SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Objective

To analyze the differences in gene expression profiles of chondrocytes in intact and damaged regions of cartilage from the same knee joint of patients with osteoarthritis (OA) of the knee.

Methods

We compared messenger RNA expression profiles in regions of intact and damaged cartilage (classified according to the Mankin scale) obtained from patients with knee OA. Five pairs of intact and damaged regions of OA cartilage were evaluated by oligonucleotide array analysis using a double in vitro transcription amplification technique. The microarray data were confirmed by real-time quantitative polymerase chain reaction (PCR) amplification and were compared with previously published data.

Results

About 1,500 transcripts, which corresponded to 8% of the expressed transcripts, showed ≥2-fold differences in expression between the cartilage tissue pairs. Approximately 10% of these transcripts (n = 151) were commonly expressed in the 5 patient samples. Accordingly, 114 genes (35 genes expressed in intact > damaged; 79 genes expressed in intact < damaged) were selected. The expression of some genes related to the wound-healing process, including cell proliferation and interstitial collagen synthesis, was higher in damaged regions than in intact regions, similar to the findings for genes that inhibit matrix degradation. Comparisons of the real-time quantitative PCR data with the previously reported data support the validity of our microarray data.

Conclusion

Differences between intact and damaged regions of OA cartilage exhibited a similar pattern among the 5 patients examined, indicating the presence of common mechanisms that contribute to cartilage destruction. Elucidation of this mechanism is important for the development of effective treatments for OA.

Osteoarthritis (OA) is the most prevalent form of arthritis in the elderly and is characterized primarily by the degeneration and loss of articular cartilage. OA is considered a heterogeneous group of disorders with a variety of pathogenic factors, all of which result in similar patterns of cartilage degeneration (1). In OA of the knee, the medial compartment of the articular cartilage is the most susceptible to degeneration, whereas the lateral compartment remains relatively unaffected (2). This phenomenon appears in a single joint despite the same OA susceptibility of the cartilage matrix and the same genetic background.

There is now abundant evidence that chondrocytes play a critical role in cartilage degeneration. For example, OA chondrocytes secrete a variety of matrix breakdown products and cytokines, and cleavage of type II collagen is observed primarily around chondrocytes (3). Therefore, changes in the gene expression patterns of chondrocytes in response to various exogenous stimuli could affect the integrity of articular cartilage. Based on this concept, many groups of investigators have reported that gene expression in certain molecules differ from region to region within a single OA joint (4–7).

Because comprehensive gene expression profiles in intact and damaged regions of human OA cartilage have never been compared, the following questions remain unresolved. What percentage of the expressed transcripts shows obvious differences in expression levels in intact and damaged regions? What percentage of the differentially expressed transcripts shows the same expression pattern in different patient samples? What are the molecular functions involved in such groups of genes?

To provide answers to these questions, we used Affymetrix high-density oligonucleotide array analysis to compare the gene expression profile of chondrocytes in intact regions of joint cartilage with the profile of chondrocytes in damaged regions of joint cartilage from the same knee. We also evaluated the validity of our oligonucleotide array data according to the results of real-time quantitative polymerase chain reaction (PCR) amplification as well as gene expression data reported by other groups of investigators (4–7).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Cartilage samples.

Specimens of human articular cartilage were obtained from a total of 15 OA patients; 5 of the initial 9 patients' samples were used for microarray analysis, based on the results of the histologic assessment, and 4 of the 6 patients' samples used to confirm the microarray results were used for real-time PCR analysis after histologic assessment. Patients were undergoing total knee replacement surgery at St. Marianna University Hospital. OA was diagnosed according to the findings of the clinical history and physical examination, as well as radiographic findings. Within 4 hours after surgery, 8 pieces of cartilage tissue (4 apparently intact cartilage pieces and 4 damaged cartilage pieces) were dissected from each specimen. Each piece of cartilage was further divided into 2 fragments. One of the fragments was rapidly frozen in liquid nitrogen and stored at −80°C until RNA isolation was performed. The other fragment was frozen in TissueTek OCT compound (Sakura, Tokyo, Japan) and was used for histologic examination.

This study was performed after obtaining approval from the Ethics Committee for Human Studies, St. Marianna University Hospital. All patients provided informed consent.

Histopathologic assessment.

Cartilage sections were stained with Safranin O–fast green. Sections were then evaluated histopathologically and scored according to the Mankin scale (8), which assesses the following 4 features: cartilage structure (0–6 scale, where 0 = normal, 1 = surface irregularity, 2 = pannus and surface irregularity, 3 = clefts to transitional zone, 4 = clefts to radial zone, 5 = clefts to calcified zone, and 6 = complete disorganization), cellularity (0–3 scale, where 0 = normal, 1 = hypercellularity, 2 = cloning, and 3 = hypocellularity), Safranin O staining (0–4 scale, where 0 = normal, 1 = slight reduction, 2 = modest reduction, 3 = severe reduction, and 4 = no dye noted), and tidemark integrity (0–1 scale, where 0 = intact and 1 = blood vessels cross the tidemark). Cartilage sections in the present study were scored for the first 3 features (structure, cellularity, and Safranin O staining). Analysis of tidemark integrity was not evaluated because of the absence of a calcified zone in the cartilage samples we examined.

Based on the Mankin scores, we defined intact and damaged regions of the OA cartilage samples for comparative analysis. Intact cartilage regions were defined as follows: normal cartilage structure (score of 0), no cell clusters (score of 0 or 1), and a total score <4. Damaged cartilage regions were defined as follows: clefts within the cartilage structure (score of 3 or 4), presence of several cell clusters (score of 2), and a total score >6. Of the 9 pairs of cartilage samples obtained, 5 met these definitions. These paired samples were obtained from 5 OA patients (4 women and 1 man; age range 64–81 years). For real-time quantitative PCR, 4 of 6 additional pairs of cartilage samples met the above definitions. These additional samples were obtained from another group of 4 OA patients (4 women; age range 64–74 years).

