The aim of this study was to identify changes in cartilage intermediate layer protein/nucleotide pyrophosphohydrolase (CILP/NTPPH) expression in articular cartilage during aging. Adult (3-4 years old) and young (7-10 days old) porcine articular hyaline cartilage and fibrocartilage were studied by Northern blot analysis, in situ hybridization, and immunohistochemistry using a complementary DNA (cDNA) probe encoding porcine CILP/NTPPH and antibody to a synthetic peptide corresponding to a CILP/NTPPH sequence. Northern blot analysis of chondrocytes showed lower expression of CILP/NTPPH messenger RNA (mRNA) in young cartilage than in adult cartilage. In adult cartilage, extracellular matrix from the surface to the middeep zone was immunoreactive for CILP/NTPPH, especially in the pericellular matrix surrounding the middeep zone chondrocytes. In young cartilage, chondrocytes were moderately immunoreactive for CILP/NTPPH throughout all zones except the calcified zone. The matrix of young cartilage was negative except in the superficial zone. In young cartilage, CILP/NTPPH mRNA expression was undetectable. In adult cartilage, chondrocytes showed strong mRNA expression for CILP/NTPPH throughout middeep zones. Protein and mRNA signals were not detectable below the tidemark. CILP/NTPPH secretion into matrix around chondrocytes increases with aging. In this extracellular site it may generate inorganic pyrophosphate and contribute to age-related calcium pyrophosphate dihydrate crystal deposition disease.
ADVANCED age is the most common risk factor for the development of calcium pyrophosphate dihydrate (CPPD) crystal deposition disease. CPPD crystal deposition is associated clinically and epidemiologically with osteoarthritis of large weight-bearing joints.(1) Thus, the complications associated with CPPD crystal deposition will increase as our population ages. Although the mechanisms involved in CPPD crystal formation remain unclear, strong clinical and experimental evidence supports our hypothesis that excessive accumulation of extracellular inorganic pyrophosphate (ePPi) favors CPPD crystal deposition.(2–5) Hyaline and fibrocartilage are the primary intra-articular sources of ePPi.(6, 7) Excess activity of the ePPi-generating ectoenzyme nucleotide pyrophosphohydrolase (NTPPH) has been shown in joint fluids(8) and cartilage extracts(9) from patients affected with CPPD deposition disease and may contribute to overproduction of ePPi.
We have identified a unique sedimentable, articular cartilage vesicle-associated, 127-kDa chondrocyte NTPPH in porcine articular cartilage organ culture conditioned media(10) and cloned a soluble 61-kDa catalytically active carboxyl domain of 127-kDa NTPPH.(11) Both immunological analysis using antibodies raised against a synthetic peptide of N-terminal sequence from 61-kDa NTPPH (anti-SNTP) and whole 127-kDa protein, and molecular analysis of messenger RNA (mRNA) and complementary DNA (cDNA) encoding porcine and human 127-kDa NTPPH confirmed that the 61-kDa NTPPH represents a soluble proteolytic fragment of the 127-kDa NTPPH. Clinically, anti-SNTP or anti-127-kDa antibodies cross-reacted with 61-kDa and 127-kDa protein fragments in human joint fluids.(12) We hypothesized that the 127-kDa NTPPH is a secreted protein and cysteine-rich 61-kDa carboxyl domain becomes active after consequent proteinase cleavage.(13) Recently, a human homologue of 127-kDa NTPPH has been identified as cartilage intermediate layer protein (CILP).(14, 15) mRNA encoding full-length CILP and the human homologue of porcine 127-kDa NTPPH are identical. Both proteins are likely cleaved at a furinlike protease cleavage site when secreted.(14) Henceforth, we refer to a full-length molecule as CILP/NTPPH. CILP/NTPPH is a gene product distinct from plasma cell membrane glycoprotein 1 (PC-1/NTPPH), which also expresses NTPPH enzyme activity in various tissues(16) and is a member of the phosphodiesterase nucleotide pyrophosphatase family.(17)
Unlike the broadly distributed PC-1/NTPPH, CILP/NTPPH expression localizes to articular tissues including hyaline cartilage, fibrocartilage, tendon, ligament,(18) and synovial membrane.(12) Also, these are the only tissues that elaborate ePPi in organ culture.(7) Among the CILP/NTPPH-expressing tissues, hyaline and fibrocartilage alone spontaneously generate quantifiable ePPi in culture without growth factor stimulation. Articular cartilages from aged donors generate more ePPi in response to transforming growth factor β (TGF-β) than do cartilages from young donors.(19, 20)
We sought to determine whether CILP/NTPPH expression increases with age in parallel with the age-dependent increase in ePPi generation in vitro and the age-dependent increased prevalence of CPPD deposition disease. We describe localization of CILP/NTPPH expression in articular tissues and report increased expression during aging.
