CCN2, amino acids

Role of CCN2 in Amino Acid Metabolism of Chondrocytes Yurika Murase,1,2 Takako Hattori,1 Eriko Aoyama,3 Takashi Nishida,1

Aya Maeda-Uematsu,1 Harumi Kawaki,1 Karen M. Lyons,4 Akira Sasaki,2

Masaharu Takigawa,1,3* and Satoshi Kubota1,3* 1Department of Biochemistry and Molecular Dentistry, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan

2Department of Oral and Maxillofacial Surgery, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan

3Advanced Research Center for Oral and Craniofacial Sciences, Okayama University Dental School, Okayama, Japan 4Department of Orthopaedic Surgery, UCLA School of Medicine, Los Angeles, California

ABSTRACT CCN2/connective tissue growth factor (CTGF) is a multi-functional molecule that promotes harmonized development and regeneration of cartilage through its matricellular interaction with a variety of extracellular biomolecules. Thus, deficiency in CCN2 supply profoundly affects a variety of cellular activities including basic metabolism. A previous study showed that the expression of a number of ribosomal protein genes was markedly enhanced in Ccn2-null chondrocytes. Therefore, in this study, we analyzed the impact of CCN2 on amino acid and protein metabolism in chondrocytes. Comparative metabolome analysis of the amino acids in Ccn2-null and wild-type mouse chondrocytes revealed stable decreases in the cellular levels of all of the essential amino acids. Unexpectedly, uptake of such amino acids was rather enhanced in Ccn2-null chondrocytes, and the addition of exogenous CCN2 to human chondrocytic cells resulted in decreased amino acid uptake. However, as expected, amino acid consumption by protein synthesis was also accelerated in Ccn2-null chondrocytes. Furthermore, we newly found that expression of two genes encoding two glycolytic enzymes, as well as the previously reported Eno1 gene, was repressed in those cells. Considering the impaired glycolysis and retained mitochondrial membrane potential in Ccn2-null chondrocytes, these findings suggest that Ccn2 deficiency induces amino acid shortage in chondrocytes by accelerated amino acid consumption through protein synthesis and acquisition of aerobic energy. Interestingly, CCN2 was found to capture such free amino acids in vitro. Under physiological conditions, CCN2 may be regulating the levels of free amino acids in the extracellular matrix of cartilage. J. Cell. Biochem. 117: 927–937, 2016. © 2015 Wiley Periodicals, Inc.

KEY WORDS: CCN2; CTGF; CARTILAGE; CHONDROCYTES; METABOLISM

The CCN family is a group of matricellular proteins with aconserved molecular structure and unique functionality that integrates extracellular signaling networks. This family name was given in 1993, by assembling the initials of the original names of its three classical members: cysteine-rich protein 61 (CYR61), connective tissue growth factor (CTGF), and nephroblastoma-overexpressed (NOV) gene product [Bork, 1993]. After the establishment of this family, three additional members were discovered by independent research groups and became widely known as Wnt-inducible secretory proteins (WISP)- 1, -2, and -3 [Leask and Abraham, 2006; Jun and Lau, 2011; Kubota

and Takigawa, 2012]. For these six family members, new names CCN1- 6 were assigned in order, based on the unified nomenclature [Brigstock et al., 2003]. All of the members consist of four conserved modules: insulin-like growth factor binding protein (IGFBP) module, von Willebrand factor type C repeat (VWC) module, thrombospondin type 1 repeat (TSP1) module, and carboxyl-terminal (CT) module. Only this last one is absent in CCN5. Every module is highly interactive with a variety of molecular counterparts. Generally, by interacting with such cofactors as cell-surface receptors and extracellular signaling molecules, the CCN family proteins play a role in the regulation of

The authors have no conflicts of interest to declare. Grant sponsor: Japan Society for the Promotion of Science; Grant numbers: 25462886, 24390415, 15H0514. *Correspondence to: Dr. Satoshi Kubota and Dr. Masaharu Takigawa, Department of Biochemistry and Molecular Dentistry, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, and Advanced Research Center for Oral and Craniofacial Sciences, Okayama University Dental School, 2-5-1 Shikata-cho, Okayama 700-8525, Japan. E-mail: [email protected] (S.K.), [email protected] (M.T.) Manuscript Received: 28 May 2015; Manuscript Accepted: 10 September 2015 Accepted manuscript online in Wiley Online Library (wileyonlinelibrary.com): 14 September 2015 DOI 10.1002/jcb.25377 � © 2015 Wiley Periodicals, Inc. 927

ARTICLE Journal of Cellular Biochemistry 117:927–937 (2016)

various cellular activities, such as cell-cycle control, adhesion, migration, and extracellular matrix (ECM) remodeling [Perbal and Takigawa, 2005; Kubota and Takigawa, 2007, 2015].

Among the six family members, CCN2 has been the best investigated upto today, and a number of reports revealed the multiple functionality of CCN2 in a variety of tissues and organs, under either physiological or pathological biological conditions. During normal animal development, CCN2 plays a critical role in the proper development of the skeleton, central nervous system, pancreas, hair, and teeth [Charrier and Brigstock, 2013; Kubota and Takigawa, 2015]. On the other hand, during abnormal tissue remodeling, CCN2 not only mediates the development of fibrotic disorders [Leask and Abraham, 2006] but also modulates the development of tumors in a variety of tissues and organs [Jun and Lau, 2011; Kubota and Takigawa, 2013]. In most cases, with the exception being ovarian, oral, and lung cancers, CCN2 is observed to promote the development of tumors including breast cancer, prostate cancer, glioma, pancreatic cancer, colon cancer, thyroid carcinoma, chondrosarcoma, gallbladder carcinoma, melanoma, and leukemia. These effects of CCN2 can be principally ascribed to the angiogenic property of the molecule [Babic et al., 1999; Shimo et al., 1999, 2001; Kubota and Takigawa, 2007].

