View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Osteoarthritis and Cartilage (2001) 9, 112–118 © 2001 OsteoArthritis Research Society International 1063–4584/01/020112+07 $35.00/0 doi:10.1053/joca.2000.0366, available online at http://www.idealibrary.com on Quantitative analysis of gene expression in human articular cartilage from normal and osteoarthritic joints I. Martin, M. Jakob, D. Scha¨fer, W. Dick, G. Spagnoli and M. Heberer Department of Surgery, Research Division, University of Basel, Switzerland Summary Objective: To quantify the expression of genes encoding extracellular matrix (ECM) proteins in human cartilage from normal and osteoarthritic (OA) joints. Design: Human cartilage samples were classified as control (CTR) or OA according to clinical evaluation and assessed histologically and biochemically to confirm the diagnosis. mRNAs encoding collagen types I, II and X, aggrecan, versican, osteopontin and osteocalcin were quantified by real-time reverse transcription-PCR assays and normalized to a reference mRNA (GAPDH). Results: RNA from native cartilage could be reproducibly and efficiently amplified by real-time PCR only if isolated using purification membranes. Primers and fluorescent probes for real-time PCR, endowed with comparable (<6% difference from GAPDH) and high (>91%) amplification efficiencies, were designed and validated for the selected ECM genes. The expression of most genes under investigation displayed large variations and was not significantly different in CTR and OA cartilage. Only osteopontin mRNA levels were significantly higher in OA than CTR specimens. mRNA ratios of collagen type II to I and of aggrecan to versican, defined as indexes of chondrocyte differentiation, were less variable within each population than the single genes and markedly higher (27.0 and 7.6-fold, respectively) in CTR than OA cartilage, with high statistical significance (P=0.00013 and P=0.00007, respectively). Conclusions: Our results provide evidence that gene patterns related to chondrocyte differentiation discriminate between CTR and OA human cartilage with higher sensitivity than single ECM genes. The method described here has the potential to improve understanding of the progression of OA and could become a valuable diagnostic tool. © 2001 OsteoArthritis Research Society International Key words: Real-time PCR, Chondrocyte, Extracellular matrix, Differentiation. Introduction OA cartilage samples7, and that collagen type I was abundantly expressed by chondrocytes of deep zones of The unique mechanical properties of articular cartilage OA fibrocartilaginous tissue8. Also the expression of differ- depend on the composition and organization of its extra- ent types of proteoglycans, investigated using a competi- cellular matrix (ECM). It is generally agreed that the pro- tive PCR technique, was not always consistent when gressive functional failure of cartilage in osteoarthrosis different sets of samples were used9,10. (OA) is not only due to enhanced degradation of the main The discordance among the different studies could be tissue components (e.g. type II collagen and aggrecan), but also to discoordinate changes in the types and amounts of due not only to variability among different individuals, but newly synthesized ECM molecules1. Thus, investigation of also to limitations inherent to the techniques used. Northern OA chondrocyte anabolic activity by quantification of the blotting and in-situ hybridization methods are not sensi- expression of mRNAs encoding specific ECM molecules tive enough to detect low-level gene expression and not should help understand the progression of the pathology, accurate enough to quantify the full range of expression. An and may ultimately improve its diagnosis and treatment. amplification step may therefore be required to quantify Despite the wealth of knowledge on the structure of mRNA amounts from limited tissue biopsies. However, different cartilage ECM molecules and their genes, there is quantification of mRNA using conventional PCR techniques relatively little and sometimes contradictory information on may be of limited accuracy, both because of the detection the quantitation of their expression in normal and OA method used (i.e. semi-quantitative image analysis), and articular cartilage. It was reported that mRNAs for collagen because analysis is often performed using a constant types I and X are absent in normal adult cartilage and that number of amplification cycles, after which the system collagen types II and X, but not I, are consistently could be in a saturation phase. up-regulated in specific layers of OA tissue2–6. However, Recently, real-time PCR has enabled the quantification independent studies performed using the same technique of mRNA