Age-Related Gene and Protein Expression in Mouse Mandibular Condyle Analyzed by Cap Analysis of Gene Expression and Immunohistochemistry Open Access

The mandibular condyle consists of the mandibular condylar cartilage (MCC), which covers its surface, and the subchondral bone, which forms the temporomandibular joint with the fibrous disc and the articular fossa of the temporal bone. The articular cartilage in most synovial joints, such as the knee joint, is hyaline cartilage, whereas the MCC is fibrocartilage. Unlike hyaline cartilage, which predominantly contains type II collagen, fibrocartilage is characterized by abundant type I collagen in addition to type II collagen and thus is more resistant to shear forces, which allows the temporomandibular joint to withstand large occlusal loads [1‒3]. In both cartilages, aggrecan is dominant among proteoglycans, the major extracellular matrix (ECM) molecules of cartilage, and can resist compressive forces [2, 4, 5].

The MCC is commonly divided into the following four zones from the surface to the subchondral bone according to the cell type: fibrous, proliferative, mature, and hypertrophic zones. The fibrous zone covers the surface of the mandibular condyle and contains flattened, fibroblast-like cells. The proliferative zone is composed of mesenchymal cells and serves as a cell reservoir to provide chondrocytes to the underlying zones. Although both the mature and hypertrophic zones consist of differentiated chondrocytes, the hypertrophic zone is distinguished from the mature zone by chondrocyte hypertrophy [2, 6].

Aging is an inevitable physiological process in all living animals and causes irreversible degenerative changes in the organs and tissues, including the mandibular condyle. Studies have reported the age-related morphological changes on the articular surface of the mandibular condyle, such as irregularity, clefts, and erosions, in humans and mice [7, 8]. Moreover, the histological changes, such as a reduction in cartilage thickness, a decrease in cellularity, and a lack of hypertrophic zone in the MCC have been reported in aged rats, mice, and humans [9‒11]. Various proteolytic enzymes, cytokines, and growth factors may be involved in chondrocyte differentiation and alteration of the cartilage ECM, resulting in these degenerative changes; however, the molecular mechanism has not yet been elucidated, and information on age-related factors for these degenerative changes in the MCC is limited.

The recent advent of next-generation sequencing technology has made it possible to sequence millions of DNA fragments of both known and unknown genes all at once. Cap analysis of gene expression (CAGE) is one of the methods for whole-transcriptome analysis using a next-generation sequencer and provides better quantification than RNA sequencing [12]. In this study, therefore, whole-transcriptome analysis of the mandibular condyle in young and aged mice was performed with CAGE to identify the differentially expressed age-related genes in the MCC, and immunohistochemical analysis was performed to localize the expression of the proteins encoded by the age-related genes identified with CAGE.

Ten male C57BL/6J mice aged 10 and 50 weeks were used. They were kept under specific pathogen-free conditions with a 12-h light-dark cycle at a temperature of 23°C ± 3°C. Stock diet and tap water were provided ad libitum.

The 10- and 50-week-old mice were euthanized by isoflurane inhalation (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan). Then, the mandibular condyle, which included the subchondral bone, the synovial membrane, and the MCC, was resected, homogenized with an ultrasonic homogenizer in QIAzol Lysis Reagent (QIAGEN, Hilden, Germany), and stored at −80°C until processed. After thawing the samples, total RNA was extracted using an RNeasy Lipid Tissue Mini Kit (QIAGEN) and treated with DNase using an RNase-free DNase set (QIAGEN).

The total RNA extracted from the mandibular condyle samples of 10- and 50-week-old mice (two biological replicates) was used for CAGE. The quality of the total RNA was examined, and CAGE was performed using DNAFORM (Yokohama, Japan), as described in our previous report [13]. Briefly, a CAGE library was prepared and sequenced using an Illumina NextSeq 500 System (Illumina, San Diego, CA, USA). The sequenced CAGE tags were mapped to the mouse mm9 genomes using BWA software (v0.5.9). Differential gene expression analysis was performed, and differentially expressed genes (DEGs) were identified as genes with a false discovery rate of less than 0.01.

