Bone 74 (2015) 125–133
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Original Full Length Article
AMP-activated protein kinase (AMPK) activity negatively regulates chondrogenic differentiation Kenjiro Bandow, Joji Kusuyama, Kyoko Kakimoto, Tomokazu Ohnishi, Tetsuya Matsuguchi ⁎ Department of Oral Biochemistry, Field of Developmental Medicine, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
a r t i c l e
i n f o
Article history: Received 1 May 2014 Revised 2 December 2014 Accepted 3 December 2014 Available online 10 December 2014 Edited by: Regis O'Keefe Keywords: Chondrocyte AMPK Metformin Sox9 Egr-1
a b s t r a c t Chondrocytes are derived from mesenchymal stem cells, and play an important role in cartilage formation. Sex determining region Y box (Sox) family transcription factors are essential for chondrogenic differentiation, whereas the intracellular signal pathways of Sox activation have not been clearly elucidated. AMP-activated protein kinase (AMPK) is a serine–threonine kinase generally regarded as a key regulator of cellular energy homeostasis. It is known that the catalytic alpha subunit of AMPK is activated by upstream AMPK kinases (AMPKKs) including liver kinase B1 (LKB1). We have previously reported that AMPK is a negative regulator of osteoblastic differentiation. Here, we have explored the role of AMPK in chondrogenic differentiation using in vitro culture models. The phosphorylation level of the catalytic AMPK alpha subunit significantly decreased during chondrogenic differentiation of primary chondrocyte precursors as well as ATDC-5, a well-characterized chondrogenic cell line. Treatment with metformin, an activator of AMPK, significantly reduced cartilage matrix formation and inhibited gene expression of sox6, sox9, col2a1 and aggrecan core protein (acp). Thus, chondrocyte differentiation is functionally associated with decreased AMPK activity. © 2015 Elsevier Inc. All rights reserved.
Introduction AMP-activated protein kinase (AMPK) is a serine–threonine kinase, which maintains the balance between production and consumption of ATP in eukaryotic cells [1]. AMPK is a heterotrimeric enzyme composed of the catalytic α subunit and regulatory β and γ subunits in a 1:1:1 stoichiometric ratio [2]. AMPK is activated by the elevation of intracellular AMP/ATP ratio caused by energy restriction. Subsequently, ATP-generating pathways are activated, while ATP-consuming mechanisms are inhibited, thereby restoring the normal cellular AMP/ATP ratio. Thus, AMPK is generally considered as an essential regulator of energy homeostasis of cells, and termed as a metabolic “energy sensor” of the biological system. The activation of AMPK is induced by phosphorylation of the catalytic α subunit [3]. The most characterized upstream AMPK kinase (AMPKK) is liver kinase B1 (LKB1), which is a reported target of metformin, a type 2 diabetes drug [4]. Recent reports have revealed that AMPK has non-metabolic functions as well. In vivo studies of Drosophila have demonstrated that AMPK is essentially involved in cell polarity and mitosis [5]. On the other hand, activation of AMPK has been proposed as one of the regulatory mechanisms of mammal longevity [6]. Notably, several lines of evidence have indicated the involvement of AMPK in the regulation of cellular differentiation. Activation of AMPK has been suggested to be inhibitory to the differentiation of adipocytes [7,8], myoblasts [9], and ⁎ Corresponding author. Fax: +81 99 275 6138. E-mail address: [email protected] (T. Matsuguchi).
http://dx.doi.org/10.1016/j.bone.2014.12.001 8756-3282/© 2015 Elsevier Inc. All rights reserved.