MATERIALS AND METHODS
Porcine articular tissue preparation
Normal hyaline cartilage and fibrocartilage (meniscus) were removed from the knee joints of adult (3-4 years) and young (7-10 days) pigs. In addition, young porcine caudal bone including intervertebral disc, vertebral growth plate, and attached skeletal muscle were examined. Longitudinal 2-mm slices of articular cartilage, meniscus, and caudal bone were fixed in 4% paraformaldehyde in 0.01 M phosphate-buffered saline (PBS, pH 7.4) overnight at 4°C. After washing three times in PBS, these tissues were dehydrated in graded concentrations of ethanol and embedded in paraffin. Consecutive sections were cut at 6 μm and then processed for immunohistochemistry and in situ hybridization.
After deparaffinization by heat in xylene and rinsing in ethanol, sections were treated with 1% hydrogen peroxide in methanol for 30 minutes to minimize endogenous peroxidase activity in the tissue and then washed in PBS. To avoid masking of epitopes, sections were predigested with 500 U/ml testicular hyaluronidase (type V; Sigma Chemical Co., St. Louis, MO, USA) in PBS for 20 minutes at 37°C and rinsed in PBS. The sections were covered with 5% normal goat serum in PBS for 30 minutes. Excess serum was removed by blotting and sections were covered with primary antibody solution and incubated overnight at 4°C. Affinity purified polyclonal antibody to synthetic peptide corresponding to the N-terminal sequence of the 61-kDa domain of CILP/NTPPH (anti-SNTP) was used at a 1:500 dilution. Nonimmune rabbit immunoglobulin G (IgG) was used as a negative control. The immunoreaction was performed using the Vectastain peroxidase rabbit ABC kit (Vector, Burlingame, CA, USA). Sections were washed in PBS and the antigenic sites were established by the sections' reacting with a mixture of 0.05% 3,3′-diaminobenzidine tetrahydrochloride (Dojin Chemical, Tokyo, Japan) in 0.05 M Tris-HCl buffer, pH 7.6, and 0.01% H2O2 for 7 minutes. Nuclei were stained with methyl green and the sections were dehydrated in ethanol, cleared in xylene, and then mounted in Permount (Fisher Scientific, Fair Lawn, NJ, USA). Consecutive sections also were stained with hematoxylin and eosin and safranin-O for histological characterization.