Metascape, an online functional annotation tool (https://metascape.org), was used for gene ontology (GO) enrichment analysis, and upregulated and downregulated DEGs were significantly categorized into functional groups using the GO biological process database. p values of less than 0.01 were used to indicate statistical significance.

The 10- and 50-week-old mice were euthanized by isoflurane inhalation (FUJIFILM Wako Pure Chemical Corporation), and the mandibular condyles were resected and fixed in 4% paraformaldehyde in 0.1-m phosphate buffer (pH 7.4) at 4°C overnight. The fixed specimens obtained from 10- and 50-week-old mice were decalcified in 10% ethylenediamine tetra-acetic acid in 0.01-m phosphate buffer at 4°C for 1.5 and 2 months, respectively. After dehydrating through a graded series of ethanol solutions, the specimens were embedded in paraffin and sectioned coronally to a thickness of 5 μm. Serial sections, including the central part of the mandibular condyle, were selected, some of which were stained with hematoxylin and eosin.

The sections close to those stained with hematoxylin and eosin were deparaffinized and processed for immunohistochemistry of matrix metalloproteinases (MMPs)-3, -9, and -13, aggrecan, type I collagen, and type II collagen, as follows. The sections for the detection of aggrecan, type I collagen, and type II collagen were pretreated with 2.5% hyaluronidase from bovine testes (Sigma, St. Louis, MO, USA) for 30 min at 37°C. The sections for the detection of all target proteins were immersed in 3% hydrogen peroxide in methanol solution for 15 min to inactivate endogenous peroxidase, treated with 5% normal goat serum (FUJIFILM Wako Pure Chemical Corporation) for 30 min to block non-specific binding, and incubated with primary antibodies against MMP-3 (1:100; ab52915; Abcam, Cambridge, UK), MMP-9 (1:1,000; ab38898; Abcam), MMP-13 (1:400; ab39012; Abcam), aggrecan (1:1,500; 13880-1-AP; Proteintech, Rosemont, IL, USA), type I collagen (1:2,000; ab21286; Abcam), or type II collagen (1:200; ab34712; Abcam) for 2 h at room temperature. After washing with phosphate-buffered saline (PBS), the sections were incubated with Histofine Simple Stain MAX PO (R) kit (Nichirei Co., Tokyo, Japan) for 30 min at room temperature. Negative control sections were treated identically, except that the primary antibodies were omitted. Immunoreactivity was visualized with 3,3′-diaminobenzidine (DAB) solution, and the sections were counterstained with 1% methyl green.

Immunohistochemically stained sections were observed with a Leica DM2500 LED microscope (Leica Microsystems GmbH, Wetzlar, Germany) at the same magnification, and TIFF images were captured with a Leica MC170 HD digital camera under the same conditions. Fiji ImageJ software (an image-processing package distribution of ImageJ that bundles plugins that facilitate scientific image analysis; https://imagej.net/imagej-wiki-static/Fiji) was used for semi-quantitative image analysis. Brown DAB-stained pixels were isolated from background tissue, the color was inverted, and regions of interest (ROIs) were manually set in the medial, middle, and lateral parts of the MCC, respectively. The mean pixel value of each ROI, which was the sum of pixel values (0–255) divided by the number of pixels, was calculated as the intensity of immunoreactivity. In addition, the mean of the pixel values of these three ROIs was calculated and taken as the intensity of immunoreactivity for the whole MCC. IBM SPSS Statistics for Windows, version 22.0 (IBM Corp., Armonk, NY, USA), was used to perform statistical analysis. The Shapiro-Wilk normality test was used to examine the data, and Student’s t test was performed to compare the intensity of immunoreactivity between the young and the aged MCCs (n = 5). p values of <0.05 were considered statistically significant.

The sections were also used for Alcian blue staining for detection of acidic glycosaminoglycans (GAGs). The GAGs were stained with 1% Alcian blue 8GX (Merk KGaA, Darmstadt, Germany) in 3% acetic acid solution at pH 2.5 for 45 min at room temperature. After washing in running tap water for 2 min, the sections were counterstained with 0.1% nuclear fast red (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan).

