Integrin α2 β 1 plays an important role in the interaction between human articular cartilage‑derived chondrocytes and atelocollagen gel

Although atelocollagen gel is used as a scaffold for culturing human articular cartilage‑derived chondrocytes, little is known about cell–gel interactions. In this study, we investigated the mechanism via which atelocollagen gel affects human articular cartilage‑derived chondrocytes. Two types of three‑dimensional cultures of human articular cartilage‑derived chondrocytes (i.e., with and without atelocollagen gel) were compared. While the amount of atelocollagen gel in culture gradually decreased with time, it promoted the expression of matrix metalloproteinases (MMPs) during the early stages of culture. Genome‑wide differential gene expression analysis revealed that cell membrane‑ and extracellular matrix‑related genes were highly ranked among up‑ and down‑regulated groups in cells cultured in the presence of atelocollagen gel. Among the integrin family of genes, the expression of integrin subunit alpha 2 and integrin subunit alpha 10 was significantly increased in the presence of atelocollagen gel. Blocking α2β1 integrin with the specific inhibitor BTT 3033 had a significant effect on cell proliferation, MMP expression, and cell shape, as well as on the response to mechanical stimulation. Taken together, our findings indicate that the α2β1 integrin pathway plays an important role in the interaction of atelocollagen gel with human articular cartilage‑derived chondrocytes and may be a potential therapeutic target for articular regarding the interactions between atelocollagen gel and human articular cartilage-derived chondrocytes, and the mechanism via which atelocollagen gel affects these cells. Thus, further elucidation of the impact of these interactions may have therapeutic implications. To this end, the present study aimed to examine the interactions between human articular cartilage-derived chondrocytes and atelocollagen gel, and their impact on cell proliferation, gene expression, and on the response to mechanical stimulation in 3D cultures with and without atelocollagen. via integrin α2β1. These results, together with the results of the integrin α2β1 inhibition experiments, strongly suggest that ITGA2 is a key molecule that mediates the effect of atelocollagen gel on human articular cartilage-derived chondrocytes. ITGA10 , another collagen-binding integrin, showed a similar expression pattern as ITGA2 ITGA10 have important in endochondral ossification it ITGA2 functions MMP13 MMP1 and MT-MMP1. function entire MMP family, a homeostatic reaction living body. Elucidation of important insights area. previous studies using ITGA2 KO mice, new vessels a healing model 40,41 and degeneration suppressed in an osteoarthritis model 42 and alterations in the expression pattern of MMP family of genes confirmed. These phenotypes are consistent with the changes in the expression of MMP family members observed in the present study. The cellular response of articular cartilage to mechanical stress, in the form of CCL, has been vigorously studied 43,44 . In the present study, CCL was cytotoxic to human articular cartilage-derived chondrocytes in the AC(−) group and promoted PGE2 production. The involvement of the α2β1 integrin pathway in this cellular response was confirmed using the specific α2β1 integrin inhibitor, BTT 3033. In contrast, this cytotoxic reaction was not observed in the AC(+) group, and the production of PGE2 was also less pronounced in the AC(+) group than in the AC(−) group. Further studies on the potential role of integrin as a mechanoreceptor and the effect of mechanical stress on cultured cells are warranted. Regenerative medicine and cell therapy, such as ACI, are gaining importance in the treatment of many diseases 45 . Although further investigation is necessary, our present findings have implications in regenerative medicine for articular cartilage disorders. Since collagen is an ECM protein of major importance in many cell types and MMP family members function as key regulators of ECM under both physiological and pathologi-cal conditions, our findings likely have implications not only in articular cartilage disorders, but also in other disorders in which cell–ECM interactions play an important role.


