Low-density lipoprotein receptor-related protein 1 (LRP1) is known to be a receptor for signal transmission and endocytosis. We have previously reported that LRP1 regulates WNT–β-catenin and protein kinase C signaling in chondrocytes, represses the hypertrophy of chondrocytes during endochondral ossification and that LRP1 is colocalized with a ligand, CCN family member 2 (CCN2; also known as connective tissue growth factor, CTGF), which conducts endochondral ossification, in chondrocytes. However, the role of LRP1 in the endocytic transport of CCN2 in chondrocytes is not yet understood. In the present study, we investigated the interaction between LRP1 and CCN2 during endocytic trafficking. Small interfering RNA (siRNA)-mediated knockdown of LRP1 in chondrocytic HCS-2/8 cells showed that the amount of exogenous CCN2 binding and/or incorporation was decreased in the LRP1 downregulated cells. Importantly, we observed that CCN2 internalization in chondrocytes was dependent on clathrin, and internalizated CCN2 was colocalized with an early or recycling endosome marker. Transcytosis of CCN2 through HCS-2/8 cells was confirmed by performing experiments with a trans-well apparatus, and the amount of transcytosed CCN2 was decreased by an LRP1 antagonist. These findings rule out possible leakage and confirm the crucial involvement of LRP1 during experimental transcytosis. Moreover, under hypoxic conditions that mimic the cartilaginous microenvironment, the level of LRP1 and the amount of transcytosed CCN2 increased, and these increases were neutralized by treatment with the LRP1 antagonist. The distribution of LRP1 and its antagonist in the growth plate in vivo was consistent with that of CCN2 in this tissue, which is produced by and transported by LRP1 from the chondrocytes in the prehypertrophic layer. These findings suggest that LRP1 mediates the transcytosis of CCN2, which might be a crucial event that determines the distribution of CCN2 in cartilage.
The low-density lipoprotein receptor (LDLR)-related protein-1 (LRP1), is a 600-kDa type I membrane protein and a member of the LDLR family (Herz and Strickland, 2001). By interacting with over 40 distinct ligands, LRP1 is thought to regulate lipid homeostasis, extracellular proteolysis, growth factor and cytokine activity, composition of the extracellular matrix (ECM) and even immune responses (Herz, 2001; Herz and Strickland, 2001; Lillis et al., 2008). A substantial part of these LRP1 functions is thought to be related to clathrin-dependent endocytosis (Hussain, 2001) and cellular signal transduction pathways (Herz, 2001; Herz and Strickland, 2001; Lillis et al., 2008), such as protein kinase C (PKC) cascades (Hayashi et al., 2007). Recently, LRP1 has been shown to interact with human frizzled-1 to downregulate the canonical WNT–β-catenin signaling pathway (Zilberberg et al., 2004). Consistent with its functional diversity, LRP1 is essential for embryonic development. It has been reported that conventional Lrp1-deficient animals failed to develop normally and died during early to mid-gestation (Herz et al., 1992). Of note, the involvement of LRP1 in the prevention of atherosclerosis has also been indicated by experiments utilizing conditional gene targeting technology (Boucher et al., 2003). In addition, the expression and function of LRP1 in the central nervous system (May and Herz, 2003), vascular smooth muscle cells (Boucher et al., 2003) and macrophages (Gardai et al., 2003) have been relatively well characterized.
We have previously reported the distribution of LRP1 in normal cartilage (Kawata et al., 2006) and shown that LRP1 initiates the hypertrophy of chondrocytes during endochondral ossification through WNT–β-catenin and PKC signaling (Kawata et al., 2010). Vertebrate cartilage is of two distinct types, permanent cartilage represented by articular cartilage, and temporary cartilage represented by the growth plate cartilage, which is where endochondral ossification occurs. During endochondral ossification, chondrocytes first proliferate and then become mature cells that produce abundant ECM components such as type II collagen. Thereafter, the cells eventually differentiate into hypertrophic chondrocytes, which produce alkaline phosphatase and type X collagen. At the terminal stage of endochondral ossification, the cartilage matrix becomes mineralized and is invaded by blood vessels; then these hypertrophic chondrocytes are thought to undergo apoptosis. Thus, through this process, cartilage is replaced by bone (Nakanishi et al., 1997; Takigawa et al., 2003).