RNA extraction.

Frozen cartilage tissue (∼50 mg) was rapidly minced with ophthalmic scissors in 1 ml of Isogen (Nippon Gene, Tokyo, Japan) and then immediately homogenized using a Polytron homogenizer (Hitachi Koki, Tokyo, Japan). Total RNA was isolated from the homogenized solution according to the manufacturer's instructions, except that the aqueous phase after initial separation was mixed with 0.25 ml of isopropanol, 0.25 ml of a high-salt precipitation solution (1.2M NaCl and 0.8M sodium citrate), and 3 μl of Ethachinmate (Nippon Gene) instead of with 0.5 ml of isopropanol. This extraction process was performed once again before RNA cleanup and DNase digestion using the RNeasy Mini kit (Qiagen, Crawley, UK).

Normalization of the amount of RNA.

Because the amount of total RNA obtained from the piece of cartilage tissue was too small to measure accurately using a spectrophotometer, a Human β-Actin Competitive PCR Set (Takara Bio, Kyoto, Japan) was used according to the manufacturer's instructions to normalize the RNA amount that would be available for microarray analysis. Total RNA isolated from cultured chondrocytes was used as a control. To confirm the integrity of the total RNA, 60 ng (normalized value) of denatured total RNA was loaded in each lane of a 1% nondenaturing (formaldehyde-free) agarose gel, subjected to electrophoresis, and stained with ethidium bromide (data not shown). Smaller amounts of total RNA can be detected in nondenaturing agarose gels than in formaldehyde-denaturing agarose gels.

High-density oligonucleotide array analysis.

Because the amount of RNA was too small to be applied to the standard protocol of the GeneChip system (Affymetrix, Santa Clara, CA), we used a double in vitro transcription technique, which can amplify the signals considerably. Biotin-labeled complementary RNA (cRNA) fragments were prepared as the target hybridized to the oligonucleotide array according to the Affymetrix protocol (GeneChip Eukaryotic Small Sample Target Labeling Technical Note). Briefly, double-stranded complementary DNA (cDNA) was synthesized from 50 ng (normalized value) of total RNA as the starting material. The cDNA was amplified by initial in vitro transcription, resulting in unlabeled cRNA. Double-stranded cDNA was also synthesized from the amplified cRNA. A second in vitro transcription was performed to produce biotin-labeled cRNA from the double-stranded cDNA. The biotin-labeled cRNA was fragmented before hybridization. Finally, 15 μg of the fragmented cRNA was hybridized to human U133A and U133B arrays, stained with streptavidin–phycoerythrin, and scanned with a GeneArray scanner.

Analysis of the data.

We normalized all intensity data by scaling the average signal intensity of 100 maintenance genes on each array to a value of 2,000 and then determined whether the transcript was detected (“present”), undetected (“absent”), or at the limit of detection (“marginal”) using Microarray Suite 5.0 software (Affymetrix). Comparison analysis was performed using Data Mining Tool 3.0 software (Affymetrix).

To identify the genes that showed the same expression pattern in different samples, the selection criteria (Figure 1) were defined as follows. First, selected transcripts had to show a ≥2-fold difference in the level of expression among all 5 samples or a ≥3-fold difference among 4 of the 5 samples. Second, selected transcripts had to be detected (“present”) and have normalized signal intensities >100 in either the intact or the damaged region. The “fold change” value represents the ratio of the average signal intensity in intact regions to the average signal intensity in damaged regions of all 5 samples. The standard error of the mean was calculated from the standard deviation of the fold change value and was determined by applying to the following formula each mean and SD of the signal intensities in the intact and damaged regions of all 5 samples:

  • equation image

where CV represents the coefficient of variation (the SD divided by the mean), CV1 represents the CV of the signal intensities in intact regions, and CV2 represents the CV of the signal intensities in damaged regions.

thumbnail image

Figure 1. Flow chart showing the procedure for identifying genes with different expression levels in intact and damaged regions of human osteoarthritic cartilage. The terms “present,” “marginal,” and “absent” represent expression levels of the transcripts described in Materials and Methods. Values are the mean ± SD number and percentage of transcripts. See Materials and Methods for details of the selection criteria.

Download figure to PowerPoint

The significance of the difference in levels of expression between the intact and damaged groups of cartilage was determined by paired t-test. Hierarchical clustering analysis was performed using the Cluster and TreeView software from Stanford University (available at http://rana.lbl.gov/EisenSoftware.htm).

Real-time quantitative PCR.

Total RNA was prepared in the same way as for the oligonucleotide array analysis. These RNA samples were converted into cDNA using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA) and random primer. Quantitative PCR reactions were performed using the ABI 7500 Real-Time PCR system (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Primer and probe sets were purchased as TaqMan Gene Expression Assays for the set of genes to be studied: Hs00248808 (CHRDL2), Hs00234422 (MMP2), Hs00155794 (APOD), Hs00182807 (FGF13), Hs00243202 (S100A4), Hs00932737 (TGFBI), Hs00610420 (PTGES), and Hs00200180 (TNFAIP6). The quantities of target genes were calculated using the standard curve method and were normalized with 18S RNA (ABI item no. 4319413E) as an internal control. The fold change and SEM values were calculated as described above.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Findings of the histopathologic assessment of cartilage samples.

First, we evaluated histopathologically the dissected cartilage fragments and scored them using the Mankin scale. We then determined whether each cartilage fragment satisfied the criteria for intact or damaged tissue as defined in Materials and Methods. Briefly, cartilage samples with a smooth surface and hyaline cartilage architecture and without fibrocartilage were defined as an intact region of OA cartilage. Cartilage samples with fibrillation, clefts, and several cell clusters were defined as a damaged region of OA cartilage. The histologic score for the damaged region (mean ± SD 7.8 ± 0.8) was significantly higher than that for the intact region (2.6 ± 0.5).