Results
Collagen gel affects the metabolic activity, but not the proliferation, of human articular cartilage-derived chondrocytes in 3D culture, as well as the expression of cell cycle-related factors and chondrogenic differentiation markers. We first examined the distribution of atelocollagen gel in atelocollagen sponge immediately after cell seeding by staining with anti-collagen type I antibody. The continuous pores of the sponge were filled with gel in the group using atelocollagen gel [AC(+) group; Fig. 1A]. Following cell seeding, 3D culture was continued in growth medium and cells were found throughout the sponge on day 29 of culture in both the AC(+) group and the group that did not use atelocollagen gel [AC(−) group] (Fig. 1B). DNA content increased during the first 4 weeks of culture, and no significant differences were observed with or without atelocollagen gel (Fig. 1C). Moreover, the expression of cell cycle-related factors (CDK4, Cyclin D1, p21) was higher in the AC(+) group than in the AC(−) group on day 4 of culture ( Fig. 1D). Comparing 2D and 3D cultures of cell proliferation, the cell metabolic activity in the 2D culture was significantly higher during the week after cell seeding (Fig. 1E). Metabolic activity was significantly lower in the AC(+) group than in the AC(−) group in the early stages of culture, but significantly higher on days 22 and 29 of culture (Fig. 1F).
Expression of multiple MMPs increases at an early stage of culture, but gradually decreases with time, in the presence of atelocollagen gel. Atelocollagen is biodegradable. When the distribution of the gel in 3D culture was examined over time, the area occupied by atelocollagen gel gradually decreased with time ( Fig. 2A). When cells were cultured with atelocollagen gel, the brown signal (corresponding to atelocollagen gel) decreased as the duration of culture increased. On day 29 of culture, dark brown spots were observed, which indicated the contraction and degradation of the gel. No change was observed in the absence of cells. When hydroxyproline in the culture supernatant was quantified for the purpose of examining collagen degradation, a significantly larger amount of product was detected in the AC(+) group compared to the 2D culture and AC(−) groups (Fig. 2B). Next, we examined the expression of several matrix metalloproteinase (MMP) family members, which are candidate collagen-degrading enzymes. The expression of MMP2, MMP3, MMP9, and MMP13 was significantly higher in the AC(+) group than in the AC(−) group on day 4 of culture, but gradually decreased with culture time (Fig. 2C). Immunostaining of the samples on day 4 of culture showed a clear expression of MMP9 in the AC(+) group, while it showed an unclear expression in the AC (−) group (Fig. 2D). No other MMPs (MMP1, MMP2, and MMP13) showed a clear expression in the AC(+) and AC(−) groups (Fig. S1). The expression of chondrogenic differentiation markers (collagen 1, collagen 11, sox9, and proteoglycan 4) also differed between the AC(+) and AC(−) groups on day 4, but did not significantly differ on day 29 (Fig. 2E). The expression of chondrocyte hypertrophic markers and hypoxia-inducible factors also showed a significant difference between the AC(+) and AC(−) groups on day 4 (Fig. S2).
Atelocollagen gel regulates the expression of cell membrane-and extracellular matrix-related genes and promotes the expression of ITGA2 and ITGA10 in 3D cultures. Genome-wide differential gene expression analysis was performed to examine the effects of atelocollagen on cultured cells. Biological replicates of the same culture conditions were clustered together using principal component analysis and cluster analysis (Fig. 3A,B). There were substantial differences in the gene expression patterns between 2 and 3D cultures. In addition, the presence of atelocollagen gel significantly affected gene expression in 3D cultures (Fig. 3C,D). In the enrichment analysis using the highest rank (|log2 FC|> 2) of differentially expressed genes (DEGs) comparing 3D cultures of the AC(+) and AC(−) groups, extracellular matrix-related factors were categorized as the highest in each category ( www.nature.com/scientificreports/ Component) of the up-/down-regulated groups (Fig. 4A,B). Among the molecules that bind to collagen as part of the extracellular matrix, we focused on the integrin family. Several genes, including integrin subunit alpha 2 (ITGA2) and integrin subunit alpha 10 (ITGA10), were found to be highly expressed in the AC(+) group and to have a low expression in the AC(−) group (Fig. 4C). The measurement of the expression of four genes (ITGA1, ITGA2, ITGA10, and ITGA11) encoding collagen-binding integrins using qPCR revealed that the expression of ITGA2 and ITGA10 was higher in the AC(+) group than in the AC(−) group and 2D cultures (Fig. 4D).