Differential gene expression analysis followed by CAGE identified 75 upregulated DEGs and 179 downregulated DEGs with aging. GO enrichment analysis was performed to assign these DEGs to the biological process GO terms, one of the ontology functional categories. The results showed that the most significantly enriched GO term in the biological processes was “extracellular matrix organization” (Fig. 1). Twenty-two DEGs were assigned to the “extracellular matrix organization,” including genes for the major ECM proteins of the MCC, such as Acan, Col1a1, Col1a2, and Col2a1, and genes for their proteinases, such as Mmp3, Mmp9, and Mmp13 (Fig. 1; Table 1). In the aged mandibular condyle, Mmp3 was upregulated, whereas Mmp9, Mmp13, Acan, Col1a1, Col1a2, and Col2a1 were downregulated (Table 1).

Four zones were distinguished – fibrous, proliferative, mature, and hypertrophic zones – in the MCC of 10-week-old mice, whereas these four zones could not be identified and the cellularity was decreased in the MCC of 50-week-old mice (Fig. 2a–d). In the aged MCC, the fibrous and hypertrophic zones disappeared, and two zones with indistinct borders were found: the superficial zone containing cells similar to those in the proliferative zone of the young MCC and the deep zone consisting of chondrocytes but without hypertrophic chondrocytes (Fig. 2b, d). Mature and hypertrophic chondrocytes in the young MCC were arranged in regular columns, whereas in the aged MCC, chondrocytes were irregularly arranged (Fig. 2a–d). In the aged MCC, chondrocytes with a small nucleus and light cytoplasm were observed on the subchondral bone side of the deep zone (Fig. 2d). Additionally, consistent with the findings of previous studies [7, 8], some clefts and detachment were occasionally seen on the surface of the MCC of 50-week-old mice only (Fig. 2e, f).

Immunoreactivity for aggrecan and type I collagen in the MCC of 10-week-old mice was observed throughout the cartilage ECM in the fibrous, proliferative, and mature zones, and the ECM in the hypertrophic zone showed weak immunoreactivity (Fig. 3a, d). In the MCC of 50-week-old mice, immunoreactivity for aggrecan and type I collagen was seen entirely in the cartilage ECM of the superficial zone and sparsely in the ECM of the deep zone (Fig. 3b, e). Immunoreactivity for type II collagen in the MCC of young mice was predominantly localized to the ECM in the mature zone and was faint in the other zones, whereas in the MCC of aged mice, type II collagen immunoreactivity was observed across the ECM in the superficial zone and partially in the deep zone (Fig. 3g, h). Interestingly, chondrocytes with a small nucleus and light cytoplasm in the deep zone of the aged MCC showed intense immunoreactivity for type I collagen but not for aggrecan and type II collagen (Fig. 3c, f, i).

The semi-quantitative immunohistochemistry results are shown in Figure 4. In the whole MCC (p = 0.014) and in the middle (p = 0.005) and medial (p = 0.012) parts of the MCC, the intensity of aggrecan immunoreactivity was significantly lower in 50-week-old mice than in 10-week-old mice. In the lateral part of the MCC, the intensity of aggrecan immunoreactivity tended to be lower in the aged MCC (p = 0.105; not significant). A significant difference in the immunoreactivity intensity was found only in the middle part of the MCC for type I collagen (p = 0.011) and only in the lateral part of the MCC for type II collagen (p = 0.030). In the whole MCC, there was no significant difference in the intensity of immunoreactivity for type I and II collagens, whereas the immunoreactivity intensity tended to be lower for type I collagen (p = 0.089) and slightly higher for type II collagen (p = 0.275) in 50-week-old mice than in 10-week-old mice.

Alcian blue histochemistry was used to examine the distribution of sulfated GAGs. The cartilage matrix throughout all four zones was positive for Alcian blue in the young MCC (Fig. 5a, b). In contrast, in the aged MCC, the entire cartilage ECM of the superficial zone was positive for Alcian blue, whereas the Alcian blue-positive ECM was markedly decreased in the deep zone (Fig. 5c, d). The finding of decreased GAGs in the deep zone, which occupies the majority of the aged MCC, led to exploration of the genes encoding GAG synthetic enzymes among the DEGs identified by CAGE data analysis, which revealed that Chst3 was significantly downregulated in the aged MCCs.