In our series of studies, we have uncovered crucial roles of CCN family member 2 (CCN2; also known as connective tissue growth factor, CTGF) in endochondral ossification (Ivkovic et al., 2003; Kubota and Takigawa, 2011; Nakanishi et al., 1997; Nakanishi et al., 2000; Nishida et al., 2002; Perbal and Takigawa, 2005; Takigawa et al., 2003) and regeneration of articular cartilage (Nishida et al., 2004). Interestingly, in vivo, CCN2 molecules are distributed in a layer different from that containing the Ccn2-mRNA-positive cells. Namely, whereas the chondrocytes expressing the Ccn2 mRNA are detected in the prehypertrophic chondrocyte zone, the CCN2 protein itself is broadly detected from the prehypertrophic zone to the hypertrophic chondrocyte zone (Oka et al., 2007). Here, it should be noted that CCN2 is one of the ligands of LRP1, which is known to mediate intracellular protein transport (Gao and Brigstock, 2003; Perbal, 2004; Segarini et al., 2001; Yang et al., 2004). We hypothesized that the difference in the distribution of Ccn2 mRNA expression and CCN2 protein in vivo was caused by the function of LRP1 in CCN2 protein trafficking. In this study, we show for the first time that LRP1 induces CCN2 transcytosis through chondrocytes in vitro.
Effect of LRP1 knockdown on CCN2 association with chondrocytic HCS-2/8 cells
First, to examine whether LRP1 participates in the association of CCN2 with chondrocytes or not, we performed RNA interference (RNAi) experiments to knockdown LRP1 in HCS-2/8 cells, a human chondrocytic cell line, as previously performed (Kawata et al., 2010). We confirmed that the production of LRP1 protein was substantially knocked down by si-1163 and si-13157 (Fig. 1). We added a recombinant CCN2 with a FLAG tag at the N-terminus to the control and LRP1 knockdown HCS-2/8 cells, and then collected the cells after 1 hour and performed immunoblotting using an anti-FLAG tag antibody. We found that the amount of recombinant CCN2 that was bound and/or incorporated was decreased in the LRP1 knockdown HCS-2/8 cells compared with that in the control cells (Fig. 1). The result of immunoblotting with anti-His tag antibody recognizing a 6×His tag fused to the CCN2 C-terminus was similar to that obtained by using the anti-FLAG tag antibody (Fig. 1). These results indicate that LRP1 participates in associating CCN2 with chondrocytes.
Effect of clathrin inhibition on CCN2 association with chondrocytic HCS-2/8 cells
Second, to examine whether the association of CCN2 with chondrocytes depends on clathrin or not, we evaluated the effect of endocytosis inhibitors. Namely, we added a recombinant CCN2 with a FLAG tag at the N-terminus to HCS-2/8 cells pretreated either with the clathrin-dependent endocytosis inhibitor chlorpromazine or with the caveolin-dependent endocytosis inhibitor methyl-β-cyclodextrin (MβCD) and then collected the cells after 1 hour and performed immunoblotting using the anti-FLAG tag antibody. We found that the bound and/or incorporated amount of recombinant CCN2 was decreased in the HCS-2/8 cells pretreated with chlorpromazine compared with that in the control cells treated with vehicle only (Fig. 2A). The result of immunoblotting with anti-His tag antibody recognizing the C-terminal 6×His tag was comparable to that obtained with the anti-FLAG tag antibody (Fig. 2A). By contrast, in the HCS-2/8 cells pretreated with MβCD, the bound and/or incorporated amount of recombinant CCN2 was unchanged compared with that in the control cells (Fig. 2B). These results indicate that the association of CCN2 with chondrocytes is dependent on clathrin.