BTT 3033, an α2β1 integrin-specific inhibitor, affects proliferation, MMP expression, and cell shape in the AC(+) group. The   www.nature.com/scientificreports/ culture and the AC(+) group, and there was no change in the AC(−) group. The change was particularly large in the AC(+) group, with MMP13 being significantly down-regulated in the presence of BTT3033, while MMP1 and MT-MMP1 were significantly up-regulated (Fig. 5D). As for the cell shape, cells in 3D culture were smaller than those in 2D culture (Fig. 6A,B) and cells in the AC(+) group cultured in the presence of BTT 3033 had a lower aspect ratio than cells cultured without BTT 3033 (Fig. 6A,C). α2β1 integrin affects the response to cyclic compressive loading stimulation in 3D culture. Cell-ECM interactions are thought to play a role in the response to mechanical stimulation 20 . 3D culture cells were subjected to cyclic compressive loading (CCL) stimulation using a bioreactor. First, we examined the concentration of lactate dehydrogenase (LDH) in the culture supernatant as a marker of cytotoxicity (Fig. 7A). In the AC(−) group, the concentration of LDH after CCL was significantly higher than that before CCL. This difference was also observed in the presence of BTT 3033. In the AC(+) group, however, the concentration of LDH did not increase in response to CCL, although a significant increase was observed in response to CCL in the presence of BTT 3033 (10 μM). We also examined the effect of CCL on the production of prostaglandin E2 (PGE2), a well-known pathogenic molecule related to OA development (Fig. 7B) 21 . In a previous report, we showed that CCL on a 3D culture of human synovial fibroblasts significantly upregulated PGE2 production 22 . In both the AC(+) and AC(−) groups, CCL significantly increased the production of PGE2. This effect was suppressed in a dose-dependent manner by BTT 3033, and a complete suppression was observed at a concentration of 10 μM. When comparing the AC(+) and AC(−) groups, the degree of increase in PGE2 production was more prominent in the AC(−) group, although the effect of BTT 3033 was similar in both groups. www.nature.com/scientificreports/

Discussion
In the current study, several lines of evidence suggest that the interactions between atelocollagen gel and human articular cartilage-derived chondrocytes play an important role in various cellular functions. First, genome-wide differential gene expression analysis revealed that atelocollagen gel regulates the expression of cell membraneand extracellular matrix-related genes, while the amount of atelocollagen gel in culture gradually decreased with time. Second, we found that the expression of ITGA2, which encodes alpha-2 integrin, was significantly increased in the presence of atelocollagen gel, and that inhibition of the α2β1 integrin pathway affected the cell shape and proliferation, the transcription of MMP family members, and the effect of mechanical stress on human articular cartilage-derived chondrocytes in 3D culture. Previous studies have shown that cells grow healthily when cultured in collagen gel 23,24 . Consistent with this, human articular cartilage-derived chondrocytes cultured in atelocollagen have been shown to proliferate and synthesize extracellular matrix in a cell density-dependent manner 25 . In the present study, we compared the growth of 3D cultured cells seeded onto atelocollagen sponges in the AC(+) and AC(−) groups, and found no significant difference in the total DNA content between the two groups over the course of a 4-week culture period. Previous studies using the same atelocollagen sponge have reported that embryonic stem cells showed good growth in 3D culture 26 , and similar findings have been reported for induced pluripotent stem cells 27 . While these previous studies used different cell types than the present study, the findings are consistent with ours in that atelocollagen gel does not interfere with cell proliferation. On the other hand, cell metabolic activity assessed using the Cell Counting Kit-8 appears to differ depending on the presence of atelocollagen gel 28 .