In relation to skeletal development, the role of CCN2 in endochondral ossification, which determines the size of long bones, is of particular note. CCN2, predominantly produced by pre- hypertrophic chondrocytes, promotes all of the processes involved by encouraging the participation of a variety of the cells that operate in endochondral ossification [Babic et al., 1999; Shimo et al., 1999; Nakanishi et al., 2000; Nishida et al., 2000; Safadi et al., 2003; Smerdel-Ramoya et al., 2008; Kawata et al., 2012; Takigawa, 2013; Kubota and Takigawa, 2015]. In fact, CCN2-deficient cartilage exhibits abnormal organization of growth plate chondrocytes with delayed ossification, which results in remarkable skeletal defects [Ivkovic et al., 2003; Kawaki et al., 2008; Takigawa, 2013]. The cartilage-regenerating activity of CCN2 and its derivatives is also of note [Nishida et al., 2004; Abd El Kader et al., 2014].

In a recent study, an interesting functional aspect of CCN2 was uncovered by metabolomic and transcriptomic investigation of Ccn2-null chondrocytes. According to the results of that study, Ccn2-null chondrocytes revealed a stable reduction in their cellular ATP level [Maeda-Uematsu et al., 2014]. The results of both microarray and quantitative RNA analyses showed down- regulation of a gene encoding one of the glycolytic enzymes, Eno1, in response to Ccn2 deletion. Therefore, it has been suggested that CCN2 plays an important role in endochondral ossification by enhancing ATP production, where it is required, by enhancing the gene expression of Eno1. Microarray analysis also picked up a number of genes that are rather up-regulated by Ccn2 deletion. Of note, among these genes were a number of ribosomal protein genes highly suspected of affecting the protein synthesis and amino acid metabolism in Ccn2-null chondrocytes. However, no further investigation to clarify the role of CCN2 therein was conducted at that time.

Based on this previous finding, here we investigated the functional impact of CCN2 on amino acid metabolism, mainly utilizing Ccn2-null chondrocytes. Taken together with subsequent

analysis in vitro, a novel role of CCN2 in amino acid metabolism in cartilage was indicated.

MATERIALS AND METHODS

CELL CULTURE Primary costal and epiphyseal chondrocytes were isolated from rib and epiphyseal cartilage of Ccn2-null mice and wild-type littermates at E17.5, 18.5, or E19.5, following an established protocol as previously described [Kawaki et al., 2008; Hattori et al., 2010]. Briefly, after careful elimination of soft tissues by digestion with 0.25% trypsin for 5 min at 37°C, the cartilage was digested with 1.5 mg/ml collagenase A (Roche, Basel, Switzerland) for 2 or 3 h at 37°C to liberate the chondrocytes. Isolation of the cells was performed according to the Guidelines for Animal Research of Okayama University and was approved by the animal committee. These cells were then inoculated at a density of 1.5 � 105 cells/dish into 3.5 cm dishes or 3 � 104 cells/well into 24-well multiwell plates containing Dulbecco0s modified Eagle0s medium (DMEM) supple- mented with 10% fetal bovine serum (FBS) and then incubated at 37°C under 5% CO2 in air. Cells of the human chondrosarcoma- derived cell line HCS-2/8 [Takigawa et al., 1989] were inoculated at a density of 5 � 105 cells/dish into 3.5 cm dishes containing DMEM supplemented with 10% FBS, followed by incubation at 37°C under 5% CO2 in air.

MICROARRAY ANALYSIS Comparative transcriptomic analysis was performed by using a mouse Panorama Micro Array (Sigma–Aldrich, St. Louis, MO) following the manufacturer0s instruction. Total RNA was extracted from eight individual mouse embryos from four different litters, as previously described [Maeda-Uematsu et al., 2014], and the RNA mixture was subjected to labeling. Signals were quantified and analyzed by use of a GenePix 4000B (Molecular Devices, Sunnyvale, CA).

METABOLOME ANALYSIS Extraction of total metabolites from murine chondrocytes was performed by the method recommended by Human Metabolome Technologies (Tsuruoka, Japan), as described previously [Maeda- Uematsu et al., 2014]. Briefly, chondrocytes isolated from mouse cartilage were inoculated into a 6-well multiwell plate at a density of 3.6 � 105 cells/well and cultured to confluence. Then, the cells were washed with 5% mannitol solution (Wako, Osaka, Japan), after which methanol containing 10 mM Internal Standard Solution (Human Metabolome Technologies) was added; and then the cells were collected. Prior to the analysis, these samples were mixed with chloroform and water and centrifuged at 5,000�g for 5 min at 4°C. After the removal of proteins by filtration (5 kDa, Millipore, Billerica, MA), the samples were subjected to cationic and anionic metabolite analysis with a capillary electrophoresis time-of-flight mass spectrometer (CE-TOFMS; Agilent CE-TOFMS system, Agilent Technologies Japan, Ltd., Tokyo, Japan). Cationic metabolites were analyzed by using a fused silica capillary (i.d. 50 mm � 80 cm), with Cation Buffer Solution (Human Metabolome

JOURNAL OF CELLULAR BIOCHEMISTRY928 CCN2 AND AMINO ACID METABOLISM

Technologies) as the electrolyte at an injection pressure of 50 mbar for 10 s under 27 kV. Electrospray ionization-mass spectrometry (ESI-MS) in the positive-ion mode was performed with the applied capillary voltage at 4,000 V, and the scan range of the spectrometer was from 50 to 1,000 mass-to-charge ratio (m/z). Anionic metabolites were analyzed with Anion Buffer Solution (Human Metabolome Technologies) as the electrolyte at an injection pressure at 50 mbar for 25 s under 30 kV. ESI-MS was performed in the negative-ion mode at a capillary voltage of 3,500 V, and the scan range of the spectrometer was from 50 to 1,000 m/z. The peaks detected by CE-TOFMS were processed by using software (MasterHands ver. 2.9.0.9, Keio University, Tokyo, Japan). The concentration of each material was calculated with reference to the concentration of the internal standard material (200 mM).