Intracellular destination of CCN2 taken up into HCS-2/8 cells
Subsequently, we added recombinant human CCN2 (rhCCN2) and followed the fate after the application onto HCS-2/8 cells. We found that exogenously added rhCCN2 and endogenous LRP1 also partially colocalized inside the HCS-2/8 cells (Fig. 2C,D). This result again suggests the contribution of LRP1 in the endocytic incorporation of CCN2. Excluding the signals from nascent CCN2 in exosomes on the way to secretion, we followed the fate of exogenous CCN2 after internalization by using the anti-FLAG antibody. As confirmed by double-staining with organelle-specific markers, incorporated CCN2 was directed to clathrin (Fig. 2E,F), early (Fig. 2G,H) and recycling endosomes (Fig. 2I,J). Particularly, exogenously added rhCCN2 and the recycling endosomes marker were predominantly colocalized in HCS-2/8 cells (Fig. 2I,J). Therefore, CCN2 that is internalized in HCS-2/8 cells is directed to recycle out of the cells in any direction, partially following the pathway directed by LRP1.
Effect of LRP1 on CCN2 transcytosis in chondrocytes
We considered that the broad localization of CCN2 protein (i.e. from the prehypertrophic zone to the hypertrophic zone of growth plate cartilage) is possibly mediated by LRP1-mediated transcytosis of this protein from the chondrocytes expressing Ccn2 in the prehypertrophic zone. To test the validity of this hypothesis, we performed a transcytosis assay using CCN2 with or without the LRP1 antagonist LRP-associated protein 1 (LRPAP1). After the addition of exogenous CCN2 with an N-terminal FLAG tag to the cells in the upper transcytosis chamber (Fig. 3A), the exogenous full-length CCN2 was detected in cell lysates by using anti-FLAG tag antibody (Fig. 3B). Moreover, the result of immunoblotting with an anti-His tag antibody, recognizing a 6×His tag fused to the CCN2 C-terminus, was similar to that obtained with the anti-FLAG tag antibody (Fig. 3B). These signals were decreased after addition of LRPAP1 (Fig. 3B). Results similar to those found with the cell lysate were obtained from the medium in the lower chamber (Fig. 3C), which indicates that the LRP1 is functionally involved in CCN2 transcytosis and rules out the possibility of substantial leakage through the uncovered part of the membrane. Because we basically use CCN2 derived from Escherichia coli, we repeated the same analysis with another recombinant CCN2 from HeLa cells, in order to rule out the possibility of the contamination with E. coli components. We found that exogenous biotin-labeled full-length CCN2 was detected in cell lysates using horseradish peroxidase (HRP)-conjugated avidin, and that the amount of protein was decreased upon addition of LRPAP1 (Fig. 3D). Results similar to those found with the cell lysate were obtained with the medium in the lower chamber (Fig. 3E). These results clearly indicate that CCN2 is transcytosed, as well as bound and incorporated in chondrocytes in a process mediated by LRP1.
Effect of hypoxia on the levels of LRP1 mRNA and protein in HCS-2/8 cells
Cartilage is an avascular tissue; therefore, it has been assumed that the low oxygen partial pressure in the chondrocytic growth plate imposes energetic limitations on the cells as they evolve from a proliferative into a terminally differentiated state (Rajpurohit et al., 1996). Suspecting the contribution of the oxygen pressure gradient to the hypertrophic-layer-specific localization of CCN2, we investigated transcytosis of CCN2 mediated by LRP1 under hypoxic conditions. First, we examined the levels of LRP1 mRNA expression and production of LRP1 protein under hypoxic conditions. Exposure to hypoxia resulted in a time-dependent increase in the LRP1 mRNA expression level (Fig. 4A). Moreover, the level of LRP1 protein was also increased under hypoxic conditions (Fig. 4B). Consistent with previous reports (Semenza and Wang, 1992; Shimo et al., 2001), the production level of CCN2 and hypoxia-inducible factor 1α (HIF1α) protein was also increased under hypoxic condition (Fig. 4B). Thus, to determine whether HIF1α mediates the hypoxia-induced LRP1 and CCN2 production, we examined the effect of the antisense oligonucleotides targeting HIF1α-encoding mRNA (HIF1A) on LRP1 and CCN2 production, respectively, under hypoxic conditions. Cells cultured in 5% O2 with antisense oligonucleotides against HIF1A (but not with sense oligonucleotides) abolished LRP1 and CCN2, as well as HIF1α induction (Fig. 4C), but β-actin levels were unaffected (Fig. 4C). Decreased CCN2 by HIF1α downregulation agrees with the results of a previous study (Hong et al., 2006). More importantly, these data indicate that HIF1α regulates not only CCN2 but also LRP1 production under hypoxic conditions. Additionally, these results agree with the results of a previous study, in which the level of LRP1 was found to be drastically decreased in the hypertrophic zone of the cartilage near the bone marrow (Kawata et al., 2006).
