody. The authors also thankDr.M. L. Pais andDr. A.Martinho (both fromCenterfor Histocompatibility, Coimbra, Portugal) for kindlysupporting this work with the necessary equipment,and A. C. Sarmento (University of Aveiro, Portugal)for critically reading this manuscript. All of the authorsdeclare no conflicts of interest.

References

BallockRT,O’Keefe RJ (2003) Physiology and pathophys-iology of the growth plate.Birth Defects Research (Part C),69, 123e143.

Burke B, Giannoudis A, Corke KP, Gill D, Wells M et al.(2003) Hypoxia-induced gene expression in humanmacrophages: implications for ischemic tissues and

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

12 I. Velada et al.

hypoxia-regulated gene therapy. American Journal of

Pathology, 163, 1233e1243.Chakraborti S,MandalM,Das S,MandalA,Chakraborti T

(2003) Regulation of matrix metalloproteinases: an over-view. Molecular and Cell Biochemistry, 253, 269e285.

D’Angelo M, Yan Z, Nooreyazdan M, Pacifici M,Sarment DS et al. (2000) MMP-13 is induced duringchondrocyte hypertrophy. Journal of Cell Biochemistry,77, 678e693.

Dan H, Simsa-Maziel S, Hisdai A, Sela-Donenfeld D,MonsonegoOrnanE (2009) Expression ofmatrixmetal-loproteinases during impairment and recovery of theavian growth plate. Journal of Animal Science, 87,3544e3555.

Folkman J, Klagsbrun M (1987) Vascular physiology. Afamily of angiogenic peptides. Nature, 329, 671e672.

Gay CV, Gilman VR, Leach RM Jr. (2007) Immunolocal-ization of vascularization factors in normal, tibial dys-chondroplasia and rachitic cartilage. Avian Pathology, 36,445e451.

Gerber HP, Vu TH, Ryan AM, Kowalski J, Werb Zet al. (1999) VEGF couples hypertrophic cartilage re-modeling, ossification and angiogenesis during endo-chondral bone formation. Nature Medicine, 5,623e628.

Hargest TE, Leach RM, Gay CV (1985) Avian tibial dys-chondroplasia. I. Ultrastructure. American Journal of

Pathology, 119, 175e190.Hasky-Negev M, Simsa S, Tong A, Genina O, Monsonego

Ornan E (2008) Expression of matrix metalloprotei-nases during vascularization and ossification of normaland impaired avian growth plate. Journal of Animal

Science, 86, 1306e1315.HortonWA(1993)Morphologyof connective tissue: cartilage.

In: Connective Tissue and its Heritable Disorders, V McCusick,Ed., Wiley-Liss Inc., New York, pp. 641e675.

Huang W, Li WQ, Dehnade F, Zafarullah M (2002) Tis-sue inhibitor of metalloproteinases-4 (TIMP-4) geneexpression is increased in human osteoarthritic femo-ral head cartilage. Journal of Cell Biochemistry, 85,295e303.

Howell DS, Dean DD (1992) The biology, chemistry, andbiochemistry of the mammalian growth plate. In:Disor-

ders of Bone and Mineral Metabolism, FL Coe, MJ Favus,Eds., Raven Press, New York, pp. 313e353.

Inada M, Wang Y, Byrne MH, Rahman MU, Miyaura Cet al. (2004) Critical roles for collagenase-3 (Mmp13) indevelopment of growth plate cartilage and in endochon-dral ossification. Proceedings of the National Academy of

Sciences USA, 101, 17192e17197.Leach JRM, LilburnMS (1992) Current knowledge on the

etiology of tibial dyschondroplasia in the avian species.Poultry Science, 4, 57e65.

Leach RM Jr., Gay CV (1987) Role of epiphyseal cartilagein endochondral bone formation. Journal of Nutrition,117, 784e790.

LeachRMJr., NesheimMC(1965)Nutritional, genetic andmorphological studies of an abnormal cartilage formationin young chicks. Journal of Nutrition, 86, 236e244.

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

Matrisian LM (1994) Matrix metalloproteinase geneexpression. Annals of the New York Academy of Sciences,732, 42e50.

Ohkubo K, Shimokawa H, Ogawa T, Suzuki S, Fukada Ket al. (2003) Immunohistochemical localization of ma-trix metalloproteinase 13 (MMP-13) in mouse mandib-ular condylar cartilage. Journal of Medical and Dental

Science, 50, 203e211.Ortega N, Behonick DJ,Werb Z (2004)Matrix remodeling

during endochondral ossification. Trends in Cell Biology,14, 86e93.