RNA EXTRACTION AND REAL-TIME REVERSE-TRANSCRIPTION POLYMERASE CHAIN REACTION (RT-PCR) ANALYSIS Total RNA was extracted and purified by using an RNeasy Mini Kit according to the manufacturer0s instructions (Qiagen, Hilden, Germany) and was then reverse-transcribed to cDNA by use of avian myeloblastosis virus reverse transcriptase with an oligo d(T) as a primer (TaKaRa RNA PCRTM Kit Ver. 3.0, Takara Shuzo, Tokyo, Japan). Quantitative real-time PCR was carried out by using TOYOBO SYBR Green PCR Master Mix (TOYOBO, Osaka, Japan) with a StepOnePlusTM Real-Time PCR Systems (Applied Biosys- tems, CA). Primers used for the amplification of each cDNA were as follow: 50- CCA AAT GGA ACA CAG AGG ATA AAG -30 (sense) and 50- AAC ACT AGG TTG ACT TAG GAG CAC -30 (antisense) for murine Pgk1; 50- TCT GGA GAC GAT CTT ATG ATG TC -30 (sense) and 50- AGC TGG TCT TCA GTA AGG TCT GC -30 (antisense) for murine Pgam1; 50- GCG AAT TCC TGC CAG TAG CAT ATG CTT G -30 (sense) and 50- GGA AGC TTA GAG GAG CGA GCG ACC AAA GG -30 (antisense) for murine r18s. Formation of proper amplicons was confirmed by melting curve analysis. Data were standardized against the level of 18s rRNA.

EVALUATION OF PROTEIN SYNTHESIS BY METABOLIC LABELING Primary chondrocytes were grown to confluence in 24-well multiwell plates containing DMEM supplemented with 10% FBS. Thereafter, the medium was replaced with that containing 0.5% FBS. These cells were pre-incubated for 10 min at 37°C under 5% CO2 in air. Then, L-[4, 5-3H (N)]-leucine (37 MBq/ml, 1.48–2.22 TBq/mmol; PerkinElmer Japan, Yokohama, Japan) was added to the culture at a final concentration of 370 kBq/ml; and the cells were then incubated for 12 h at 37°C under 5% CO2 in air. After labeling, the medium was removed; and the cells were then washed with cold Dulbecco0s phosphate-buffered saline (PBS). The suspension of cells was thereafter solubilized by incubation for 1 h at 37°C with 0.5 M NaOH and subsequently neutralized with 5 M HCl, as previously described [Takigawa et al., 1980]. Part of the neutralized material was saved for total protein quantification by the bicinchoninic acid (BCA) method. The remainder was precipitated with trichloroacetic acid (TCA), and the radioactivity incorporated into the precipitates was determined by using a liquid scintillation counter, as previously described [Akeda et al., 2006].

AMINO ACID UPTAKE ASSAY Primary chondrocytes or HCS-2/8 cells were grown to confluence in 3.5 cm dishes containing DMEM supplemented with 10% FBS. Amino acid uptake was measured by following a previously described established protocol [Ellory et al., 1981; Barker et al., 1999]. Briefly, after a wash, the medium was replaced with isosmotic saline buffer (pH 7.4), and the cells were pre-incubated at 37°C under 5% CO2 in air. Then, the uptake reaction was started by the addition of a tracer amino acid and stopped by washing the cells rapidly twice with cold isotonic MgCl2 buffer on ice. To HCS- 2/8 cells, the mixture of the tracer amino acid and recombinant human CCN2 (BioVender Laboratory, Brno, Czech Republic) at a final concentration of 500 ng/ml was added. The uptake of L-[3H]- leucine (74 kBq/ml) at 190 mM or L-[3H]-lysine (74 kBq/ml) (PerkinElmer Japan) at 90 mM was measured for 10 min at 37°C under 5% CO2 in air. The cells were then lysed with lysis buffer (1% [w/v] Triton X-100, 20 mM Tris–HCl [pH 8.0], 25 mM NaCl, and 0.1 mM phenylmethylsulfonyl fluoride). The lysate was collected and centrifuged (15,000 rpm, 10 min, 4°C), after which the radioactivity of the supernatant was determined by using a liquid scintillation counter.

SODIUM DODECYL SULFATE POLYACRYLAMIDE GEL ELECTROPHORESIS (SDS–PAGE) AND SILVER STAINING Primary chondrocytes were grown to confluence in 3.5 cm dishes containing DMEM supplemented with 10% FBS. Total cellular proteins were prepared by lysing the cells in a radioimmunopreci- pitation analysis (RIPA) buffer (50 mM Tris–HCl, 0.15 M NaCl, 4 mM EDTA, 1% Nonidet P-40, 0.1% sodium deoxycholate) supplemented with 0.5% protease inhibitor cocktail (Sigma–Aldrich). One micro- gram of the proteins was separated by SDS–PAGE in a 10% polyacrylamide gel and stained with a commercial silver-staining reagent (Daiichi Pure Chemicals, Tokyo, Japan), following the manufacturer0s instructions.

AMINO ACID BINDING ASSAY Wells of an ELISA plate were coated with 10 mg/ml of rCCN2 or bovine serum albumin (BSA) in 50 mM NaHCO3 buffer (pH 9.6) at 4°C overnight. After being blocked with 100 ml of binding buffer (50 mM Tris/HCl [pH 7.4], 0.15 mM NaCl and 2% [w/v] BSA) for 2 h at 37°C, the wells were washed with a wash buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl). Then, 50 ml of L-[3H]-leucine or L-[3H]- lysine (74 kBq/well), which had been diluted with binding buffer, was added; and incubation was conducted for 4 h at 37°C. The wells were washed with wash buffer three times, and then the material left behind was lysed with 0.3 M NaOH. The lysate was incubated for 1 h at 37°C and neutralized with 6 M HCl. The radioactivity of the neutralized material was determined by using a liquid scintillation counter.