Effect of hypoxia on CCN2 transcytosis in chondrocytes
Second, we evaluated the effect of hypoxia on actual CCN2 transport in the same Transwell system. Immunoblotting with anti-FLAG or His tag antibody revealed that the amount of exogenous CCN2 in the cell lysate was increased under the hypoxic conditions compared with that under normoxic conditions (Fig. 5A). Results similar to those for the cell lysate were obtained when using the medium in the lower chamber of the system (Fig. 5B). In both cases these signals were decreased by LRPAP1 (Fig. 5), again confirming the functional involvement of LRP1. These results indicate that the transcytosis of CCN2 in chondrocytes by LRP1 is increased under the hypoxic conditions compared with that under normoxic conditions.
Higher mRNA and protein levels of LRPAP1 in the chondrocytic cell line
We formerly reported that the LRP1 levels are higher in HCS-2/8 cells than in other types of cells (Kawata et al., 2006). On the basis of this finding, we next compared mRNA and protein levels of the LRP1 antagonist LRPAP1 in the chondrocytic HCS-2/8 cells with those in the other cell lines. LRP1 and LRPAP1 were analyzed comparatively in HCS-2/8 cells, breast-cancer-derived MDA-MD-231 cells (MDA-231), and cervical-carcinoma-derived HeLa cells by use of real-time quantitative RT-PCR and immunoblotting. We found that the expression level of LRP1 was certainly higher in HCS-2/8 cells than in HeLa and MDA-231 cells, as previously reported (Kawata et al., 2006). Similarly, the mRNA level of LRPAP1 was higher in the HCS-2/8 cells than in the other cells (Fig. 6A). Furthermore, although LRPAP1 protein was detected in all of the cell lines tested, it was more abundant in HCS-2/8 cells than in HeLa and MDA-231 cells (Fig. 6B). This result indicates that LRPAP1 was specifically induced in chondrocytic HCS-2/8 cells.
Distribution of LRPAP1 in growth-plate cartilage in vivo
Next, to investigate the production of LRPAP1 in growth-plate cartilage, we performed immunostaining analysis using anti-LRPAP1 antibody and tibial sections prepared from mice. As a result, LRPAP1 was clearly detected, particularly in resting chondrocytes, where CCN2 was absent (Fig. 7A). We previously reported that LRP1 is present in the growth-plate cartilage, but is drastically decreased along the hypertrophic zone of the cartilage (Kawata et al., 2006). Therefore, LRP1-mediated transcytosis was supposedly suppressed by LRPAP1 in resting chondrocytes.
Changes in the expression of LRPAP1 mRNA in chondrocytes at various differentiation stages in vitro
Following the in vivo analysis, we analyzed the changes in the levels of LRPAP1 mRNA in chondrocytes during differentiation in vitro. To do this, we employed an established differentiation system using primary chicken chondrocytes (Iwamoto et al., 1995). For the evaluation of gene expression, real-time quantitative RT-PCR was performed (Fig. 7B). We initially confirmed that lower sternum (LS), upper sternal peripheral (USP) and upper sternal core (USC) cells represented resting, proliferating and hypertrophic chondrocytes, respectively. Indeed, the COL2A1 mRNA level was the highest in the LS cells, whereas the COL10A1 mRNA level was the highest in the USC cells. Moreover, the ALP mRNA level was lower in the LS cells than in the other cells. Under this condition, the expression of LRPAP1 mRNA was the highest in LS cells, confirming strong distribution of LRPAP1 protein in resting chondrocytes in vivo.