Findings of the genome-wide screening.

We used the Affymetrix GeneChip system to compare gene expression profiles in intact and damaged regions of the same OA knee joint. There are 44,692 probe sets on the Affymetrix human U133 chip set, covering >39,000 transcript variants that represent 33,000 human genes. The expression analysis detected ∼40% (mean ± SD 18,742 ± 968) of all probe sets corresponding to the transcripts in each cartilage sample (Figure 1), but little difference was noted between the percentage of transcripts expressed in intact and damaged regions of cartilage (intact 42.1 ± 2.5% [18,813 ± 1,106] versus damaged 41.8 ± 2.1% [18,671 ± 933]). About 3.5% (1,546 ± 328) of all probe sets indicated a ≥2-fold difference in expression per sample (intact > damaged 1.9 ± 1.1% [858 ± 481]; intact < damaged 1.5 ± 0.5% [688 ± 234]). This corresponds to ∼8% of the expressed transcripts (1,546/18,742 = 0.082). Furthermore, ∼10% of differentially expressed transcripts were commonly detected in all 5 OA samples (151/1,546 = 0.098).

Thus, we identified 114 genes, which consisted of 35 genes expressed in intact > damaged cartilage and 79 genes expressed in intact < damaged cartilage. These genes were then subdivided into functional categories: the 35 genes with higher expression in intact regions than in damaged regions were subdivided into 10 categories, and the 79 genes with higher expression in damaged regions than in intact regions were subdivided into 17 categories. Since ∼18% of all selected genes (21 of 114) have unknown function, these were assigned to the miscellaneous group, representing the largest category in the list. (Complete data for all of the selected genes are available upon request from the corresponding author.)

Differentially expressed genes in intact versus damaged regions.

Tables 1 and 2 show a representative gene from each functional subcategory: 10 genes that were highly expressed in intact regions and 17 genes that were highly expressed in damaged regions, respectively.

Table 1. Representative genes with higher expression in intact regions than in damaged regions of human osteoarthritic cartilage*
Category, gene nameGene symbolPublic IDFold changeSEMP
  • *

    See Materials and Methods for the definition of “fold change” and for the method of calculating the SEM. Complete data for all of the selected genes are available upon request from the corresponding author. Public ID = accession number in public databases (RefSeq or GenBank).

  • All P values were significant (<0.05), as determined by paired t-test, except for the P values for genes CHRDL2, C4BPA, and EST.

Transcription factors     
 v-ets erythroblastosis virus E26 oncogene homolog 1ETS1NM_0052383.641.350.007
Signal transduction     
 Guanylate cyclase 1, soluble, α3 subunitGUCY1A3NM_0008568.172.600.028
Membrane proteins/receptors     
 Glypican 5GPC5NM_0044664.821.460.010
Transporters/carrier proteins     
 Apolipoprotein DAPODNM_0016476.241.150.004
Cytokines/growth factors     
 Fibroblast growth factor 13FGF13NM_0041144.470.730.001
Proteases     
 Matrix metalloproteinase 2MMP2NM_0045307.302.900.017
Secreted inhibitors/antagonists     
 Chordin-like 2CHRDL2NM_0154248.955.350.060
Immunity/defense     
 Complement component 4 binding protein α subunitC4BPANM_0007156.443.530.121
Extracellular matrix proteins     
 VitrinVITNM_0532769.343.090.031
Miscellaneous     
 Expressed sequence tagESTBF00262514.016.140.056
Table 2. Representative genes with higher expression in damaged regions than in intact regions of human osteoarthritic cartilage*
Category, gene nameGene symbolPublic IDFold changeSEMP
  • *

    See Materials and Methods for the definition of “fold change” and for the method of calculating the SEM. Complete data for all of the selected genes are available upon request from the corresponding author. Public ID = accession number in public databases (RefSeq or GenBank).

  • All P values were significant (<0.05), as determined by paired t-test, except for the P value for gene EPB41L3.

Transcription factor     
 Sex determining region Y–type high mobility group box 11SOX11NM_00310830.197.660.009
Cell cycle     
 Cyclin D1CCND1NM_0530566.962.250.006
Metabolism     
 Phosphoserine aminotransferase 1PSAT1NM_0581799.824.770.009
Collagen synthesis     
 Prolyl 4-hydroxylase α polypeptide IIIP4HA3NM_1829049.663.440.021
Transporters/carrier proteins     
 Uncoupling protein 2UCP2NM_00335510.565.880.017
Glycosylation     
 UDP-N-acetyl-α-D-galactosamine:polypeptideGALNTL1XM_03110412.162.970.010
  N-acetylgalactosaminyltransferase–like 1     
Cell motility/invasion     
 S100 calcium binding protein A4S100A4NM_0029615.331.730.007
Signal transduction     
 Down syndrome critical region gene 1DSCR1NM_0044144.690.820.008
Membrane proteins/receptors     
 Triggering receptor expressed on myeloid cells 1TREM1NM_01864315.384.430.020
Secreted inhibitors/antagonists     
 Tumor necrosis factor α–induced protein 6TNFAIP6NM_00711525.345.780.001
Growth factors/cytokines     
 Cytokine receptor–like factor 1CRLF1NM_00475022.907.980.031
Immunity/defense     
 Prostaglandin E synthasePTGESNM_00487812.443.080.013
Proteases     
 A disintegrin and metalloproteinase domain 12ADAM12NM_0034744.431.590.024
Structural proteins     
 Erythrocyte membrane protein band 4.1–like 3EPB41L3NM_0123075.362.880.070
Adhesion proteins     
 Transforming growth factor β–induced 68 kdTGFBINM_00035812.053.740.001
Extracellular matrix proteins     
 C-type lectin domain family 3, member BCLEC3BNM_00327811.165.450.030
Miscellaneous     
 Hypothetical protein FLJ37034FLJ37034H162587.652.460.019