In situ hybridization
The hybridization procedures used in this study were essentially the same as those described elsewhere.(21, 22) An 880-base pair (bp) cDNA fragment from clone im202 (2.5 kilobases [kb] partial cDNA encoding porcine 127-kDa CILP/NTPPH)(11) was used as an in situ hybridization probe. Deparaffinized sections mounted on microscope slides were treated with 0.25 mg/ml pronase E type XXV (Sigma Chemical Co.) in 50 mM Tris-HCl, pH 7.6, and 5 mM EDTA for 10 minutes at room temperature and then acetylated with freshly diluted acetic anhydride (0.25% in 0.1 M triethanolamine buffer, pH 8.0) for 10 minutes. The slides were washed twice in double-strength SSC (0.15 M NaCl, 0.015 M trisodium citrate, pH 7.0) for 5 minutes each, dehydrated in ethanol, and air-dried. The treated sections were processed for hybridization in a mixture containing cDNA probe (1 μg/ml), yeast transfer RNA (tRNA) (500 μg/ml), 50% formamide, 10 mM Tris-HCl (pH 7.0), 0.15 M NaCl, 1 mM EDTA, 1× Denhardt's mixture, and 10% dextran sulfate. The cDNA was labeled with [35S]thymidine 5′-[α-thio] triphosphate (TTP) (Dupont, Boston, MA, USA) by nick-translation to a specific activity of 1.8 × 108 cpm/μg DNA. After hybridization, slides were washed under conditions of high stringency as previously described.(21) The dried tissue sections were dipped into Kodak NTB-2 emulsion (Eastman Kodak, Rochester, NY, USA) and exposed for 3-7 days at 4°C. The sections were counterstained with hematoxylin for image enhancement. As a negative control, parallel sections also were digested before hybridization with 2 mg/ml of RNAse for 1 h at room temperature.
Normal hyaline articular cartilage was removed from the distal femoral surface of knee joints from adult and young pigs. Both articular cartilages were full thickness (superficial through deep). Chondrocytes were isolated by sequential digestion as described,(23) plated at high density, maintained as primary monolayer culture in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin + streptomycin + fungizone (PSF; Life Technologies, Grand Island, NY, USA), and used within 3-10 days after plating so that cells recovered from digestion and maintained their phenotype. Total RNA was isolated from chondrocytes using a standard nonionic detergent method in the presence of 1000 U/ml of human placental RNAase inhibitor (Amersham, Arlington Heights, IL, USA).(24) RNA was quantitated by spectrophotometric absorbance at 260 nm. Ten micrograms of total RNA was electrophoresed in formaldehyde/agarose gels,(24) vacuum-transferred (VacuGeneXL; Pharmacia Biotech, Piscataway, NJ, USA) to a nylon membrane (Micron Separations, Inc., Westborough, MA, USA), and UV cross-linked. The membrane was prehybridized for 2 h at 42°C in 50% formamide, 5× saline-sodium phosphate-EDTA buffer (SSPE),(24) 5× Denhardt's reagent, 0.1% sodium dodecyl sulfate (SDS), and 100 μg/ml denatured salmon sperm DNA, and then hybridized overnight at 42°C in 50% formamide, 5× SSPE, 1× Denhardt's, 0.1% SDS, and 20 μg/ml denatured sheared salmon sperm DNA, with a [32P]-labeled cDNA probe containing an internal portion of the 880-bp fragment encoding the CILP/NTPPH. The DNA probe was radiolabeled by32P using a random-primer labeling kit (Ready-To-Go DNA labeling Beads; Pharmacia).(25) After hybridization with the probe, membrane was washed under high stringency (0.25× SSC and 0.1% SDS, at 65°C) and subjected to autoradiography. As a control, chicken β-actin cDNA was hybridized to the same membrane after removal of the CILP/NTPPH cDNA probe. Relative CILP/NTPPH mRNA expression was determined as percent β-actin mRNA expression. Densitometric measurement was carried out by scan analysis software Image Quant, version 5.0 (Molecular Dynamics, Sunnyvale, CA, USA).
Adult porcine articular hyaline and fibrocartilage
In adult cartilage, CILP/NTPPH was intensely immunostained in chondrocytes and cartilage extracellular matrix from superficial to deep zones (Fig. 1B). In addition to the cytoplasmic immunoreactivity, intense immunostaining for CILP/NTPPH was observed in the pericellular cartilage matrix (Fig. 1G). CILP/NTPPH mRNA was observed throughout mid- to deep-zone chondrocytes (Figs. 1D and 1H). Signals for CILP/NTPPH mRNA were not detectable in the cells beneath the calcified tidemark (data not shown).