CCN2 acts in a harmonized manner on all cells involved in the promotion of endochondral ossification. In the growth-plate in vivo, CCN2 protein is distributed in a pattern different from that of its mRNA. Namely, whereas Ccn2 mRNA expression is found in the prehypertrophic chondrocytes, CCN2 protein is detected broadly, from the prehypertrophic to the hypertrophic zone (Oka et al., 2007). This broad distribution enables CCN2 to contact all the target cells, and thus is of crucial importance. Here, we should note that CCN2 is one of the ligands of LRP1 (Gao and Brigstock, 2003; Perbal, 2004; Segarini et al., 2001; Yang et al., 2004). We considered that such a distribution of CCN2 protein was achieved through LRP1-mediated transcytosis. To test the validity of this hypothesis, we performed this study and found that not only CCN2 binding/incorporation but also CCN2 transcytosis was indeed mediated by LRP1 (Figs 3, 5).
Because the growth-plate is an avascular tissue, it has been assumed that the low oxygen partial pressure in the chondrocytic growth-plate imposes energetic limitations on the cells as they differentiate from a proliferative to a terminally differentiated state (Rajpurohit et al., 1996). Therefore, mimicking such in vivo conditions, we examined the levels of LRP1 mRNA and LRP1 protein production under hypoxic conditions in vitro. We found that exposure to hypoxia increased both mRNA expression and protein production levels of LRP1 (Fig. 4). These results support a previous finding where the level of LRP1 is drastically decreased along the hypertrophic zone of the cartilage (Kawata et al., 2006). Moreover, in a previous study, increased mRNA expression of the Lrp1 gene in carcinoma-derived cells under hypoxia has been demonstrated (Koong et al., 2000). Consistent with these findings, transcytosis of CCN2 by LRP1 in chondrocytes was increased under hypoxic conditions compared with under normoxic conditions (Fig. 5). Thus, we propose that the distribution of CCN2 in growth-plate cartilage is possibly controlled by LRP1 regulated by hypoxia (Fig. 8).
Although cartilage is avascular, the cartilage matrix is invaded by blood vessels at the terminal stage of endochondral ossification (Nakanishi et al., 1997; Takigawa et al., 2003). Therefore, the level of the oxygen tension might be different according to the difference in the quantity of the oxygen supply between the entirely avascular layer of resting chondrocytes and layer of hypertrophic chondrocytes that is invaded by blood vessels. According to such an oxygen tension gradient, it would be suspected that LRP1 production was repressed in the late hypertrophic layer, as described previously, which might prevent the flow-through of CCN2 into the bone marrow.
Finally, we should comment on the distribution of LRPAP1, the natural LRP1 antagonist, in the growth plate. We previously reported that the LRP1 levels were higher in HCS-2/8 cells than in the other cells (Kawata et al., 2006). As was shown here, the mRNA and protein levels of LRPAP1 were also higher in HCS-2/8 cells than in HeLa and MDA-231 cells (Fig. 6). These results indicate that LRPAP1 was specifically induced in chondrocytes, and concomitant expression and production of LRP1 and LRPAP1 suggests a particular role for LRPAP1 in the endocytic pathway in chondrocytes. Importantly, LRPAP1 was detected particularly in resting chondrocytes, where CCN2 was absent (Fig. 7). Thus, in the growth-plate cartilage, CCN2 produced in the prehypertrophic chondrocyte layer might not be transcytosed to the resting chondrocyte layer, being prevented by the higher level of LRPAP1. Collectively, localization of CCN2 in the growth plate would supposedly be maintained by LRP1 under the interaction with LRPAP1 and the hypoxic gradient therein (Fig. 8).