The following genes were highly expressed in the intact region of OA cartilage (Table 1). The v-Ets-1 erythroblastosis virus E26 oncogene homolog (ETS1; 3.64-fold) is a member of the Ets transcription factor family and can activate the expression of many matrix metalloproteinases (MMPs) (9, 10). Soluble guanylate cyclase 1α3 subunit (GUCY1A3; 8.17-fold) functions as the main receptor for nitric oxide. Glypican 5 (GPC5; 4.82-fold) is a glycosyl phosphatidylinositol (GPI)–anchored heparan sulfate proteoglycan and functions as a “coreceptor” for several “heparin-binding” growth factors (11). Apolipoprotein D (APOD; 6.24-fold) is a member of the lipocalin superfamily of transporter proteins that bind progesterone and arachidonic acid (12). Fibroblast growth factor 13 (FGF13; 4.47-fold), which is also called fibroblast growth factor homologous factor 2, is an intracellular protein that unlike other FGFs, cannot activate any FGF receptors (13). MMP-2 (MMP2; 7.3-fold) is an enzyme known to degrade denatured collagen. Chordin-like 2 (CHRDL2; 8.95-fold) is a novel chordin-like bone morphogenetic protein inhibitor expressed preferentially in chondrocytes of developing cartilage and OA cartilage (14). C4 binding protein α subunit (C4BPA; 6.44-fold) is a regulatory protein of the classical pathway of complement. Vitrin (VIT; 9.34-fold) is a member of the von Willebrand factor A superfamily, and it might serve as a bridge between collagen fibrils and the hyaluronan network (15). The function of gene BF002625 (14.01-fold) remains unknown. Hierarchical clustering of this group of miscellaneous genes did not show characteristic changes (Figure 2A).

thumbnail image

Figure 2. Hierarchical clustering of A, 35 genes that were highly expressed in intact regions of human osteoarthritic (OA) cartilage (green) and B, 79 genes that were highly expressed in damaged regions of human OA cartilage (red). Black areas indicate no difference in the level of expression between intact and damaged regions of human OA cartilage.

Download figure to PowerPoint

The following genes were highly expressed in the damaged region of OA cartilage (Table 2). Sex determining region Y–type high mobility group box 11 (SOX11; 30.19-fold) plays an important role in tissue remodeling, including craniofacial and skeletal tissue remodeling (16). Cyclin D1 (CCND1; 6.96-fold) is an essential component of chondrocyte proliferation in the growth plate (17). Phosphoserine aminotransferase 1 (PSAT; 9.82-fold) is an enzyme involved in the biosynthesis of L-serine, and the messenger RNA (mRNA) level for PSAT appears to be up-regulated to support cell proliferation (18). Prolyl 4-hydroxylase α polypeptide III (P4HA3; 9.66-fold) is a component of prolyl 4-hydroxylase, a key enzyme in collagen synthesis (19). Uncoupling protein 2 (UCP2; 10.56-fold) is a mitochondrial anion carrier protein that also protects against oxidative stress–induced cell death (20). Polypeptide N-acetylgalactosaminyltransferase–like 1 (GALNTL1; 12.16-fold) is an enzyme that catalyzes the initial reaction in the biosynthesis of O-linked oligosaccharide. S100 calcium binding protein A4 (S100A4; 5.33-fold) is involved in cell proliferation and tumor metastasis (21). Down syndrome critical region gene 1 (DSCR1; 4.69-fold) acts as a negative regulator of calcineurin signaling, which is known to induce chondrogenesis (22, 23). Triggering receptor expressed on myeloid cells 1 (TREM1; 15.38-fold) plays a role in acute inflammation. Tumor necrosis factor α–induced protein 6 (TNFAIP6; 25.34-fold) forms complexes with inter-α-inhibitor (a serine protease inhibitor) and inhibits MMP activation via antiplasmin activity (24, 25). Cytokine receptor–like factor 1 (CRLF1; 22.9-fold) is expressed in chondrocytes and osteoblasts and plays a role in cartilage and bone formation (26). Prostaglandin E (PGE) synthase (PTGES; 12.44-fold), which is also called microsomal PGE synthase 1 (27), is responsible for the production of PGE2, which has a wide variety of biologic activities. A disintegrin and metalloproteinase domain 12 (ADAM12; 4.43-fold) is one of the candidate genes in OA susceptibility and progression (28). Erythrocyte membrane protein band 4.1–like 3 (EPB41L3; 5.36-fold) is a structural protein that links transmembrane proteins to the actin cytoskeleton. Transforming growth factor β–induced 68 kd (TGFBI; 12.05-fold) may play an important role in cell–collagen interactions in cartilage (29). C-type lectin domain family 3, member B (COLEC3B; 11.16-fold) is a plasminogen-binding protein that plays a role in tissue growth and remodeling (30). The function of gene FLJ37034 (7.65-fold) remains unknown.

There is a characteristic change in the group of genes with higher expression in damaged cartilage. Three interstitial collagen genes (collagen, type I, α1 [COL1A1]; collagen, type I, α2 [COL1A2]; and collagen, type V, α1 [COL5A1]) together with 4 collagen biosynthetic enzymes (P4HA3 [19]; lysyl oxidase–like 2 [LOXL2] [31]; leprecan-like 1 [LEPREL1] [32], and LOXL3 [33]) were expressed at significantly higher levels in chondrocytes from the damaged region of OA cartilage than those from the intact region (Complete data for all of the selected genes are available upon request from the corresponding author.) Furthermore, hierarchical clustering analysis indicated that COL1A2, COL5A1, LOXL2, and LOXL3 as well as other minor collagen genes belonged to the same cluster in the tree diagram (Figure 2B). This cluster also contained 2 proliferation-associated genes (S100A4 and PSAT1).

Results of real-time quantitative PCR of selected 8 genes.