Immunoreactivity for MMP-3 was observed in cells in the fibrous, proliferative, mature, and hypertrophic zones of the young MCC (Fig. 6a, c). In the aged MCC, MMP-3 immunoreactivity was particularly localized to both the lateral parts of the mandibular condyle, where it was found in cells in the superficial zone and cells near the superficial zone in the deep zone (Fig. 6b, d). Additionally, the cells in the synovial membrane around the condylar neck were immunoreactive for MMP-3 in both 10- and 50-week-old mice (Fig. 6b–d). Immunoreactivity for MMP-9 was identified only in large multinucleated cells in the subchondral bone at the interface with the MCC of 10-week-old mice, whereas no MMP-9 immunoreactivity was detected in the MCC of 50-week-old mice (Fig. 6e, f). Most cells in the fibrous, proliferative, mature, and hypertrophic zones of the young MCC were immunoreactive for MMP-13, and among them, the cells around the surface of the mandibular condyle and hypertrophic chondrocytes tended to show strong immunoreactivity (Fig. 6g). In the aged MCC, MMP-13 immunoreactivity was observed in cells in the superficial zone and cells near the superficial zone in the deep zone (Fig. 6h).

We demonstrated that the genes related to the “extracellular matrix organization,” including Mmp3, Mmp9, and Mmp13, are most differentially expressed in the mandibular condyle with aging. MMPs, endopeptidases containing a catalytic zinc ion in their active site, are the main enzymes for ECM degradation. So far, 23 MMPs have been identified in humans and are divided into six groups based on their domain structure and substrate specificity: collagenases, gelatinases, stromelysins, matrilysins, membrane-type MMPs, and nonclassified MMPs [14‒16]. Our whole-transcriptome analysis indicated that MMP-3 was the only proteolytic enzyme whose gene expression increased with age. MMP-3 is classified as a stromelysin and can degrade aggrecan and type II collagen, the major components of the cartilage ECM [15, 16]. In this study, the protein expression of MMP-3 was prominent in the synovial membrane in both the young and aged. Studies have shown that MMP-3 is highly expressed in synovial cells in osteoarthritis (OA) and rheumatoid arthritis (RA) with articular cartilage destruction and that MMP-3 levels are elevated in the synovial fluid of patients with OA and RA [17‒20]. The low-grade chronic systemic inflammation, referred to as inflammaging, occurs during physiological aging [21‒23]. In the aged mandibular condyle as well as in the joints of patients with OA and RA, the release of proinflammatory cytokines is increased due to inflammaging and may promote the secretion of MMP-3, which might then facilitate the degradation of the articular cartilage. The gene expression of MMP-3 has been also shown to be elevated in other aged connective tissues, such as human skin dermis, human gingiva, and mouse periodontal tissues [13, 24, 25], which supports the idea that MMP-3 is a crucial age-related factor. In our immunohistochemical investigation, MMP-3 immunoreactivity was detected in the medial and lateral parts of the aged MCC, not in the middle part. Both sliding and rotating movements occur in the temporomandibular joint, resulting in mechanical stresses on the surface of the articular cartilage: compressive, tensile, and shear stresses. The MCC particularly has been reported to have weak resistance to medio-lateral shear stress [2]. Additionally, mechanical forces have been shown to induce upregulation of the MMP-3 gene and protein expression levels in the MCC [26]. Consequently, we presumed that shear stress applied to the aged MCC would elevate MMP-3 expression in the medial and lateral parts of the MCC, which are less resistant to such stress.

In this study, the gene expression of MMP-13 in the mandibular condyle, in contrast, decreased with age. MMP-13 is a collagenase that can degrade a wide range of ECM proteins, including type I collagen, type II collagen, and aggrecan [16, 27, 28]. The production and activity of MMP-13 are known to enhance in pathological conditions, such as cancer progression, metastasis, and OA [16, 28, 29]. Studies have shown that the expression of MMP-13 is increased in human OA cartilage and in tibial OA model mice [30‒32] and that Mmp13 knockout mice with induced OA exhibit less cartilage destruction in the tibia than wild-type mice [33]. In our study, the gene expression of MMP-3 was upregulated with aging, consistent with previous reports of its expression in OA; however, the gene expression of MMP-13 was downregulated, conversely to previous reports in OA. One of the possible reasons for this is that our study focused on the expression of MMPs in physiological aging, not pathological aging. Presumably, increased MMP-13 expression may reflect pathological conditions. Thus, MMP-13 could be a potent marker of pathological cartilage destruction.