Materials and Methods
Antibodies and reagents
For immunoblotting and/or immunofluorescence microscopy, anti-FLAG M2 MONOCLONAL (Sigma-Aldrich, St Louis, MO), anti-6-His (BETHYL, Montgomery, TX), monoclonal 5A6 (Progen, Heidelberg, Germany) recognizing the 85-kDa LRP1 light chain, a rabbit polyclonal H-80 antibody (Santa Cruz Biotechnology, Santa Cruz, CA), which recognizes amino acids 206–285 of LRP1, anti-β-Actin AC-74 (Sigma-Aldrich), anti-clathrin heavy chain P1663 (Cell Signaling Technology, Inc., Danvers, MA), anti-EEA1 (Cell Signaling Technology, Inc.), anti-Rab11 (C-19; Santa Cruz Biotech), anti-CTGF ab6992 (abcam, Cambridge, England), anti-hypoxia inducible factor (HIF) 1α clone H1α67 (Millipore, Billerica, MA), and anti-LRPAP1 rabbit monoclonal (Epitomics, Inc., Burlingame, CA) antibodies were employed. As secondary antibodies, horseradish peroxidase (HRP)-conjugated anti-mouse-IgG antibody was purchased from GE Healthcare, HRP-conjugated anti-rabbit-IgG antibody, from BETHYL; and Alexa-Fluor-488-conjugated goat anti-rabbit-IgG, Alexa-Fluor-568-conjugated goat anti-mouse-IgG, Alexa-Fluor-488-conjugated donkey anti-goat-IgG, from Molecular Probes (Eugene, OR). Streptavidin–HRP conjugate was purchased from Zymed Laboratories (San Francisco, CA). Chlorpromazine and methyl-β-cyclodextrin (MβCD) were purchased from LKT Laboratories (St. Paul, MN) and Sigma-Aldrich, respectively.
HCS-2/8 cells (a human chondrocytic cell line) (Takigawa et al., 1991; Takigawa et al., 1989) HeLa cells (a human cervical cancer cell line), and MDA-MB-231 cells (MDA-231; a human breast cancer cell line) were cultured in Dulbecco’s modification of minimum essential medium (D-MEM) containing 10% fetal bovine serum (FBS). The cells were cultured at 37°C in humidified air with 5% CO2. Hypoxia experiments were performed for the desired times in a humidified triple gas model BL-40M incubator (BIO-LABO, Tokyo, Japan) calibrated to deliver 5% CO2, 5% O2, and 90% N2 at 37°C for 8–48 hours. Primary chicken chondrocytes were isolated from the caudal one-third portions (LS) of the sterna, the peripheral regions (USP) and central core regions (USC) of the cephalic portions of the sterna of day-17 chick embryos by using the method described earlier (Iwamoto et al., 1995). All chicken experiments were performed according to approved guidelines.
LRP1 siRNA transfection
To knockdown LRP1 protein production, we used RNA interference technology. Two designed pairs of RNA oligoduplexes targeting human LRP1 (gene accession no. NM_002332) were purchased from Hokkaido System Science Co., Ltd. (Sapporo, Japan). The target nucleotide sequences of those oligoduplexes were 5′-UGGACUAUAUUGAAGUGGUGGACUAAG-3′ and 5′-CCUGUACCAUGAACAGCAAAAUGAUAG-3′. The former was termed LRP-1163, and the latter LRP-13157. A nonspecific oligoduplex (nonsilencing control, targeting 5′-UUAGGGGAUAAGUACGGUUGAAUCUAG-3′) was used as a negative control at the same final concentrations as used for the human LRP1-targeting RNA duplexes. Prior to transfection, the cells were transferred to each well in 6-well plates (density: 4×105 cells/well). Transient transfection with a 70 nM concentration of siRNA was performed by using siPORT NeoFXTM Transfection Agent (Applied Biosystem, Foster City, CA) according to the manufacturer's protocol. At 24 hours after the transfection, the medium was exchanged for fresh medium; and the cells were then cultured for another 48 hours.