STATISTICAL ANALYSIS The results obtained from quantitative experiments were reported as the mean values � standard deviation (SD). Statistical comparisons between the groups were performed by using Student0s t-test. Unless otherwise specified, all experiments were repeated at least twice, and comparable results were obtained.

JOURNAL OF CELLULAR BIOCHEMISTRY CCN2 AND AMINO ACID METABOLISM 929

RESULTS

REDUCED LEVELS OF INTRACELLULAR ESSENTIAL AMINO ACIDS IN CCN2-NULL CHONDROCYTES Suspecting changes in amino acid metabolism caused by Ccn2 deficiency, we comparatively analyzed the intracellular amino acid levels of normal and Ccn2-null chondrocytes by metabolome analysis. According to the results of the initial evaluation, the levels of most amino acids in the Ccn2-null chondrocytes were reduced in comparison with those in the wild-type ones (Fig. 1). Interestingly, in spite of the overall decrease in the amino acid levels, cellular proline and glycine levels remained high in the Ccn2-null chondrocytes. Since these two amino acids are major constituents of collagen, which is one of the extracellular matrix components in cartilage, these findings suggest a relatively sufficient supply and inefficient consumption of the amino acids required for collagen synthesis. Among the protein-constituting amino acids examined, cysteine was the only one that was under the level of detection.

Comprehensive interpretation of the results of a second metabolome analysis further unveiled an interesting outcome owing to Ccn2 depletion. In Fig. 2, the relative levels of amino acids that showed more than a 25% decrease by Ccn2 deletion in the initial analysis are presented as mean values of two independent analyses from different animals. Here, it should be noted that Ccn2-null chondrocytes presented a stable reduction in all essential amino acids therein, whereas a few non-essential amino acids displayed large variance between the two analyses. These results indicate a significant role of CCN2 in essential amino acid metabolism in chondrocytes.

ENHANCEMENT OF AMINO ACID UPTAKE BY CCN2 DELETION IN CHONDROCYTES According to the results of metabolome analysis indicating a total reduction in the levels of essential amino acids due to CCN2 deficiency, we first anticipated attenuated amino acid uptake in Ccn2-null chondrocytes. Therefore, we evaluated the transport activity of free amino acids in normal and Ccn2-null chondrocytes. Utilizing a radiolabeled free leucine, which is one of the essential amino acids, we comparatively quantified the amino acid incorpo- rated into Ccn2-null and wild-type chondrocytes. However, unexpectedly, the uptake of the amino acids was rather enhanced in the Ccn2-null chondrocytes (Fig. 3A). This finding was also confirmed by the experiments with lysine, which is another essential amino acid with molecular properties different from those of leucine (Fig. 3B).

Additionally, in order to examine the integrity of the amino acid transport systems in Ccn2-null chondrocytes, we performed tran- scriptome analysis with a DNA microarray. Consistent with the results shown in Fig. 3A and B, the results of microarray analysis showed no striking effect on amino acid transporter genes by Ccn2 deletion (Supporting Information Fig. S1). For the ones with more than a 30% decrease in Ccn2-null chondrocytes, real-time quantitative PCR analysis was carried out to confirm these findings from the microarray analysis, but no significant difference was observed between Ccn2-null and wild-type chondrocytes in the expression levels of these genes (data not shown). Collectively, the

total reduction found in the levels of the essential amino acids in Ccn2-null chondrocytes (Fig. 2) may not be ascribed to their reduced uptake, but rather to accelerated consumption of these amino acids.

EFFECT OF ADDITION OF EXOGENOUS CCN2 ON AMINO ACID UPTAKE BY HUMAN CHONDROCYTIC CELLS The unexpected effect of Ccn2 depletion conversely suggested the possibility that CCN2 may have repressed the transport of essential amino acids into the cells. To examine this hypothesis, we next evaluatedtheeffectofexogenousCCN2ontheaminoacidmetabolism of chondrocytic cells. For this objective, we employed a human cell line, HCS-2/8, whose cells stably retain the chondrocytic phenotype. As expected from the result shown in Fig. 3A, the addition of exogenous CCN2 to HCS-2/8 cell cultures resulted in a significant reduction in amino acid uptake as determined by the incorporation of radiolabeled leucine (Fig. 3C). Together with the data in panels A and B, these data revealed a novel extracellular function of CCN2 in the uptake of essential amino acids, the mechanism of which was further investigated later on in this study.

ACCELERATED PROTEIN SYNTHESIS IN CCN2-NULL CHONDROCYTES The results above together indicated a remarkable acceleration of amino acid consumption caused by Ccn2 deficiency. Since a major part of free amino acids is utilized for protein synthesis, we suspected that protein synthesis might have been up-regulated in the Ccn2- null chondrocytes, a notion supported by the results of a previous study showing a remarkable increase in the expression of a number of ribosomal protein genes [Maeda-Uematsu et al., 2014]. Therefore, we pulse labeled the nascent proteins with radiolabeled leucine and evaluated the protein synthesis in normal and Ccn2-null chon- drocytes. Firstly, although we seeded the same number of the cells for each experiment, the total protein level was higher in the Ccn2-null chondrocytes than in the wild-type ones at the end of the experiment (Fig. 4A). Secondly, the radiolabeled amino acid used for protein synthesis during labeling period was also higher in amount in the Ccn2-null chondrocytes (Fig. 4B). Finally, even when the level of incorporated radiolabeled amino acid was standardized against the total protein level, a significant increase was still observed in the Ccn2-null chondrocytes (Fig. 4C). Thus, as expected, accelerated amino acid consumption due to heightened protein synthesis took place in the Ccn2-null chondrocytes.