osteoblasts [10]. In contrast, it has also been reported that AMPK is promotive of the differentiation of endothelial progenitor cells [11]. In the earliest event of chondrogenesis, recruited mesenchymal cells proliferate and aggregate. Subsequently, the cells mature into chondrocytes, producing cartilage matrix proteins. After differentiation into hypertrophic cells, they are finally replaced by bone tissue [12]. Deregulation of chondrocyte differentiation has been described in various chronic skeletal diseases including osteoarthritis (OA) [13–15]. Chondrocytes produce a characteristic cartilage matrix that consists of collagen and proteoglycan [16]. During mouse embryogenesis, a transcription factor, Sox9, is expressed in all chondrocytes and their progenitor cells [12]. Heterozygous sox9 mutants exhibited delayed chondrogenic mesenchymal condensation and enlargement of the hypertrophic zone [17]. Sox9 regulates col2a1 gene, which encodes one component of collagen type II, in the early stage of chondrogenesis [18]. Two other Sox family transcription factors, Sox5 and Sox6, are co-expressed with Sox9 [19] and are required for the expression of aggrecan, which is a cartilage-specific proteoglycan [20]. Inactivation of Sox9 before mesenchymal condensation results in the absence of Sox5 and Sox6, as well as other chondrogenic marker gene expression [21], indicating that Sox9 is the essential master regulator of chondrocyte differentiation. In this present study, we have explored the role of AMPK in chondrogenic differentiation by in vitro chondrocyte differentiation models. It was found that the phosphorylation level of AMPKα was progressively decreased during chondrogenic differentiation. Stimulation with metformin during chondrogenic differentiation significantly
126
K. Bandow et al. / Bone 74 (2015) 125–133
decreased gene expression of sox9 and sox6 along with other chondrogenic differentiation markers including col2a1, and aggrecan core protein (acp). Conversely, knock-down of AMPKα expression by siRNA increased sox9, col2a1, and acp mRNA Thus, our present data indicate that differentiation of chondrocytes is functionally associated with decreased AMPK activity.
Western blot analysis For Western blot analyses, total cellular lysate preparation and immunoblotting procedures were performed as previously described [27]. Quantitative polymerase chain reaction (qPCR) analysis
Materials and methods Reagents and antibodies Sodium selenite, transferrin human, metformin, sodium salicylate, 5-aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR) and insulin from bovine pancreas were purchased from Sigma-Aldrich (St. Louis, MO). Antibodies specifically recognizing AMPKα1/2, AMPKβ1/2, phospho-AMPKα (Thr172), and EGR1 were purchased from Cell Signaling Technology (Danvers, MA). Antibodies against AMPKγ1/ 2/3 and β-Actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
Monolayer and micro-mass culture ATDC-5, a well-characterized chondrogenic cell line derived from mouse teratocarcinoma [22], was obtained from Riken Cell Bank (Tsukuba, Japan), and maintained in Dulbecco's modified Eagle medium/Ham's F-12 (DMEM/F-12) (Wako Pure Chemical Industries, Osaka, Japan) containing 5% fetal bovine serum (FBS), 50 U/ml penicillin, and 50 mg/ml streptomycin. The cells were induced to differentiate by IST (1.7 mM insulin, 10 nM sodium selenite, and 125 μM transferrin) for 4 weeks. C57BL/6 mice (wild-type) were obtained from Japan Clea (Tokyo, Japan). The animals were maintained in accordance with protocols approved by the Animal Care and Use Committee at Kagoshima University, Japan. Primary chondrogenic progenitor cells were isolated from murine limb buds at embryonic day 10.5 (E10.5) and cultured as previously described [23].
Transfection of small interference RNA (siRNA) Chemically synthesized siRNAs, purified by HPLC, were purchased from Sigma-Aldrich (MO, USA). The following sequences were synthesized as prkaa1 siRNA: sense sequence: 5′-GAU GCA AUA AGC AUG GAU ATT-3′ and anti-sense sequence: 5′-UAU GCA UGC UUA UUG CAU CTT-3′. Non-targeting control siRNA duplexes (Control siRNA-A) were purchased from Santa Cruz Biotechnology. ATDC-5 cells were seeded in a well of a six-well plate at a density of 5.0 × 104 cells, and then cultured for 12 h before each experiment. The duplex siRNA was transfected using HilyMax reagent (Dojindo Laboratories, Kumamoto, Japan), according to the manufacturer's instructions.