Orth MW (1999) The regulation of growth plate cartilageturnover. Journal of Animal Science, 77(Suppl. 2),183e189.

Poulos PW Jr., Reiland S, Elwinger K, Olsson SE (1978)Skeletal lesions in the broiler,with special reference todys-chondroplasia (osteochondrosis). Pathology, frequencyand clinical significance in two strains of birds on highand low energy feed. Acta Radiologica, 358, 229e275.

Rath NC, Huff WE, Huff GR (2007) Thiram-inducedchanges in the expression of genes relating to vasculari-zation and tibial dyschondroplasia. Poultry Science, 86,2390e2395.

Rath NC, Richards MP, Huff WE, Huff GR, Balog JM(2005) Changes in the tibial growth plates of chickenswith thiram-induced dyschondroplasia. Journal of

Comparative Pathology, 133, 41e52.Reginato AM, Bashey RI, Rosselot G, Leach RM, Gay CV

et al. (1998) Type X collagen biosynthesis and expres-sion in avian tibial dyschondroplasia. Osteoarthritis andCartilage, 6, 125e136.

Shapiro IM, Mansfield KD, Evans SM, Lord EM,Koch CJ (1997) Chondrocytes in the endochondralgrowth cartilage are not hypoxic. American Journal of

Physiology, 272, C1134eC1143.Simsa S, Genina O, Ornan EM (2007a) Matrix metallo-

proteinase expression and localization in turkey(Meleagris gallopavo) during the endochondral ossifica-tion process. Journal of Animal Science, 85, 1393e1401.

Simsa S, Hasdai A, DanH,Ornan EM (2007b) Differentialregulation of MMPs and matrix assembly in chickenand turkey growth-plate chondrocytes. American Journalof Physiology, 292, R2216eR2224.

Stickens D, BehonickDJ,OrtegaN,Heyer B,Hartenstein Bet al. (2004) Altered endochondral bone development inmatrix metalloproteinase 13-deficient mice.Development,131, 5883e5895.

Thorp BH, Farquharson C, Kwan APL, Loveridge N(1993) Osteochondrosis/dyschondroplasia: a failure ofchondrocyte differentiation. Equine Veterinary Journal,16, 13e18.

Tong A, Reich A, Genin O, Pines M,Monsonego-Ornan E(2003) Expression of chicken 75-kDa gelatinase B-likeenzyme in perivascular chondrocytes suggests its rolein vascularization of the growth plate. Journal of Boneand Mineral Research, 18, 1443e1452.

van der Eerden BC, KarperienM,Wit JM (2003) Systemicand local regulation of the growth plate. EndocrinologyReviews, 24, 782e801.

coding Extracellular Matrix Macromolecules and Metalloproteinases in

doi:10.1016/j.jcpa.2010.12.008

Avian Tibial Dyschondroplasia 13

van der Rest M, Mayne R (1988) Type IX collagen pro-teoglycan from cartilage is covalently cross-linked totype II collagen. Journal of Biological Chemistry, 263,1615e1618.

Vandesompele J, De Preter K, Pattyn F, Poppe B,Van Roy N et al. (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averagingofmultiple internal control genes.GenomeBiology,3, 1e12.

Vaughan L, Mendler M, Huber S, Bruckner P,Winterhalter KH et al. (1988) D-periodic distribution ofcollagen type IX along cartilage fibrils. Journal of Cell

Biology, 106, 991e997.Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA et al.

(1998) MMP-9/gelatinase B is a key regulator of growthplate angiogenesis and apoptosis of hypertrophic chon-drocytes. Cell, 93, 411e422.

Please cite this article in press as: Velada I et al., Expression of Genes En

Avian Tibial Dyschondroplasia, Journal of Comparative Pathology (2011),

Vu TH, Werb Z (2000) Matrix metalloproteinases: effec-tors of development and normal physiology. Genes andDevelopment, 14, 2123e2133.

Wardale RJ, Duance VC (1996) Collagen expression inchicken tibial dyschondroplasia. Journal of Cell Science,109, 1119e1131.

Wilson CL, Matrisian LM (1996) Matrilysin: an epithelialmatrix metalloproteinase with potentially novel func-tions. International Journal of Biochemistry and Cell Biology,28, 123e136.

coding Extracell

doi:10.1016/j.jc

½ R

A

ula

pa.

eceived, July 9th, 2010

ccepted, December 13th, 2010

r Matrix Macromolecules and Me

2010.12.008

�

talloproteinases in