with high accuracy, reproducibility and sensitivity in a wide dynamic range, without the need of post-PCR (i.e. in situ hybridization) reported that collagen type II 11 mRNA was detectable in only about 50% of the processed processing . In this paper, we developed and vali- dated real-time PCR assays to quantify the expression of genes encoding the most abundant molecules in the ECM Received 28 January 2000; accepted 16 May 2000 Address correspondence to: Ivan Martin, Research Division of of normal articular cartilage (i.e. collagen type II and the Department of Surgery, University of Basel, Hebelstrasse 20, aggrecan) and of degenerated fibrocartilage (i.e. collagen ZLF, Room 405, 4031 Basel, Switzerland. Tel.: + 41 61 265 2384; type I and versican), as well as of typical markers Fax: + 41 61 265 3990; E-mail: [email protected] of chondrocyte hypertrophy (i.e. collagen type X and 112 Osteoarthritis and Cartilage Vol. 9 No. 2 113 osteopontin) and of bone cells (i.e. osteocalcin). Our (1 mg/ml) containing 50 mM Tris, 1 mM EDTA, 1 mM iodo- hypothesis was that real-time PCR can lead to the identifi- acetamide, and 10 g/ml pepstatin A at 56°C for 15 h. cation of quantitative patterns of ECM gene expression, Proteinase K was then inactivated by boiling the digest for typical for normal and OA adult human cartilage tissue. 12 min. The GAG content was determined spectrophoto- metrically using dimethylmethylene blue dye and bovine chondroitin sulfate as a standard15. Collagen type II was Methods quantified using a monoclonal antibody-based inhibition ELISA14. PATIENTS AND SAMPLES Patient’s characteristics REAL-TIME QUANTITATIVE PCR Human articular OA cartilage was obtained from femoral heads, femoral condyles, and tibial plateaus of patients Theoretical basis operated for knee, hip or ankle replacement (10 individuals; Real-time quantitative PCR monitors the degradation of mean age 68.4, range 55–87) in the University Hospital of a sequence-specific, dual-labeled fluorescent probe after Basel. The diagnosis of OA was based on clinical and each cycle of PCR amplification. During the extension radiographic evaluations, according to standard criteria12 . phase, the 5′-exonuclease activity of Taq DNA polymerase Control (CTR) cartilage samples (seven individuals; mean cleaves the probe, separating the 5′-reporter fluorescent age 59.3, range 25–85) were collected from the hip or ankle dye from the 3′-quencher fluorescent dye, resulting in an of patients with no history of joint disease, undergoing increase in the emission spectra of the reporter fluorescent joint replacement following femoral neck fracture or foot dye. After subtraction of the background fluorescence, amputation for tumor resection. calculated during the first 15 amplification cycles, the measured fluorescence is graphed as an amplification plot. Cartilage specimens Each reaction is characterized by a value, Ct, defined as the fractional number of cycles at which the reporter Non-calcified cartilage (i.e. superficial, middle, and deep fluorescent emission reaches a fixed threshold level in the layers above the tidemark)3. was dissected from the bone exponential region of the amplification plot. The Ct value is immediately after surgery. Pieces of cartilage were correlated to input target mRNA amount: a larger starting obtained from the weight-bearing area of the joints, which quantity of mRNA target results in a lower number of PCR was macroscopically degenerated in OA patients, and cycles required for the reporter fluorescent emission to processed for the different assays as follows. Samples reach the threshold, and therefore a lower Ct value. Thus, were fixed in 4% buffered formalin at 4°C for 24 h for the method is not based on the measurement of the total histological analysis, frozen in pre-weighed vials for amount of amplified product after a fixed number of cycles, biochemical assessment, or homogenized in Trizol (Life as in conventional PCR, and does not require post-PCR Technologies, Basel, Switzerland) for mRNA quantitation. processing of the product11. Some samples were also fixed in acetone at 4°C for 10 min, or digested with 0.15% collagenase for 22 h to isolate primary chondrocytes. Primers and probes Primers and probes for human GAPDH, collagen types I, Osteoblast isolation II, and X, aggrecan, versican, osteopontin and osteocalcin were designed with the assistance of the Primer Express Bone chips from the iliac crest of patients undergoing computer program (Perkin-Elmer Applied Biosystems, spinal fusion were cultured in Petri dishes in medium Foster City, CA), in order to display minimal internal struc- supplemented with 10% fetal bovine serum, ascorbic acid, ture (i.e. primer–dimer formation) and similar melting tem- dexamethasone