Expression and purification of full-length recombinant human CCN2 (rhCCN2)
Association and transcytosis assay of CCN2
The HCS-2/8 cells were washed three times on ice with cold phosphate-buffered saline (PBS), and LRPAP1 was added to the cells. After 15 minutes, the cells were allowed to associate with CCN2 in serum free D-MEM containing 2 µg/ml recombinant CCN2 at 37 °C for 1 hour. For inhibitor assays, prior to addition of CCN2, HCS-2/8 cells were preincubated for 5 minutes at 37°C in medium lacking FBS and with 5 µM chlorpromazine or 3 mM MβCD to specifically inhibit endocytic pathways. Then, the cells were washed three times on ice with cold PBS. After that, cell layers (total binding samples) were harvested in 100 µl of lysis buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM Na3VO4, 5% glycerol, 40 mM ammonium molybdate, and 1 mM phenylmethylsulfonyl fluoride). For the transcytosis assays of CCN2, HCS-2/8 cells were seeded in Transwell chambers with the pore size of 0.4 µM (Millipore, Billerica, MA), which had been inserted into 6-well culture plates (density: 6×105 cells/well) containing D-MEM supplemented with 10% FBS and incubated at 37°C for 1 week. The cells were washed 3 times on ice with cold PBS and then allowed to associate with CCN2 at 37°C for 1 hour in serum-free D-MEM containing 2 µg/ml recombinant CCN2. All cells were subsequently washed three times on ice with cold PBS. Cell lysates (total binding samples) on upper chambers were harvested in 100 µl of lysis buffer. The medium in the lower chamber (transcytosis sample) was harvested, and then anti-FLAG M2 affinity gel freezer-safe (Sigma-Aldrich) or Ni-NTA (Ni2+-nitrilotriacetate)–agarose gel (Qiagen, Hilden, Germany) was added to it, which mixture was subsequently incubated for 2 hours with gentle rotation to capture the FLAG or His-tagged protein in the medium. After the removal of the supernatant, then PBS and 1× SDS sample buffer (50 mM Tris-HCl pH 6.8, 2% SDS, 5% glycerol, 2% bromphenol blue) with 2-mercaptoethanol was added to the gel to elute the bound proteins.
HCS-2/8 cells were lysed in the lysis buffer. The lysate, diluted in 1× SDS sample buffer with or without 2-mercaptoethanol, was boiled for 3 minutes and was then subjected to SDS–PAGE on 9% or 12% polyacrylamide gels. Proteins were transferred onto polyvinylidene difluoride (PVDF) membranes with a blotting apparatus. The membranes were then incubated for 1 h in a blocking buffer (3% dry non-fat milk in PBS) and subsequently incubated overnight with anti-FLAG (1∶1000), anti-GAPDH (1∶100), anti-6-His (1∶1000), anti-LRP1 5A6 (1∶100), anti-β-actin (1∶5000), anti-CCN2 (1∶1000), anti-HIF1α (1∶1000), anti-LRPAP1 (1∶500) antibody, or Streptavidin HRP conjugate (1∶2000) in the blocking buffer. Next, the membrane was washed 5 times in PBS and then incubated for 2 h with HRP-conjugated anti-mouse (1∶5000) or anti-rabbit (1∶5000) IgG in the blocking buffer. After extensive washes with PBS, immunoreactive proteins were detected by using an ECL Western Blotting Detection System (Amersham Biosciences, Piscataway, NJ).
HCS-2/8 cells were cultured on glass coverslips, fixed in 4% paraformaldehyde (w/v) in a phosphate buffer (PB) for 15 minutes and permeabilized with 0.2% Triton X-100 for 15 minutes. Primary anti-LRP1 H-80 (1∶40), anti-FLAG (1∶200), anti-clathrin (1∶300), anti-EEA1 (1∶100) and anti-Rab11 C-19 (1∶40) antibodies were used for detection. Alexa-Fluor-labeled secondary antibodies were also utilized at 1∶500.
Confocal laser-scanning microscopy
Confocal laser microscopy was performed using a ZEISS Confocal Laser Scaning Microscope Model LSM510 (Carl Zeiss, Oberkochen, Germany) belonging to Central Research Laboratory, Okayama University Medical School.
Biotin labeled CCN2 protein derived from HeLa cells
CCN2 protein derived from HeLa cells was biotin-labeled by a commercially available kit, following the manufacturer's instructions (Biotin Labeling Kit-NH2; Dojindo Molecular Technologies, Kamimashiki-Gun, Japan).