Atelocollagen is a naturally occurring collagen obtained by removing telopeptides from type 1 collagen and it is biodegradable 12 . Immunostaining and the measurement of collagen degradation products suggest that the atelocollagen gel and sponge are biodegradable. MMP family members are known to function as important elements of collagen degradation 12 . In the present study, the expression of MMP2, MMP3, MMP9, and MMP13 in human articular cartilage-derived chondrocytes increased in the AC(+) group. Among these MMPs, MMP13, which showed the greatest expression change, was classified as a collagenase and has a broad and strong proteolytic ability 29,30 . MMP13 plays a role in nascent bone growth, remodeling, and in the maintenance of articular www.nature.com/scientificreports/ cartilage homeostasis, and is also recognized as a major MMP overexpressed in osteoarthritis 31,32 . MMP9 was confirmed to have significant transcription promotion and an increased protein content in the AC(+) group. Similar to MMP13, MMP9 is also reported to have an increased expression in OA cartilage and to function as a cartilage-destroying factor. It is unclear which of these MMP families plays a central role, but it is likely that the microenvironment of cells differs significantly between early and late cultures using atelocollagen due to the enzymatic function of promoted MMPs. Chondrogenic markers were more highly expressed in cells of 3D cultures than in 2D culture, particularly in the early stages of culture, when atelocollagen gel is abundantly present. This result is consistent with a previous report showing that collagen gel contributes to both the differentiation and dedifferentiation of chondrocytes [33][34][35] . These findings collectively suggest that the presence of atelocollagen gel modulates gene expression in human articular cartilage-derived chondrocytes in 3D culture, leading to changes in the cell properties, such as in cell metabolic activity. Consistent with previous reports, we found that dimensionality had a major impact on cells, with differences being observed between 2 and 3D cultures [36][37][38] . Genome-wide gene expression analysis also revealed that the presence of atelocollagen gel promotes the transcription of genes related to cell membrane-ECM interactions, particularly those related to growth factor, cytokine, and MMP activity. Among these molecules, ITGA2, a gene involved in cell-ECM interactions, was highly expressed in human articular cartilage-derived chondrocytes in the AC(+) group. 3D culture of human skin fibroblasts with type I collagen gel has been shown to induce integrin-mediated MMP13 expression, and p38 MAPK (mitogen-activated protein kinase) activation is important in this process 39 . Similarly, in the present study, atelocollagen gel induced the expression of MMP13, likely  www.nature.com/scientificreports/ The cellular response of articular cartilage to mechanical stress, in the form of CCL, has been vigorously studied 43,44 . In the present study, CCL was cytotoxic to human articular cartilage-derived chondrocytes in the AC(−) group and promoted PGE2 production. The involvement of the α2β1 integrin pathway in this cellular response was confirmed using the specific α2β1 integrin inhibitor, BTT 3033. In contrast, this cytotoxic reaction was not observed in the AC(+) group, and the production of PGE2 was also less pronounced in the AC(+) group than in the AC(−) group. Further studies on the potential role of integrin as a mechanoreceptor and the effect of mechanical stress on cultured cells are warranted.
Regenerative medicine and cell therapy, such as ACI, are gaining importance in the treatment of many diseases 45 . Although further investigation is necessary, our present findings have implications in regenerative medicine for articular cartilage disorders. Since collagen is an ECM protein of major importance in many cell types and MMP family members function as key regulators of ECM under both physiological and pathological conditions, our findings likely have implications not only in articular cartilage disorders, but also in other disorders in which cell-ECM interactions play an important role.

Materials and methods
Ethics statement. This study was conducted using cells derived from human articular cartilage and was approved by the Osaka University Institutional Ethical Committee (approval ID 16085-4). Written informed consent was obtained from all subjects, and all methods were performed in accordance with the relevant guidelines and regulations.

Scaffolds.
Commercially available atelocollagen sponge (Mighty, KKN-CSM-50, Koken, Tokyo, Japan) and 1% atelocollagen gel (Koken) were used. Atelocollagen primarily consists of type I collagen derived from bovine dermis. The sponge has an interconnected pore size of 30-200 μm and can withstand a compressive loading of up to 40 kPa.

Isolation of human articular cartilage-derived chondrocytes and 2D culture.
Human knee articular cartilage was aseptically obtained from eight patients aged 58-83 years who underwent total knee arthroplasty to treat osteoarthritis. The cell isolation protocol was essentially the same as previously described 22 . Briefly, cartilage specimens were rinsed with phosphate-buffered saline (PBS), minced meticulously, and digested with 0.1% collagenase in Dulbecco Modified Eagle's Medium/high glucose (DMEM, D6429) (Sigma-Aldrich, St. Louis, MO, USA) with 1% penicillin/streptomycin (Thermo Fisher Scientific Cat# 15140122) for 6 h at 37 °C. Cells were cultured in DMEM with 10% fetal bovine serum (FBS, Lot# 174012) (Nichirei Bioscience, Tokyo, Japan) and 1% penicillin/streptomycin using commercial tissue culture polystyrene (TCPS) dishes or plates and used at passages two to four. 3D culture in atelocollagen sponges. Cells were seeded onto atelocollagen sponges using two methods.