MMP-13 is a known marker for hypertrophic chondrocytes, and consistent with this, our immunohistochemistry showed MMP-13 immunoreactivity in hypertrophic chondrocytes of MCCs in young mice. During embryonic development, the growth plate cartilage of Mmp13-null mice increases relative to that of wild-type mice due to an increase in the thickness of the hypertrophic cartilage layer, demonstrating that MMP-13 is essential for normal endochondral ossification [34]. These findings suggest that MMP-13 is involved in apoptosis of hypertrophic chondrocytes during endochondral ossification. In this study, the hypertrophic zone was observed in the young MCC but not in the aged MCC; instead, chondroid bone-like tissue-containing chondrocytes with a small nucleus that strongly expressed type I collagen were found in the aged MCC, which is discussed below. As in endochondral ossification of the growth plate, cartilage-to-bone transformation might occur in the mandibular condyle with aging, and MMP-13 could contribute to this process.

The gene expression of MMP-9 was also downregulated with aging in this study. MMP-9 is classified as a gelatinase and highly expressed in osteoclasts [15, 16, 35]. Since the sample for CAGE contained subchondral bone as well as MCC, it is understandable that MMP-9 gene expression in osteoclasts was detected. Indeed, our immunohistochemical investigation showed that MMP-9 was expressed in multinucleated bone-resorbing osteoclasts at the interface between MCC and subchondral bone in 10-week-old mice. In contrast, no MMP-9 immunoreactivity was observed in 50-week-old mice. This finding indicates that osteoclasts actively secrete MMP-9 to degrade the subchondral bone ECM for remodeling in the young, whereas ECM remodeling at the bone-cartilage interface tends to be less active with age, and osteoclast activity would be reduced in the aged. Interestingly, both Mmp9- and Mmp13-null mice exhibit the expanded hypertrophic zone of the growth plate cartilage in the developing long bone due to delayed apoptosis of hypertrophic chondrocytes [34, 36, 37]. Furthermore, this phenotype has been shown to be even more severe in Mmp9- and Mmp13-double-null mice [37]. Indeed, studies reported that the expression of MMP-13 was observed in hypertrophic chondrocytes of the growth plate cartilage in long bones and of the articular cartilage in the mandibular condyle [10, 37, 38]; additionally, in this study, the hypertrophic chondrocytes strongly expressed MMP-13. Furthermore, pro-MMP-9, a latent form, is activated by cleavage by MMP-13 [39], suggesting that these two MMPs are synergistically involved in ECM remodeling in the transition from cartilage to bone in the mandibular condyle.

In this study, differential gene expression analysis and semi-quantitative immunohistochemical analysis revealed a significant decrease in the aggrecan gene and protein expression levels in the aged MCC. The analyses also showed downregulated gene expression of type I collagen and a tendency toward a decreased type I collagen protein expression level in the aged MCC. A reduction in gene and protein levels of aggrecan and type I collagen may be associated with age-related inactivity of the cartilage ECM proteins’ synthesis. However, the type II collagen gene expression was downregulated, whereas, contrary to our prediction, its protein expression tended to increase, albeit only slightly, with aging. A previous study used an immunoblot technique to also demonstrate an increased type II collagen protein level in the porcine MCC with aging, which is consistent with our results [40]. Type II collagen in the articular cartilage is known to be one of the longest-lived proteins, with an estimated half-life of 117 years, and has a higher potential for cross-linking than type I collagen because it contains more amino acid residues for cross-linking, presumably resulting in a highly stable structure by binding strongly within and between fibrils [41]. Properties of type II collagen, such as longevity and stability through cross-linking, might have contributed to the increase in its protein expression level despite its decreased gene expression level with aging. Another possible reason for the increased type II collagen protein expression is the effect of MMP-13, which has a substrate preference for type II collagen over type I collagen [41]. MMP-13 expression decreases during aging, which may lead to reduced degradation of type II collagen, consequently increasing its protein expression level in the aged MCC. Previous studies have reported that the correlation between gene and protein expression levels gradually disassociates with aging and that this decoupling is mediated by several posttranscriptional regulators, including micro RNAs and RNA-binding proteins, and reduced proteostasis [42]. However, the mechanism underlying the inconsistency between gene and protein expression in aged tissues has not been elucidated yet and requires further investigation.