RNA extraction and cDNA synthesis
Cells were collected, and total RNA was extracted by following the manufacturer's instructions (RNeasy kit, Qiagen). Total RNA (500 ng) was reverse-transcribed by AMV Reverse Transcriptase (Takara, Ohtsu, Japan) at 42°C for 30 minutes, according to the manufacturer's protocol.
Real-time PCR was performed by using TOYOBO SYBR Green PCR Master Mix (TOYOBO, Osaka, Japan) in a LightCyclerTM system (Roche, Basel, Switzerland). Reactions were performed in a 10-µl reaction mixture containing 1 µl of cDNA, 0.4 µl of each primer (5 µM), and 5 µl of 1× SYBR Green master mix. Primer sets and optimized conditions for the PCR of each target are listed in supplementary material Table S1. Absence of non-specific PCR products was checked by melting curve and electrophoresis analyses. Relative copy numbers were computed based on data obtained with a serial dilution of a representative sample for each target gene.
To inhibit the expression of HIF1α, we prepared an antisense phosphorothioate oligonucleotide (AS-HIF) and a sense oligonucleotide (S-HIF: control) according to the nucleotide sequence of the human HIF1α gene (Caniggia et al., 2000). The nucleotide sequences of the AS-HIF and S-HIF were 5′-GCCGGCGCCCTCCAT-3′ and 5′-ATGGAGGGCGCCGGC-3′, respectively. These oligonucleotides were added directly to medium in HCS-2/8 cell culture at a concentration of 10 µM.
Animals and preparation of tissue
After Balb/cj mice (2 weeks of age) had been anesthetized with sodium pentobarbital (Nembutal, Abbott laboratories, North Chicago, IL; 25 mg/kg), proximal tibiae were harvested and immersed in 4% paraformaldehyde (w/v) in phosphate buffer (PB: 0.1 M NaH2PO4, 0.1 M Na2HPO4, pH 7.4) at 4°C overnight. After having been rinsed in PBS, the tibiae were decalcified in 0.5 M EDTA, pH 7.4, at 4°C and then embedded in paraffin wax. The sections were prepared at a thickness of 7 µm and mounted on silane-coated slides. The Animal Committee of Okayama University approved all of the procedures.
Tibial sections were dewaxed in xylene and rehydrated through a graded series of ethanol to water, blocked in a blocking buffer (5% dry non-fat milk in Tris-buffered saline), and incubated overnight at 4°C with the primary anti-LRPAP1 antibody (1∶100) and subsequently with an HRP-conjugated anti-rabbit-IgG (1∶1000) for 1 hour at room temperature. Color development was performed by using 3, 3′-diaminobenzidine tetrachloride (Dojindo, Tokyo, Japan). The sections were also counterstained with hematoxylin and mounted. Control samples were processed with the omission of the primary antibody.
Data were presented as means ± standard deviations, and the statistical significance of differences in mean values was assessed by performing Student's unpaired t-test. Differences among the mean values were considered significant at a P value of <0.05. All experiments were repeated at least twice, and similar results were obtained.
Masaharu Takigawa would like to dedicate this article to the memories of his father, Dr Masami Takigawa, and Professor Emeritus Yoshiro Takeda, who passed away while this research was being conducted. The authors thank Drs Takano Hattori and Eriko Aoyama for valuable discussion and Ms Eri Yashiro for secretarial assistance.
This work was supported by the programs Grants-in-Aid for Scientific Research (S) [grant number 19109008 to M.T.] and (C) [grant number 21592360 to S.K.] and Grants-in-Aid for Exploratory Research [grant number 23659872 to M.T.] from Japan Society for the Promotion of Science; by Grants-in-Aid for Fellows of the Japanese Society for the Promotion of Science [grant number 03J02535 to T.E.]; and by grants from the Sumitomo Foundation (to M.T.), Foundation of Sanyo Broadcasting (to S.K.), Terumo Life Science Foundation [grant number 10-312 to S.K.], and Ryobi Teien Memory Foundation (to K.K.).