To validate the oligonucleotide array data, real-time quantitative PCR was performed on 8 selected genes using additional 4 pairs of cartilage samples (Figure 3). For these 8 genes, the region of high expression as determined by real-time quantitative PCR was consistent with the region of high expression as determined by oligonucleotide array analysis. We found that the levels of expression of mRNA for FGF13, TGFBI, PTGES, and TNFAIP6 genes in intact and damaged regions of OA cartilage were significantly different as determined by both methods. In the present study, however, the fold change in gene expression estimated using real-time quantitative PCR tended to be smaller than that estimated using oligonucleotide array analysis.

thumbnail image

Figure 3. Histogram showing levels of expression of 8 selected genes, as measured by oligonucleotide array (open bars; n = 5) and real-time quantitative polymerase chain reaction (solid bars; n = 4). Values are the mean ± SEM (see Materials and Methods for further details). ∗ = P < 0.05; † = P < 0.01 for intact region versus damaged region, by paired t-test.

Download figure to PowerPoint

Expression patterns of genes reported to be differentially expressed in intact versus damaged regions of OA cartilage and in normal versus OA cartilage.

We assessed the microarray data for genes that in previous studies were shown to be differentially expressed in intact versus damaged regions of OA cartilage from the same joint or in normal versus OA cartilage from different patients (4–7, 14, 34–38). The name and symbol of the reported genes are shown in Table 3. Table 3 also shows the sample characterization (comparison being made) and the gene expression pattern reported in previous studies, as well as the results of oligonucleotide array analysis for each gene.

Table 3. Expression patterns of genes reported to be differentially expressed in intact versus damaged regions of cartilage as well as in normal versus OA cartilage*
Comparison, expression patternRef.Gene name (alias)Gene symbolFold changeSEMP
  • *

    OA = osteoarthritic; ND = not detected.

  • Positive numbers represent high expression in damaged regions; negative numbers represent high expression in intact regions.

  • Included in the list of 114 genes selected by our criteria. Complete data for all of the selected genes are available upon request from the corresponding author.

Intact versus damaged regions of cartilage from the same joint      
 Intact < damaged4Nerve growth factor, β polypeptide (NGF)NGFB5.081.540.011
 4Neurotrophic tyrosine kinase receptor type I (p140 TrkA)NTRK1ND  
 5Insulin-like growth factor binding protein 3IGFBP35.942.710.056
 5Insulin-like growth factor binding protein 4IGFBP43.931.200.006
 5Insulin-like growth factor binding protein 5IGFBP51.170.330.611
 6Heparan sulfate proteoglycan 2 (perlecan)HSPG22.220.800.074
 Intact > damaged7Matrix metalloproteinase 13MMP13−9.519.340.32
 7Hyaluronan and proteoglycan link protein 1 (link protein)HAPLN1−1.940.450.043
 7B cell CLL/lymphoma 2BCL21.050.340.828
 7Sex determining region Y–type high mobility group box 9SOX91.080.180.685
Normal versus OA cartilage      
 Normal < OA34Tumor necrosis factor α–induced protein 6 (TSG-6)TNFAIP625.345.780.001
 35Serine protease 11 (HtrA1)PRSS113.080.770.012
 36Prostaglandin E synthasePTGES12.443.080.013
 37S100 calcium binding protein A4S100A45.331.730.007
 37Fibronectin 1FN16.161.490.002
 37Transforming growth factor β–induced 68 kd (BIGH3)TGFBI12.053.740.001
 37Collagen, type I, α2COL1A25.673.190.027
 37Matrix metalloproteinase 2MMP2−7.302.900.017
 14Chordin-like 2 (CHL2)CHRDL2−8.955.350.06
 Normal > OA38Serine proteinase inhibitor, clade E, member 1 (PAI-1)SERPINE114.388.090.016

Consistent with studies that compared intact and damaged regions of OA cartilage from the same joint, we confirmed high expression of NGFB (4), IGFBP3, IGFBP4 (5), and HSPG2 (6) in the damaged region, as well as high expression of MMP13 and HAPLN1 (7) in the intact region of OA cartilage. However, we did not observe NTRK1 gene expression itself or differential expression of IGFBP5, BCL2, and SOX9.

We also evaluated whether the results of comparisons between intact and damaged regions of cartilage from the same joint resembled the results of comparisons between normal and OA cartilage obtained from different individuals. At least 7 genes reported to be highly expressed in OA cartilage compared with normal cartilage (TNFIP6 [34], PRSS11 [35], PTGES [36], S100A4, FN1, TGFBI, and COL1A2 [37]) were highly expressed in the damaged region compared with the intact region of OA cartilage. However, there were some exceptions. According to our data, the 2 genes reported to be highly expressed in OA cartilage (MMP2 [37] and CHRDL2 [14]) were highly expressed in the intact region, and SERPINE1 (38), which was reported to be expressed at low levels in OA cartilage, was highly expressed in the damaged region of OA cartilage.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

In this study, we compared the gene expression profiles in intact versus damaged regions of OA cartilage. Transcripts with a ≥2-fold difference in mRNA expression between these 2 regions accounted for an average of 8% of all expressed transcripts per OA cartilage tissue sample, ∼10% of which were commonly detected in the 5 patient samples. The former observation indicates that the gene expression profile of chondrocytes in the intact region is quite different from that of chondrocytes in the damaged region, even though both regions are in contact with the same synovial fluid and have the same genetic background. The latter finding suggests that the gene expression profiles of chondrocytes in OA cartilage change in a region-specific manner. This also suggests the presence of common molecular mechanisms of OA development, assuming that the changes in gene expression patterns of chondrocytes lead to cartilage degeneration.