MULTIPLE DEFICIENCIES IN GLYCOLYTIC ENZYME GENE EXPRESSION IN CCN2-NULL CHONDROCYTES Under an energy-shortage condition, cells also consume amino acids for biological energy production. Indeed, it was earlier clarified that Ccn2 deficiency leads to ATP deficiency, in which the expression of a glycolytic enzyme gene, Eno1, is repressed [Maeda-Uematsu et al., 2014]. Here, we newly found that, in addition to Eno1, the expression of two genes encoding two other glycolytic enzymes, Pgk1 and Pgam1, was also repressed in Ccn2-null chondrocytes to a comparable level. We made this discovery by reviewing the results of the microarray analysis (Fig. 5A), which were confirmed to be significant by quantitative real-time RT-PCR analysis (Fig. 5B). It should be noted that these three enzymes, including enolase 1, serially catalyze the glycolytic reaction from glyceraldehyde 3-

JOURNAL OF CELLULAR BIOCHEMISTRY930 CCN2 AND AMINO ACID METABOLISM

phosphate to phosphoenol pyruvic acid, resulting in the formation of pyruvic acid (Fig. 5C). Therefore, subsequent tricarboxylic acid cycle in collaboration with oxidative phosphorylation may have been in short of pyruvic acid supply from glycolysis in Ccn2-null

chondrocytes. Under such a condition, a significant amount of amino acids may well have been catabolized into metabolites that could be used for aerobic ATP production, which also would have led to decreased levels of intracellular amino acids.

Fig. 1. Metabolomic analysis of amino acids in normal and Ccn2-null chondrocytes. Evaluated amino acids are represented by three-letter codes at the top of each graph. Amino acid levels in wild-type (WT) and Ccn2-null (KO) chondrocytes are shown in pmol/1 million cells (ordinate). A representative set of data of two independent experiments is shown. N.D.: not detectable.

JOURNAL OF CELLULAR BIOCHEMISTRY CCN2 AND AMINO ACID METABOLISM 931

RETENTION OF HIGH-ENERGY ELECTRON-BEARING COENZYME LEVELS IN CCN2-NULL CHONDROCYTES In a previous study, it was unexpectedly shown that mitochondrial activity in Ccn2-null chondrocytes was as high as that in normal ones [Maeda-Uematsu et al., 2014]. This observation indicates that even in Ccn2-null chondrocytes, aerobic ATP production was as active as that in wild-type chondrocytes. In order to confirm this finding from a different aspect, we compared the level of nicotine amide adenine dinucleotide hydride (NADH), which supplies the highest energy-bearing electron to mitochondria for energy production through the respiratory chain. As shown in Fig. 6A, no significant difference was detected in the cellular NADH level between Ccn2-null and wild-type chondrocytes. In addition, NADPH level in the Ccn2-null chondrocytes was also comparable to that in the wild type (Fig. 6B). Moreover, NADH/NADþ and NADPH/NADPþ

ratios in Ccn2-null chondrocytes were no less than those in the wild type (Fig. 6C and D). These data suggest that, despite the impaired glycolysis, aerobic ATP production was still retained at a normal level, utilizing other metabolites instead of pyruvate originating from glycolysis.

CCN2 AS AN EXTRACELLULAR RETAINER OF FREE AMINO ACIDS The results in Fig. 3 show, although no substantial changes were observed in the gene expression levels of amino acid transporters, amino acid uptake depended on CCN2. Namely, CCN2 was found to interfere with amino acid uptake. Thus, we hypothesized that CCN2 might be sequestering amino acids in the microenvironment, which resulted in decreased availability of amino acids. To gain more insight into this possibility, we performed simple experiments in vitro. By a conventional solid-phase binding assay approach, we investigated whether CCN2 could bind to radiolabeled free amino acids. As a result, CCN2 coated on a plate was found to capture free leucine efficiently (Fig. 7A), and similar results were obtained with lysine (Fig. 7B). Taken together, extracellular CCN2 was suggested to capture different types of free amino acids, regulating their supply in a matricrine manner.

DISCUSSION

According to a number of studies in the past, CCN2 promotes both proliferation and differentiation of chondrocytes in vitro and in vivo [Nakanishi et al., 2000; Kubota and Takigawa, 2015], which is, at least in part, supported by the function of CCN2 to enhance energy metabolism. In fact, previous studies indicated that Ccn2-deficient growth plate cartilage exhibits disorganized structure with delayed endochondral ossification, which is accompanied by loss of ECM integrity and impaired chondrocyte proliferation [Ivkovic et al., 2003; Kawaki et al., 2008]. As a result of these changes, skeletal development is remarkably affected in Ccn2-null mice, which consequently leads to their lethal phenotype upon delivery. Nevertheless, the skeletal size of Ccn2-null mice is not significantly different from that of the wild type [Ivkovic et al., 2003]. Here, one should note that a long-term deficiency of a particular protein may lead to complex compensatory reactions in order to maintain homeostasis. Such reactions are associated with a number of genes that are up-regulated by the deficiency. Specifically, in the case of CCN2, up-regulation of several ribosomal protein genes by Ccn2 deletion was highly suspected to be this kind of compensatory reaction. Overproduction of ribosomal proteins leads to an increase in the activity and/or number of ribosome complexes, which structures serve as the basic intracellular apparatus required for enhanced protein synthesis. The decreased levels of essential amino acids, enhanced uptake of essential amino acids, and accelerated protein synthesis observed in Ccn2-null chondrocytes may be accounted for comprehensively from this point of view. Of note, addition of exogenous CCN2 to Ccn2-null chondrocytes did not slow down the accelerated protein synthesis (Supporting Information Fig. S2). This finding supports the idea that ribosomal activity increased by long-term CCN2 deficiency led to increased protein synthesis, and thus decreased amino acids in Ccn2-null chondro- cytes also may not be redeemed by transient addition of CCN2.

In this study, we showed that cellular levels of all of the essential amino acids were stably reduced in Ccn2-null

Fig. 2. Stable reduction in levels of essential amino acids in Ccn2-null chondrocytes. Relative amino acid levels in Ccn2-null (KO) chondrocytes versus wild-type (WT) ones are displayed. Mean values from two independent experiments are presented with error bars indicating standard deviations. Essential amino acids are underlined.