Aggrecan, a proteoglycan, is a major component of cartilage composed of a core protein and a large number of GAGs that bind to it. In human cartilage, aggrecan has two kinds of GAGs, chondroitin sulfate and keratan sulfate, whereas mouse and rat cartilage do not contain keratan sulfate in aggrecan [43]. Our CAGE data showed that Chst3, which encodes chondroitin 6-0-sulfotransferase 1 (C6ST-1), was significantly downregulated in the aged mandibular condyle. The enzyme C6ST-1 catalyzes the modification of chondroitin sulfate synthesis, and mutations in Chst3 systemically lacking sulfation of the chondroitin sulfate chain reportedly cause severe chondrodysplasia [44]. Alcian blue histochemical analysis can detect acidic GAGs such as chondroitin sulfate. In this study, Alcian blue-positive ECM was markedly reduced in the aged MCC, indicating a decrease in aggrecan chondroitin sulfate content in the MCC. One of the reasons for this may be the decrease in GAG synthesis due to downregulation of Chst3 expression.

In this study, we found chondrocytes with a small nucleus and light cytoplasm in the aged MCC. These cells were present on the near side of the subchondral bone in the deep zone and intensely expressed type I collagen, rather than aggrecan and type II collagen. A tissue, known as chondroid bone, has features in-between cartilage and bone, and the chondroid bone ECM contains both type I and type II collagens [45]. Mizoguchi et al. reported the presence of chondroid cells that more strongly express type I collagen than type II collagen in the MCC and suggested the possibility of the osteogenic trans-differentiation of chondrocytes [46]. Considering that they are present near the subchondral bone and have the appearance intermediate between osteocytes and chondrocytes and that they strongly express type I collagen, it is possible that the cells with a small nucleus and light cytoplasm observed in the aged MCC in this study might be chondroid cells. Further investigation is required to determine what these cells are.

Our study revealed that the genes related to the GO term “extracellular matrix organization” were most differentially expressed in the mandibular condyle during aging and identified the age-related genes, such as Acan, Col1a1, Col1a2, Col2a1, Mmp3, Mmp9, and Mmp13. Among these seven genes, only Mmp3 was upregulated, whereas the others were downregulated with aging. We localized the protein expression encoded by these age-related genes, and our immunohistochemical analysis showed the time- and site-specific localization of aggrecan, type I collagen, type II collagen, MMP-3, MMP-9, and MMP-13 in the MCC. In this study, we observed a significant decrease in protein expression of aggrecan and a reduction in GAGs in the aged MCCs. These results suggest that MMP-3, MMP-9, and MMP-13 contribute to the ECM remodeling of the MCC and the subchondral bone during aging through degradation of ECM proteins at specific times and sites under regulation of their production and secretion.

We thank Dr. Miyuki Mayanagi and Mr. Yasuto Mikami, Division of Craniofacial Development and Tissue Biology, Tohoku University Graduate School of Dentistry, for their great assistance.

All experimental procedures were reviewed and approved by the Institutional Laboratory Animal Care and Use Committee of Tohoku University (approval number 2019 DnA-001-07). This study followed the Animal Research: Reporting of In Vivo Experiments’ guidelines.

The authors have no conflicts of interest to declare.

This work was supported by JSPS KAKENHI (Grant No. JP18K17110, JP21K09825, and JP21K10061) and Tohoku University Center for Gender Equality Promotion Support Project.

M.-C.Y. performed the experiments and contributed to the analysis and interpretation of the data; Y.K. and Y.I. performed the experiments; M.N. conceived the idea, designed the study, and contributed to the analysis and interpretation of the data; M.-C.Y. and M.N. drafted the original manuscript; and Y.S. contributed to the interpretation of the data and revised the manuscript. All authors reviewed the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.