Using our selection criteria, we identified 114 genes that were differentially expressed in the intact region versus damaged region of OA cartilage. Thirty-five of these genes were up-regulated in the intact region, and 79 genes were up-regulated in the damaged region. The validity of our data was confirmed using real-time quantitative PCR and comparing our findings with those of previous studies. The possible roles of these genes with significantly altered mRNA expression are discussed here in terms of the reaction patterns of chondrocytes in order to determine the underlying mechanisms that may participate in the pathogenesis or progression of OA. Sandell and Aigner (39) described 5 categories of cellular reaction patterns related to OA development: phenotypic modulation of articular chondrocytes, formation of osteophytes, chondrocyte proliferation and apoptosis, matrix synthetic activity of chondrocytes, and matrix degradation activity of chondrocytes. The latter 3 are discussed below.

With regard to chondrocyte proliferation and apoptosis, the CCND1, PSAT, and S100A4 genes are known to be involved in cell proliferation, as mentioned above (17, 18, 21). These genes were more highly expressed in damaged regions than in intact regions of OA cartilage. This finding is consistent with the pathologic feature showing some clusters of chondrocytes in the surface layer of damaged cartilage, and it supports the hypothesis that changes in gene expression patterns of chondrocytes lead to the pathologic condition of OA in cartilage. Yet, among the 114 genes we identified, no genes were clearly related to apoptosis. (Complete data for all of the selected genes are available upon request from the corresponding author.) However, we cannot suggest that there are no differences in the activation of the apoptosis signaling pathway between intact and damaged regions of OA cartilage, because the proteolytic cascade, rather than transcriptional regulation, is important in apoptotic signaling (40).

With regard to the matrix synthetic activity of chondrocytes, our microarray data showed that although the signal intensities of type II collagen and aggrecan were very high in both intact and damaged regions of OA cartilage, there were no differences in gene expression between these 2 regions (data not shown). However, we found that 3 interstitial collagen genes (COL1A1, COL1A2, and COL5A1) and 4 genes for enzymes involved in the collagen biosynthetic pathway (P4HA3, LOXL2, LEPREL1, and LOXL3) were highly expressed in the damaged region of OA cartilage. Interestingly, hierarchical clustering analysis showed that 4 of these 7 genes (COL1A2, COL5A1, LOXL2, and LOXL3) belonged to the same cluster as proliferation-associated genes (S100A4 and PSAT1). This suggests that wound healing, including the process of cell proliferation and interstitial collagen synthesis, occurs in damaged OA cartilage, where the expression of genes related to wound healing might be regulated in the same manner.

With regard to the matrix degradation activity of chondrocytes, the most obvious pathologic feature in the damaged region is advanced cartilage destruction. Therefore, we predicted that expression levels of protease genes would be higher in damaged regions than in intact regions. Consistent with this expectation, we identified 4 proteases (DKFZP586H2123, ADAMTS6, ADAM12, and PRSS11) with high expression in the damaged region of OA cartilage among the selected 114 genes. However, MMP-2, which can degrade the extracellular matrix, is only one protease showing high expression in the intact region of OA cartilage. In particular, the PRSS11 (HtrA1) gene was recently reported to enhance cartilage degeneration via digestion of major cartilage components (41), and the single-nucleotide polymorphisms of the ADAM12 gene are associated with the progression of knee OA (28). However, there is a discrepancy between the function of the detected genes and the histopathologic features. The expression levels of 3 genes known to inhibit degradation of the extracellular matrix (TNFAIP6, SERPINE1, and TIMP3) were significantly high in the damaged region compared with the intact region of OA cartilage. These molecules are probably up-regulated to prevent the progression of cartilage destruction in the damaged areas of OA cartilage.

We subsequently examined whether the results of comparisons between intact and damaged regions of OA cartilage from the same joint resembled the results of comparisons between normal and OA cartilage obtained from different persons. We found that comparisons between normal and OA cartilage and between intact and damaged regions of OA cartilage yielded similar results with regard to the expression pattern of 7 of the 10 genes examined (Table 3). This suggests that during the transition from normal cartilage to OA lesional cartilage, the gene expression profile changes before there is any apparent damage to the cartilage. Because the expression levels of some genes probably change during the transitional period from normal to normal-appearing cartilage, our study design might have allowed us to miss some important genes that show no differences in expression levels between intact and damaged regions. To overcome this problem, further studies comparing these samples with normal cartilage samples from normal joints are needed.

What causes the OA-specific pattern of gene expression? Several possible mechanisms have been investigated, such as mechanical stress, cytokine stimulation, cell–matrix interaction, hypoxia, and reoxygenation. One of the strongest candidates is mechanical stress, because the damaged cartilage region is usually subjected to mechanical loading, whereas the intact region is not. In particular, chondrocytes residing in damaged regions are susceptible to mechanical stress because the tensile properties of the damaged cartilage are lost as a result of destruction of the collagen network (42). Proinflammatory cytokines, especially interleukin-1 (IL-1) and TNFα, are also closely related to the development of OA (3). We thought that these cytokines might be accessible to the chondrocytes in damaged cartilage. However, some IL-1–induced genes were detected in both regions (in the intact region, ETS1, GUCY1A3, C4BPA, PBEF, and APOD; in the damaged region, TNFAIP6, PTGES, FN1, NGFB, and TNFSF11). Furthermore, although MMP-2 mRNA is not significantly up-regulated by treatment with IL-1 (43), our microarray data showed that MMP-2 mRNA was expressed 7.3-fold higher in intact regions than in damaged regions of OA cartilage. Therefore, the effects of the cytokine alone could not account for these complex conditions of OA in vivo. Up-regulation of MMP-2 mRNA might be the result of other factors, including hypoxia/reoxygenation, cell–matrix interactions, and intermittent hydrostatic pressures (44–46).