For 3D culture with atelocollagen, cells (1-5 × 10 5 /sponge) were suspended in growth medium and mixed with an equal volume of 1% atelocollagen gel to produce a cell suspension in a 0.5% collagen solution. The solution was carefully dripped onto the sponge and allowed to soak. For the 3D culture without atelocollagen, cells sus-   RNA extraction and quantitative real-time RT-PCR. Total RNA was extracted using Trizol (Invitrogen) and a PureLink RNA Purification kit (Thermo Fisher Scientific), and reverse transcribed into cDNA using a High-Capacity RNA-to-cDNA kit (Thermo Fisher Scientific). Quantitative PCR was performed using Power SYBR Green Master Mix and QuantiStudio 7 Pro Real-Time PCR System (Thermo Fisher Scientific), according to the manufacturer's instructions. Nucleotide sequences of primers are described in Table S1. The expression of the target genes was normalized to that of the reference gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Measurement of hydroxyproline in cell culture supernatants.
We performed a colorimetric method using the QuickZyme kit (QuickZyme Biosciences, Leiden, the Netherlands) for hydroxyproline measurement.
In brief, the supernatant on day 3 of culture was harvested and hydrolyzed in 6 M HCl in a safety centrifuge tube for 16 h at 95 °C. Reconstituted samples were centrifuged for 10 min at 13,000×g, and the supernatants were used for the measurement according to the manufacturer's protocol.
Microarray analysis. Experiments were conducted in triplicates for all three culture conditions (2D culture, 3D culture with atelocollagen gel, and 3D culture without atelocollagen gel). Cells in the growth phase in 2D culture or cells on day 4 of 3D culture were harvested. The isolated total RNA was amplified and labeled with Cy3 using a Low Input Quick Amp Labeling Kit (Agilent Technologies). Hybridized arrays were scanned using an Agilent DNA Microarray Scanner (G2565CA). Microarray analysis was carried out using Takara Bio. Data analyses were carried out using the Subio platform version 1.24 (Subio, Inc., Kagoshima, Japan, http://www. subio .jp). Microarray datasets are available at the National Center for Biotechnology Information Gene Expression Omnibus (GEO) database (Accession no. GSE 154723).
Cell shape analysis. Samples were stained with 4′ 6-diamidino-2-phenylindole (DAPI) and phalloidin to show the nuclei and F-actin filaments, respectively. Fluorescence microscopy images were obtained using a BZ-X 700 fluorescent microscope system (Keyence, Osaka, Japan). For cell shape analysis, randomly selected cells (more than 60 cells per group) were used. Aspect ratio analysis was performed using a Hybrid Cell Counting application (Keyence) by dividing the major axis by the minor axis of cells.
Cyclic compression loading stimulation. Cyclic compressive loading was applied to sponges using a custom-designed apparatus, a cyclic load bioreactor (CLS-5J-Z, Technoview, Osaka, Japan), as previously described 22,46,47 . Briefly, cells were cultured in sponges on 96-well culture plates and a cyclic load compressive load of 40 kPa was applied to the sponges for 1 h at a rate of 0.5 Hz using metal pistons. Medium replacement was performed 12 h before, immediately before and after, and 12 h after stimulation. The collected culture supernatant was used to measure the protein concentration.
In vitro cytotoxicity assay. The amount of LDH released from the cells was measured using a Cytotoxicity LDH Assay Kit-WST (CK1205, Dojindo Molecular Technologies) according to the manufacturer's protocol. The culture supernatant 12 h before stimulation and the culture supernatant 12 h after stimulation were compared, and the ratio of the amount of protein contained was calculated.
Quantitative PGE2 protein analysis. An enzyme immunoassay was performed to measure the concentration of PGE2 in the culture supernatant using a HTRF human PGE2 assay kit (CIS Bio International, Saclay, France). The culture supernatant 12 h before stimulation and the culture supernatant 12 h after stimulation were compared, and the ratio of the amount of protein contained was calculated.
Statistical analysis. Data are presented as means and standard deviations (SDs). Single comparisons were performed using Student's t test, and multiple comparisons were performed using one-way analysis of variance (ANOVA) and a post hoc Tukey-Kramer test. Statistical analyses were performed using Ekuseru-Toukei (Social Survey Research Information Co., Ltd., Tokyo, Japan) and a p < 0.05 was considered statistically significant.