In meniscus, CILP/NTPPH immunostaining was strong in fibrocartilage matrix but weak intracellularly (Fig. 1J). As in adult hyaline cartilage chondrocytes, strong expression of CILP/NTPPH mRNA was noted in fibrocartilage cells (Fig. 1K).
Young porcine caudal bone and intervertebral disc
Strong CILP/NTPPH immunoreactivity was observed in perichondrium, periosteum, and skeletal muscle (Fig. 2C). In the intervertebral disc, CILP/NTPPH protein stained weakly in fibrocartilage cells and matrix (Fig. 2E); however, the signal for CILP/NTPPH mRNA was very high in these cells. In the surface zone of vertebral cartilage, the CILP/NTPPH immunoreactivity was strong both within chondrocytes and in cartilage matrix (Fig. 2H). However, expression of CILP/NTPPH mRNA was undetectable (Fig. 2I). In the proliferating zone of the caudal vertebral cartilage growth plate, CILP/NTPPH was moderately stained. Only faint staining was seen in hypertrophic chondrocytes (Fig. 2K). No immunoreactivity for CILP/NTPPH was seen in cartilage matrix (Fig. 2K). In addition, no definite signal for CILP/NTPPH mRNA was seen within these chondrocytes (Fig. 2L).
Young porcine meniscus and articular cartilage
In meniscus, CILP/NTPPH immunoreactivity was strong in fibrocartilage matrix and weak within fibrocartilage cells (Fig. 3B). Very low signal for CILP/NTPPH mRNA was detected in the outer layer (Fig. 3C). In articular cartilage, CILP/NTPPH staining was strong in the surface layer of cartilage matrix and weak in resting chondrocytes (Fig. 3E). A moderate localized signal for CILP/NTPPH mRNA was identified only in surface chondrocytes (Fig. 3F).
Summarized histological data are shown in Table 1.
Table Table 1.. Summary of the CILP/NTPPH Expression in Various Articular Tissues
Northern blot analysis
Northern blot analysis of RNA extracted from cultured adult and young articular chondrocytes using a cDNA probe encoding CILP/NTPPH confirmed about 50% lower CILP/NTPPH mRNA expression in young chondrocytes than adult (Fig. 4).
Previous pathological studies described the smallest and presumably the earliest CPPD crystals at the lacunar margin of chondrocytes in the midzone of articular hyaline cartilage.(26) Because PPi is the anionic component of CPPD crystals, it is logical to hypothesize that the highest concentrations of ePPi produced from chondrocytes occur in the pericellular area where CPPD crystals form. The mechanisms by which chondrocytes elaborate ePPi either in normal or in diseased cartilage remain unclear. However, previous studies indicated that one of the pathways for ePPi synthesis involves the ePPi-generating ectoenzyme activity NTPPH.(1, 13) Here, we report the localization of one type of ecto-NTPPH protein and mRNA in articular tissues using nucleic acid and protein probes that recognize CILP/NTPPH. Another protein possessing NTPPH enzyme activity, PC-1/NTPPH, accounts for more than half of the cell-associated NTPPH activity of human articular chondrocytes(27) but does not have tissue specificity.(16, 17, 28) CILP/NTPPH is relatively specific to the articular tissues(18) that are able to generate ePPi.(7) The current study shows strong immunoreactivity to CILP/NTPPH antibody in cells and matrix of adult hyaline articular cartilage, adult meniscus, young meniscus, young perichondrium, and young periosteum. mRNA encoding CILP/NTPPH was present in adult chondrocytes, fibrocartilage chondrocytes (adult meniscus and young intervertebral disc), and terminally differentiated chondrocytes in young cartilage. Physiologically, ePPi is a potent inhibitor of apatite crystal nucleation and growth.(29) We speculate that these noncalcified articular and periarticular tissues may require ePPi to prevent mineralization, which would alter their physiological functions. In diseased tissues, excess production of ePPi might account for the pathological deposition of CPPD crystals. Further pathological analysis is ongoing using diseased human articular tissues.