Measurement of cell viability Cell viability was measured by 3-(4,5-di-methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, yellow tetrazole (MTT) assay as previously described [24].
Alcian blue staining Chondrogenic matrix was visualized by Alcian blue staining as previously described [25]. The stained area was photographically measured by ImageJ program [26].
The isolation of total RNA and qPCR was conducted as previously described [27]. The primer sequences were described in Table 1. Two housekeeping genes, ribosomal protein L13a (rpl13a) and TATA box binding protein (tbp), were used as endogenous controls in the qPCR. Luciferase reporter gene assay The promoter regions of mouse sox9 (− 953/+ 19) and sox6 (− 842/+66) genes containing the putative transcriptional initiation sites were cloned by PCR from C57BL/6 genomic DNA using pairs of primers, 5′-TCA ACC CCG GAG TAG TTT TG-3′ and 5′-CTC TCC GAC TTC CAG CTC AG-3′ for sox9 and 5′-CTG CTA CAG GCT CCT TTG CT-3′ and 5′-TCA CAT TGG CAA CTG CCT TA-3′ for sox6, and ligated into the luciferase reporter vector pGL4.19 (Promega, WI) to generate pGL-Sox9953 + 19 and pGL-Sox6-842 + 66, respectively. Deletion (pGL-Sox9457 + 19, pGL-Sox9-207 + 19, and pGL-Sox6-447 + 66) and mutation (pGL-Sox9-953 + 19mt) constructs were created by methods using PCR as previously described [28]. The primer sequences were described in Table 2. Each of the reporter plasmid or the vector control, pGL4.19, was transfected into 70% confluent ATDC-5 cells using HilyMax reagent, according the manufacturer's instruction. At 24 h after transfection, the
Table 1 Primers used in this study. Gene symbol prkaa1
Primers [5′–3′]
TGAGAACGTCCTGCTTGATG ATAATTGGGTGAGCCACAGC prkaa2 CCCCTGGTCTCTGTCTTTCT GCTGGATTTCTGTGTTTTCG prkab1 TCCGATGTGTCTGAGCTGTC CCCGTGTCCTTGTTCAAGAT prkab2 GAAGGAGGCAAGGAGGTCTT ACTGATGCTCTCCCTCTGGA prkag1 CTCCGCCTTACCTGTAGTGG AAGTAGTGGGACCGATGCTG prkag2 GTCCATCGGCTGGTTGTAGT TCTCCTTCTGTTTGGCACCT prkag3 CTCCCAATGACAGCCTGTTT GGAACTTGAGTAGCCGCTTG sox6 ATCTCCCACCCAGAACCTCT CAGGGCAGGAGAGTTGAGAC sox9 CCATGTGGCCAGCAGATG TTTTAGCACATGGGATGTCTTG AA col2a1 AGAACAGCATCGCCTACCTG CTTGCCCCACTTACCAGTGT col9a1 CAGGATTGGCCAAGATGACT TTCCCAGCTTGTAAGCCACT col10a1 CATCTCCCAGCACCAGAATC GTGTCTTGGGGCTAGCAAGT acp AACTTCTTTGCCACCGGAGA GGTGCCCTTTTTACACGTGAA alp TCCTG CCAAAAACCTCAAAGG TGCTTCATGCAGAGCCTGC mmp13 GCCATTTCATGCTTCCTGAT CTCTGGTGTTTTGGGATGCT rpl13a GCTTACCTGGGGCGTCTG ACATTCTTTTCTGCCTGTTTCC tbp CAAACCCAGAATTGTTCTCCTT ATGTGGTCTTCCTGAATCCCT
GenBank
Amplication length
NM_001013367 112 bp (467–578 nt) NM_178143
159 bp (1767–1925 nt)
NM_031869
140 bp (699–838 nt)
NM_182997
131 bp (433–553 nt)
NM_016781
113 bp (742–885 nt)
NM_145401
154 bp (2094–2205 nt)
NM_153744
129 bp (968–1096 nt)
NM_011445
145 bp (1492–1636 nt)
NM_011448
99 bp (3049–3147 nt)
NM_031163
161 bp (4412–4572 nt)
NM_007740
161 bp (386–516 nt)
NM_009925
152 bp (48–199 nt)
NM_007424
111 bp (6161–6271 nt)
NM_007431
101 bp (1138–1238 nt)
NM_008607
131 bp (793–890 nt)
NM_009438
149 bp (437–585 nt)
NM_013684
131 bp (876–1006 nt)