JOURNAL OF CELLULAR BIOCHEMISTRY932 CCN2 AND AMINO ACID METABOLISM

Fig. 3. Effects of Ccn2 depletion and CCN2 addition on amino acid uptake by chondrocytes. A and B: Leucine uptake by wild-type (WT) and Ccn2-null (KO) chondrocytes. Primary chondrocytes were grown to confluence in 3.5 cm dishes. After the cells had been pre-incubated in isosmotic saline buffer at 37°C under 5% CO2 in air, [

3H]-leucine (74 kBq/ml)(A) or [3H]-lysine (74 kBq/ml)(B) uptake was allowed to proceed for 10min at 37°C.Collected cell lysates were centrifuged, and radioactivity of the supernatants was determined. C: Leucine uptake by human chondrocytic HCS-2/8 cells in the absence (PBS) or presence of recombinant human CCN2 (rCCN2). After HCS-2/8 cells had reached confluence in 3.5 cm dishes, uptake of [3H]-leucine (74kBq/ml) in the presence of rCCN2 (500 ng/ml)orPBSwasstarted;and radioactivitywasdeterminedasdescribedfor panel A. Incorporated [3H]-leucine or lysine is presented as absolute radioactivity (Bq). Results of three independent samples are shown with error bars (standard deviations). Asterisks (�) and (��) indicate a statistical significance of difference versus the control at P < 0.05 and P < 0.01, respectively.

Fig. 4. Enhanced protein synthesis in Ccn2-null chondrocytes in culture. A: Total protein levels in wild-type (WT) and Ccn2-null (KO) chondrocytes at the time of evaluation. After primary chondrocytes had reached confluence in 24- well plates containing DMEM supplemented with 10% FBS, the medium was replaced with that with 0.5% FBS, and the cells were pre-incubated for 10 min at 37°C under 5% CO2 in air. Then, [

3H]-leucine (370 kBq/ml) was added to the culture. After incubation for 12 h, the cells were lysed in 0.5 M NaOH, neutralized with 5 M HCl; and then total protein amount was quantified by use of the BCA method. B: Levels of radiolabeled leucine incorporated into protein by WT and KO chondrocytes. Chondrocyte lysates were prepared as described for panel A, precipitated with TCA; and then radioactivity in the precipitate was measured. Incorporated [3H]-leucine is represented as molar number (pmol) per cell lysate of equal amount. C: Levels of protein-incorporated leucine standardized by the total protein in WT and KO chondrocytes. Relative amounts of incorporated [3H]-leucine (B) per total protein (A) are displayed. Results of three independent samples are shown with error bars (standard deviations). Asterisks (�) and (��) indicate a statistical significance of difference versus the control (WT) at P < 0.05 and <0.01, respectively.

JOURNAL OF CELLULAR BIOCHEMISTRY CCN2 AND AMINO ACID METABOLISM 933

chondrocytes. In contrast, we obtained an unexpected finding that proline and glycine, as major constituents of collagen, showed no difference between Ccn2-null and wild-type chondrocytes. Never- theless, considering the activated amino acid uptake together with the reduced production of collagens by Ccn2-null chondrocytes [Kawaki et al., 2008], the retention of proline and glycine in Ccn2- null chondrocytes at high levels is understandable. In addition, since proline and glycine are not essential amino acids, both can be biochemically synthesized inside the cells, which may have yielded the difference from the essential amino acids. Moreover, the unusual retention of these two amino acids after Ccn2 depletion may rather suggest a specific role of CCN2 in collagen synthesis.

As the observed up-regulation of amino acid uptake and reduced levels of essential amino acids appeared to be contradictory, we suspected accelerated amino acid consumption in Ccn2-null chondrocytes. Consequently, we found that both the level of total cellular proteins and protein synthesis evaluated as radiolabeled leucine incorporated in protein were rather enhanced in Ccn2-null chondrocytes. However, Ccn2-null growth plate cartilage is abnormal, being characterized by delayed endochondral ossification with reduced production of collagens; whereas the size of growing bones and growth plate is not smaller than that in the wild type [Ivkovic et al., 2003]. These findings suggest that some unknown protein may be overproduced and deposited to the ECM in place of collagens as one of the compensatory reactions. To examine this

Fig. 5. Reduction in the expression levels of multiple glycolytic enzymes. A: Reduction in expression of genes related to aerobic and anaerobic ATP production, as revealed by transcriptomic analysis. Relative gene expression levels in Ccn2-null chondrocytes (KO) versus those in wild-type ones (WT) are given. B: Results of quantitative real-time RT-PCR analysis confirming the reduction in the expression of glycolytic enzyme genes. Data from three independent samples are standardized against those of 18s rRNA, and mean values are presented with error bars denoting standard deviations. Asterisks (�) and (��) indicate a statistical significance of difference versus the control (WT) at P < 0.05 and <0.01, respectively. C: Metabolic steps catalyzed by the enzymes shown in panel B and enolase 1 during glycolysis.

JOURNAL OF CELLULAR BIOCHEMISTRY934 CCN2 AND AMINO ACID METABOLISM

hypothesis, we extracted total proteins from Ccn2-null chondrocytes and normal ones, separated them by SDS–PAGE, visualized them by silver staining for comparative analysis. We observed that several proteins were more abundantly produced in the Ccn2-null ones (data not shown), indicating excessive synthesis of specific proteins as a result of long-term CCN2 deficiency. Considering that increase in the incorporation of leucine into proteins was more drastic than that in the total protein level in Ccn2-null chondrocytes (Fig. 4), such proteins may contain leucine-rich ones. We strongly suspect the involvement of ribosomal proteins but believe that of ECM- associated proteins therein as well. Further detailed examinations on these proteins are currently underway. In addition to protein synthesis, amino acids can be consumed through other catabolic biochemical pathways to obtain bioenergy. This issue is especially critical for Ccn2-null chondrocytes, since such cells suffer from ATP shortage, which is further emphasized here by the present results indicating multiple down-regulation of glycolytic enzymes. Inter- estingly, a previous study revealed that mitochondrial activity in Ccn2-null chondrocytes showed no significant difference from that in normal ones, in spite of depression of pyruvic acid supply [Maeda- Uematsu et al., 2014]. This finding was also confirmed by the present data showing that the cellular levels of high-energy electron-bearing