There were discrepancies between mRNA and protein expression in some tissues. In young intervertebral disc, protein expression in cells was weak despite strong localized mRNA expression. This finding was striking because most other cells in young tissues failed to express mRNA encoding CILP/NTPPH. Failure to accumulate immunoreactive protein in this area may be caused by the rapid turnover of protein, alternative splicing of mRNA, or repressed translation. Further study is needed whether developmental or aging-related change in intervertebral disc correlate to the expression of CILP/NTPPH.
Young cartilage, except in surface layer articular chondrocytes, failed to show localized mRNA expression despite moderate protein expression. Low-abundance mRNA may be producing a stable protein accumulating in sufficient quantities to be immunostained by specific antibody.
In normal adult porcine cartilage, we observed strong mRNA expression for CILP/NTPPH within chondrocytes in middeep zone and strong immunoreactivity for the CILP/NTPPH protein within and surrounding these cells. The CILP/NTPPH produced in these adult chondrocytes appeared to be secreted from chondrocytes into the pericellular area. The strong pericellular immunolocalization of this CILP/NTPPH may be explained by the elaboration of vesicles into the pericellular area. Articular cartilage vesicles enriched in CILP/NTPPH have been described and characterized.(10) These vesicles form CPPD crystals in the presence of exogenous adenosine triphosphate (ATP) substrate for NTPPH.(30) In contrast, the absence of extracellular CILP/NTPPH around immature chondrocytes may reflect their inability to express and produce CILP/NTPPH or inability to secrete CILP/NTPPH into matrix.
Most of our findings agree with the prominent immunostaining of CILP localized in the midzone as described by Lorenzo et al.(15) However, their antibody stained more in the interterritorial matrix compared with our strong signal in pericellular matrix of adult cartilage. They also observed less prominent staining in the superficial zone. Antibody used for their immunohistochemical study was raised against purified CILP from human cartilage and recognized a 91.5-kDa protein on Western blot. Our antibody was raised against synthetic peptides encoding N-terminal 61-kDa porcine NTPPH downstream of the furinlike protease cleavage site. It recognizes 61-kDa and 127-kDa molecules in human and porcine cartilage extract, in vesicles,(10) and in human joint fluids,(12) and a 100-kDa molecule in human serum.(31) The epitope-specific immunolocalization will be solicited to study the cell trafficking and how this molecule is processed in the extracellular matrix.
Despite low abundance of CILP/NTPPH in young cartilage and growth plate matrix vesicles, there is ample NTPPH enzyme activity in these sites.(32) PC-1/NTPPH is the predominant NTPPH involved in mineralization by osteoblasts(33) and may play an important role in growth plate mineralization. Our preliminary Western blot analysis suggests that the dominant NTPPH enzyme in young chondrocytes is PC-1/NTPPH, not CILP/127-kDa NTPPH.(34)
Previous studies have shown that young cartilage and chondrocytes produce less ePPi than do adult cartilage and chondrocytes.(19) The weak mRNA expression of CILP/NTPPH in young chondrocytes by immunostaining, Northern blot, and in situ hybridization supports the notion that CILP/NTPPH may be relatively unimportant in ePPi production in young cartilage compared with PC-1/NTPPH. With aging, chondrocytes express relatively more CILP/NTPPH than PC-1/NTPPH. Accompanying this switch in ectoenzyme activity is an increased capacity to generate ePPi and increased sensitivity to stimulation of ePPi elaboration in response to TGF-β.
We are grateful to Ms. Michiko Nojima and the staffs in Department of Surgical Pathology, Kumamoto University Hospital for excellent technical assistance. Johnsonville Foods, Watertown, WI, kindly provided adult pig bones. This work was supported by grants from NIH NIAMS AR38656 (L.M.R.) and AR44862 (I.M.).