In conclusion, our study demonstrated a clear difference in the gene expression profile in damaged regions of human OA cartilage compared with that in intact regions. The pattern of differences between these 2 regions was similar among 5 pairs of OA cartilage samples. This finding implies that there is a common mechanism responsible for the destruction and maintenance of the articular cartilage in OA. Elucidation of this mechanism is important for the development of effective treatments for OA.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank Dr. A. Shibakawa for providing the human cartilage specimens, and N. Furuya, S. Asada, N. Takagi, S. Shinkawa, Y. Nakagawa, T. Mogi, H. Ogasawara, and M. Yamanashi for excellent technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    Dieppe P, Kirwan J. The localization of osteoarthritis. Br J Rheumatol 1994; 33: 2013.
  • 2
    Ahlback S. Osteoarthrosis of the knee: a radiographic investigation. Acta Radiol Diagn (Stockh) 1968; 227 Suppl: 772.
  • 3
    Poole AR. Cartilage in health and disease. In: KoopmanWJ, editor. Arthritis and allied conditions: a textbook of rheumatology. 14th ed. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 22684.
  • 4
    Iannone F, De Bari C, Dell'Accio F, Covelli M, Patella V, Lo Bianco G, et al. Increased expression of nerve growth factor (NGF) and high affinity NGF receptor (p140 TrkA) in human osteoarthritic chondrocytes. Rheumatology (Oxford) 2002; 41: 14138.
  • 5
    Iwanaga H, Matsumoto T, Enomoto H, Okano K, Hishikawa Y, Shindo H, et al. Enhanced expression of insulin-like growth factor-binding proteins in human osteoarthritic cartilage detected by immunohistochemistry and in situ hybridization. Osteoarthritis Cartilage 2005; 13: 43948.
  • 6
    Tesche F, Miosge N. Perlecan in late stages of osteoarthritis of the human knee joint. Osteoarthritis Cartilage 2004; 12: 85262.
  • 7
    Yagi R, McBurney D, Laverty D, Weiner S, Horton WE Jr. Intrajoint comparisons of gene expression patterns in human osteoarthritis suggest a change in chondrocyte phenotype. J Orthop Res 2005; 23: 112838.
  • 8
    Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteo-arthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971; 53: 52337.
  • 9
    Wasylyk C, Gutman A, Nicholson R, Wasylyk B. The c-Ets oncoprotein activates the stromelysin promoter through the same elements as several non-nuclear oncoproteins. EMBO J 1991; 10: 112734.
  • 10
    Sun HB, Yokota H. Messenger-RNA expression of matrix metalloproteinases, tissue inhibitors of metalloproteinases, and transcription factors in rheumatic synovial cells under mechanical stimuli. Bone 2001; 28: 3039.
  • 11
    Saunders S, Paine-Saunders S, Lander AD. Expression of the cell surface proteoglycan glypican-5 is developmentally regulated in kidney, limb, and brain. Dev Biol 1997; 190: 7893.
  • 12
    Morais Cabral JH, Atkins GL, Sanchez LM, Lopez-Boado YS, Lopez-Otin C, Sawyer L. Arachidonic acid binds to apolipoprotein D: implications for the protein's function. FEBS Lett 1995; 366: 536.
  • 13
    Goldfarb M. Fibroblast growth factor homologous factors: evolution, structure, and function. Cytokine Growth Factor Rev 2005; 16: 21520.
  • 14
    Nakayama N, Han CY, Cam L, Lee JI, Pretorius J, Fisher S, et al. A novel chordin-like BMP inhibitor, CHL2, expressed preferentially in chondrocytes of developing cartilage and osteoarthritic joint cartilage. Development 2004; 131: 22940.
  • 15
    Mayne R, Ren ZX, Liu J, Cook T, Carson M, Narayana S. VIT-1: the second member of a new branch of the von Willebrand factor A domain superfamily. Biochem Soc Trans 1999; 27: 8325.
  • 16
    Sock E, Rettig SD, Enderich J, Bosl MR, Tamm ER, Wegner M. Gene targeting reveals a widespread role for the high-mobility-group transcription factor Sox11 in tissue remodeling. Mol Cell Biol 2004; 24: 663544.
  • 17
    Beier F, Ali Z, Mok D, Taylor AC, Leask T, Albanese C, et al. TGFβ and PTHrP control chondrocyte proliferation by activating cyclin D1 expression. Mol Biol Cell 2001; 12: 385263.
  • 18
    Baek JY, Jun Do Y, Taub D, Kim YH. Characterization of human phosphoserine aminotransferase involved in the phosphorylated pathway of L-serine biosynthesis. Biochem J 2003; 373: 191200.
  • 19
    Kukkola L, Hieta R, Kivirikko KI, Myllyharju J. Identification and characterization of a third human, rat, and mouse collagen prolyl 4-hydroxylase isoenzyme. J Biol Chem 2003; 278: 4768593.
  • 20
    Teshima Y, Akao M, Jones SP, Marban E. Uncoupling protein-2 overexpression inhibits mitochondrial death pathway in cardiomyocytes. Circ Res 2003; 93: 192200.
  • 21
    Sherbet GV, Lakshmi MS. S100A4 (MTS1) calcium binding protein in cancer growth, invasion and metastasis. Anticancer Res 1998; 18: 241521.
  • 22
    Fuentes JJ, Genesca L, Kingsbury TJ, Cunningham KW, Perez-Riba M, Estivill X, et al. DSCR1, overexpressed in Down syndrome, is an inhibitor of calcineurin-mediated signaling pathways. Hum Mol Genet 2000; 9: 168190.
  • 23
    Tomita M, Reinhold MI, Molkentin JD, Naski MC. Calcineurin and NFAT4 induce chondrogenesis. J Biol Chem 2002; 277: 422148.
  • 24