coenzymes supporting mitochondrial membrane potential in Ccn2- null chondrocytes were comparable to those in the wild type. Respiratory chain activity is regulated by the coupled ATP synthesis. Thus, we highly suspect that the ATP shortage forces activation of the mitochondrial respiratory chain, which requires an energy source other than glycolytic products in Ccn2-null chondrocytes for maintaining aerobic catabolism. Since all of these amino acids can be eventually catabolized into metabolites processed through the tricarboxylic acid cycle, impaired glycolysis is also considered as a causative factor of decreased amino acid levels in Ccn2-null chondrocytes as well as enhanced protein synthesis. In this context, it is worth noting that the intracellular level of 2-oxoglutaric acid, which is a major intermediate in the tricarboxylic acid cycle and can be metabolized from a number of amino acids, is stably retained, compared to that of acetyl-CoA (Supporting Information Fig. S3).

Collectively, our present results suggest that Ccn2 deficiency induced a reduction in the levels of essential amino acids in chondrocytes through accelerated amino acid consumption by compensatory protein synthesis and amino acid catabolism (Fig. 8). However, Ccn2-null mice are characterized by impaired ECM integrity, which characteristic may appear contradictory to these results obtained in vitro. This apparent discrepancy may be ascribed

Fig. 6. Retention of the high-energy-bearing coenzymes in Ccn2-null chondrocytes. Cellular NADH content (A), NADPH content (B), NADH/NADþ ratio (C), and NADPH/NADPþ

ratio (D) in wild-type (WT) and Ccn2-null (KO) chondrocytes are shown. Mean values of two independent metabolome analyses are given with error bars (standard deviations).

JOURNAL OF CELLULAR BIOCHEMISTRY CCN2 AND AMINO ACID METABOLISM 935

to the altered microenvironment in cartilage caused by Ccn2 deletion. Since CCN2 is a matricellular protein, this molecule is dominantly present in the ECM around chondrocytes. In this context, we found an interesting molecular property of CCN2 as a retainer of amino acids, which is particularly important for ECM-embedded chondrocytes that survive in an avascular microenvironment with a limited amino acid supply. Although this finding may not be directly related to the intracellular events observed in Ccn2-null chondro- cytes in vitro, it provides a critical idea in understanding a role of CCN2 in microenvironment in vivo. In the absence of CCN2, a sufficient amount of free amino acids may not be stored in the ECM, and thus increased amino acid demand of Ccn2-null chondrocytes may not be fulfilled, resulting in further repression of ECM synthesis in vivo. Not only the biological significance of interactions of CCN2 with protein counterparts, but also that of those with small molecules such as amino acids, needs to be further characterized to clarify the entire molecular functions of CCN family proteins.

ACKNOWLEDGEMENTS The authors thank Drs. Kazumi Ohyama, Mitsuhiro Hoshijima, and Danilo Janune for their helpful suggestions, as well as Ms. Yoshiko Miyake for her secretarial assistance.

REFERENCES Abd El Kader T, Kubota S, Nishida T, Hattori T, Aoyama E, Janune D, Hara ES, Ono M, Tabata Y, Kuboki T, Takigawa M. 2014. The regenerative effects of CCN2 independent modules on chondrocytes in vitro and osteoarthritis models in vivo. Bone 59:180–188.

Akeda K, An HS, Pichika R, Attawia M, Thonar EJ, Lenz ME, Uchida A, Masuda K. 2006. Platelet-rich plasma (PRP) stimulates the extracellular matrix metabolism of porcine nucleus pulposus and anulus fibrosus cells cultured in alginate beads. Spine 31:959–966.

Babic AM, Chen CC, Lau LF. 1999. Fisp12/mouse connective tissue growth factor mediates endothelial cell adhesion and migration through integrin alphavbeta3, promotes endothelial cell survival, and induces angiogenesis in vivo. Mol Cell Biol 19:2958–2966.

Fig. 7. Retention of essential amino acids by CCN2. Direct binding of leucine (A) and lysine (B) was measured as radioactivity from labeled amino acids bound to CCN2 in vitro. Microplate wells were coated with rCCN2 or BSA (10 mg/ml). After blocking, the wells were incubated with [3H]-leucine or [3H]- lysine (74 kBq/well). After washes, determination of radioactivity was performed as described in the Materials and Methods section. The results of five independent samples are shown with error bars (standard deviations). Asterisks (�) and (��) indicate a statistical significance of difference versus the control (BSA) at P < 0.05 and <0.01, respectively.

Fig. 8. Schematic representation of amino acid metabolism in chondrocytes altered by Ccn2 depletion. Around normal chondrocytes, free essential amino acids (represented as small spheres) are captured by CCN2 in the ECM, which stores them and restricts their cellular uptake. In the absence of CCN2, increased availability of free amino acids enhances the incorporation of these amino acids, which are over-consumed by accelerated protein synthesis. At the same time, impaired glycolysis leads to ATP shortage and subsequent amino acid consumption for energy production. AA: amino acid.

JOURNAL OF CELLULAR BIOCHEMISTRY936 CCN2 AND AMINO ACID METABOLISM

Barker GA, Wilkins RJ, Golding S, Ellory JC. 1999. Neutral amino acid transport in bovine articular chondrocytes. J Physiol 514:795–808.

Bork P. 1993. The modular architecture of a new family of growth regulators related to connective tissue growth factor. FEBS Lett 327:125–130.

Brigstock DR, Goldschmeding R, Katsube KI, Lam SC, Lau LF, Lyons K, Naus C, Perbal B, Riser B, Takigawa M, Yeger H. 2003. Proposal for a unified CCN nomenclature. Mol Pathol 56:127–128.

Charrier A, Brigstock DR. 2013. Regulation of pancreatic function by connective tissue growth factor (CTGF, CCN2). Cytokine Growth Factor Rev 24:59–68.

Ellory JC, Jones SEM, Young JD. 1981. Glycine transport in human erythrocytes. J Physiol 320:403–422.

Hattori T, M€uller C, Gebhard S, Bauer E, Pausch F, Schlund B, B€osl MR, Hess A, Surmann-SchmittC,vonderMarkH,deCrombruggheB,vonderMarkK.2010. SOX9 is a major negative regulator of cartilage vascularization, bone marrow formation and endochondral ossification. Development 137:901–911.

Ivkovic S, Yoon BS, Popoff SN, Safadi FF, Libuda DE, Stephenson RC, Daluiski A, Lyons KM. 2003. Connective tissue growth factor coordinates chondrogenesis and angiogenesis during skeletal development. Development 130:2779–2791.

Jun JI, Lau LF. 2011. Taking aim at the extracellular matrix: CCN proteins as emerging therapeutic targets. Nat Rev Drug Discov 10:9459–9463.

Kawaki H, Kubota S, Suzuki A, Lazar N, Yamada T, Matsumura T, Ohgawara T, Maeda T, Perbal B, Lyons KM, Takigawa M. 2008. Cooperative regulation of chondrocyte differentiation by CCN2 and CCN3 shown by a comprehensive analysis of the CCN family proteins in cartilage.J Bone Miner Res23:1751–1764.

Kawata K, Kubota S, Eguchi T, Aoyama E, Moritani NH, Kondo S, Nishida T, Takigawa M. 2012. Role of LRP1 in transport of CCN2 protein in chondrocytes. J Cell Sci 125:2965–2972.

Kubota S, Takigawa M. 2007. CCN family proteins and angiogenesis: From embryo to adulthood. Angiogenesis 10:1–11.

Kubota S, Takigawa M. 2012. CCN. In: Choi S, editor. Encyclopedia of signaling molecules. Dordrecht, Netherlands: Springer. pp 273–281.

Kubota S, Takigawa M. 2013. The CCN family acting throughout the body: Recent research developments. Biomol Concepts 4:477–494.

Kubota S, Takigawa M. 2015. Cellular and molecular actions of CCN2/CTGF and its role under physiological and pathological conditions. Clin Sci 128:181–196.

Leask A, Abraham DJ. 2006. All in the CCN family: Essential matricellular signaling modulators emerge from the bunker. J Cell Sci 119:4803–4810.

Maeda-Uematsu A, Kubota S, Kawaki H, Kawata K, Miyake Y, Hattori T, Nishida T, Moritani N, Lyons KM, Iida S, Takigawa M. 2014. CCN2 as a novel molecule supporting energy metabolism of chondrocytes. J Cell Biochem 115:854–865.

Nakanishi T, Nishida T, Shimo T, Kobayashi K, Kubo T, Tamatani T, Tezuka K, Takigawa M. 2000. Effects of CTGF/Hcs24, a product of a hypertrophic chondrocyte-specific gene, on the proliferation and differentiation of chondrocytes in culture. Endocrinology 141:264–273.

Nishida T, Nakanishi T, Asano M, Shimo T, Takigawa M. 2000. Effects of CTGF/Hcs24, a hypertrophic chondrocyte-specific gene product, on the proliferation and differentiation of osteoblastic cell in vitro. J Cell Physiol 184:197–206.

Nishida T, Kubota S, Kojima S, Kuboki T, Nakao K, Kushibiki T, Tabata Y, Takigawa M. 2004. Regeneration of defects in articular cartilage in rat knee joints by CCN2 (connective tissue growth factor). J Bone Miner Res 19:1308–1319.

Perbal B, Takigawa M. 2005. CCN protein—A new family of cell growth and differentiation regulators. London: Imperial College Press. p. 311.

Safadi FF, Xu J, Smock SL, Kanaan RA, Selim AH, Odgren PR, Marks SC, Jr., Owen TA, Popoff SN. 2003. Expression of connective tissue growth factor in bone: Its role in osteoblast proliferation and differentiation in vitro and bone formation in vivo. J Cell Physiol 196:51–62.

Shimo T, Nakanishi T, Nishida T, Asano M, Kanyama M, Kuboki T, Tamatani T, Tezuka K, Takemura M, Matsumura T, Takigawa M. 1999. Connective tissue growth factor induces the proliferation, migration, and tube formation of vascular endothelial cells in vitro, and angiogenesis in vivo. J Biochem 126:137–145.

Shimo T, Nakanishi T, Nishida T, Asano M, Sasaki A, Kanyama M, Kuboki T, Matsumura T, Takigawa M. 2001. Involvement of CTGF, a hypertrophic chondrocyte-specific gene product, in tumor angiogenesis. Oncology 61:315–322.

Smerdel-Ramoya A, Zanotti S, Deregowski V, Canalis E. 2008. Connective tissue growth factor enhances osteoblastogenesis in vitro. J Biol Chem 283:22690–22699.

Takigawa M, Ishida H, Takano T, Suzuki F. 1980. Polyamine and differentiation: Induction of ornithine decarboxylase by parathyroid hormone is a good marker of differentiated chondrocytes. Proc Natl Acad Sci 77:1481–1485.

Takigawa M, Tajima K, Pan HO, Enomoto M, Kinoshita A, Suzuki F, Takano Y, Mori Y. 1989. Establishment of a clonal human chondrosarcoma cell line with cartilage phenotypes. Cancer Res 49:3996–4002.

Takigawa M. 2013. CCN2: A master regulator of the genesis of bone and cartilage. J Cell Commun Signal 7:191–201.

SUPPORTING INFORMATION

Additional supporting information may be found in the online version of this article at the publisher0s web-site.

JOURNAL OF CELLULAR BIOCHEMISTRY CCN2 AND AMINO ACID METABOLISM 937