    Maier R, Wisniewski HG, Vilcek J, Lotz M. TSG-6 expression in human articular chondrocytes: possible implications in joint inflammation and cartilage degradation. Arthritis Rheum 1996; 39: 5529.
  • 25
    Wisniewski HG, Hua JC, Poppers DM, Naime D, Vilcek J, Cronstein BN. TNF/IL-1-inducible protein TSG-6 potentiates plasmin inhibition by inter-α-inhibitor and exerts a strong anti-inflammatory effect in vivo. J Immunol 1996; 156: 160915.
  • 26
    Clancy BM, Johnson JD, Lambert AJ, Rezvankhah S, Wong A, Resmini C, et al. A gene expression profile for endochondral bone formation: oligonucleotide microarrays establish novel connections between known genes and BMP-2-induced bone formation in mouse quadriceps. Bone 2003; 33: 4663.
  • 27
    Jakobsson PJ, Thoren S, Morgenstern R, Samuelsson B. Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci U S A 1999; 96: 72205.
  • 28
    Valdes AM, Hart DJ, Jones KA, Surdulescu G, Swarbrick P, Doyle DV, et al. Association study of candidate genes for the prevalence and progression of knee osteoarthritis. Arthritis Rheum 2004; 50: 2497507.
  • 29
    Hashimoto K, Noshiro M, Ohno S, Kawamoto T, Satakeda H, Akagawa Y, et al. Characterization of a cartilage-derived 66-kDa protein (RGD-CAP/β ig-h3) that binds to collagen. Biochim Biophys Acta 1997; 1355: 30314.
  • 30
    Iba K, Durkin ME, Johnsen L, Hunziker E, Damgaard-Pedersen K, Zhang H, et al. Mice with a targeted deletion of the tetranectin gene exhibit a spinal deformity. Mol Cell Biol 2001; 21: 781725.
  • 31
    Vadasz Z, Kessler O, Akiri G, Gengrinovitch S, Kagan HM, Baruch Y, et al. Abnormal deposition of collagen around hepatocytes in Wilson's disease is associated with hepatocyte specific expression of lysyl oxidase and lysyl oxidase like protein-2. J Hepatol 2005; 43: 499507.
  • 32
    Vranka JA, Sakai LY, Bachinger HP. Prolyl 3-hydroxylase 1, enzyme characterization and identification of a novel family of enzymes. J Biol Chem 2004; 279: 2361521.
  • 33
    Jourdan-Le Saux C, Tomsche A, Ujfalusi A, Jia L, Csiszar K. Central nervous system, uterus, heart, and leukocyte expression of the LOXL3 gene, encoding a novel lysyl oxidase-like protein. Genomics 2001; 74: 2118.
  • 34
    Bayliss MT, Howat SL, Dudhia J, Murphy JM, Barry FP, Edwards JC, et al. Up-regulation and differential expression of the hyaluronan-binding protein TSG-6 in cartilage and synovium in rheumatoid arthritis and osteoarthritis. Osteoarthritis Cartilage 2001; 9: 428.
  • 35
    Hu SI, Carozza M, Klein M, Nantermet P, Luk D, Crowl RM. Human HtrA, an evolutionarily conserved serine protease identified as a differentially expressed gene product in osteoarthritic cartilage. J Biol Chem 1998; 273: 3440612.
  • 36
    Li X, Afif H, Cheng S, Martel-Pelletier J, Pelletier JP, Ranger P, et al. Expression and regulation of microsomal prostaglandin E synthase-1 in human osteoarthritic cartilage and chondrocytes. J Rheumatol 2005; 32: 88795.
  • 37
    Aigner T, Zien A, Gehrsitz A, Gebhard PM, McKenna L. Anabolic and catabolic gene expression pattern analysis in normal versus osteoarthritic cartilage using complementary DNA-array technology. Arthritis Rheum 2001; 44: 277789.
  • 38
    Martel-Pelletier J, Faure MP, McCollum R, Mineau F, Cloutier JM, Pelletier JP. Plasmin, plasminogen activators and inhibitor in human osteoarthritic cartilage. J Rheumatol 1991; 18: 186371.
  • 39
    Sandell LJ, Aigner T. Articular cartilage and changes in arthritis: an introduction: cell biology of osteoarthritis. Arthritis Res 2001; 3: 10713.
  • 40
    Morgan D, Murray A, Hunt T, Nurse P. The cell cycle and programmed cell death. In: AlbertsB, JohnsonA, LewisJ, RaffM, RobertsK, WalterP, editors. Molecular biology of the cell. 4th ed. New York: Garland Science; 2002. p. 9831026.
  • 41
    Tsuchiya A, Yano M, Tocharus J, Kojima H, Fukumoto M, Kawaichi M, et al. Expression of mouse HtrA1 serine protease in normal bone and cartilage and its upregulation in joint cartilage damaged by experimental arthritis. Bone 2005; 37: 32336.
  • 42
    Kempson GE. Age-related changes in the tensile properties of human articular cartilage: a comparative study between the femoral head of the hip joint and the talus of the ankle joint. Biochim Biophys Acta 1991; 1075: 22330.
  • 43
    Duerr S, Stremme S, Soeder S, Bau B, Aigner T. MMP-2/gelatinase A is a gene product of human adult articular chondrocytes and is increased in osteoarthritic cartilage. Clin Exp Rheumatol 2004; 22: 6038.
  • 44
    Martin G, Andriamanalijaona R, Grassel S, Dreier R, Mathy-Hartert M, Bogdanowicz P, et al. Effect of hypoxia and reoxygenation on gene expression and response to interleukin-1 in cultured articular chondrocytes. Arthritis Rheum 2004; 50: 354960.
  • 45
    Zemmyo M, Meharra EJ, Kuhn K, Creighton-Achermann L, Lotz M. Accelerated, aging-dependent development of osteoarthritis in α1 integrin–deficient mice. Arthritis Rheum 2003; 48: 287380.
  • 46
    Trindade MC, Shida J, Ikenoue T, Lee MS, Lin EY, Yaszay B, et al. Intermittent hydrostatic pressure inhibits matrix metalloproteinase and pro-inflammatory mediator release from human osteoarthritic chondrocytes in vitro. Osteoarthritis Cartilage 2004; 12: 72935.