K. Bandow et al. / Bone 74 (2015) 125–133 Table 2 Primers for luciferase reporter vectors.
sox6 promoter −842 + 66 −447 + 66 sox9 promoter −953 + 19 −457 + 19 −207 + 19 For mutagenesis −953 + 219mt −228mt + 19
Forward primers [5′–3′]
Reverse primers [5′–3′]
GGTACCTGCTACAGG CTCCTTTGCT GGTACCGCCAGGACA GTTGGTTGTTT
AGATCTCACATTGGCA ACTGCCTTA AGATCTCACATTGGCA ACTGCCTTA
CTCGAGTCAACCCCG GAGTAGTTTTG GGTACCACGGAGAC AGCATC GGTACCCCTCACCCC ACCATCCA C
AGATCTCTCCGACTTC CAGCTCAG AGATCTCTCCGACTTC CAGCTCAG AGATCTCTCCGACTTC CAGCTCAG
CTCGAGTCAACCCCG GAGTAGTTTTG GCCTTTCGCCCCCCT CCAAG
GGCGAAAGGCAAGC GGGGGA AGATCTCTCCGACTT CCAGCTCAG
127
differentiation of ATDC-5 cells (Figs. 1A and B). On the other hand, the phosphorylation level of the AMPKα subunit was significantly decreased during differentiation in ATDC-5 cells (Fig. 1A). Metformin inhibited differentiation of ATDC-5 cells
cells were treated with IST and metformin, followed by further incubation for 48 h. Luciferase activities were measured and normalized by protein contents of the cell lysates.
Results The phosphorylation level of AMPKα decreased during chondrogenic differentiation In order to assess AMPK activity during the course of chondrocyte differentiation, we analyzed phosphorylation of AMPKα and the expression levels of AMPK α/β/γ subunits during chondrogenic differentiation by Western blot and qPCR analyses. ATDC-5 cells were induced to differentiate in the presence of IST, and total cell lysates and RNA were isolated. It was found that the protein levels of AMPKα, β, and γ subunits and mRNA expression levels of prkaa, prkab, and prkag (respectively coding AMPKα, β, and γ subunits) remained constant during the chondrogenic
In order to investigate how forced activation of AMPK affects chondrocyte differentiation, we sought to use metformin and AICAR, known activators of AMPK. We first examined the dose-dependent effect of metformin on the phosphorylation level of AMPKα in ATDC-5 cells. As expected, significant phosphorylation of AMPKα was induced by metformin (Fig. 2A). The maximum phosphorylation was observed at 2 mM of metformin. On the other hand, the cell viability of ATDC-5 cells was not significantly affected by metformin at up to 2 mM, but was decreased at 5 mM metformin (Fig. 2B). Unlike metformin, AICAR failed to induce sufficient phosphorylation of AMPKα in ATDC-5 cells at non-toxic concentrations (Figs. 2C, D). We thus evaluated the effects of 2 mM metformin treatment on the cartilage matrix synthesis by ATDC-5 cells. The matrix synthesis induced by IST was significantly inhibited by 2 mM metformin (Figs. 3A and B), indicating that the forced AMPKα activation is correlated with the inhibition of matrix synthesis in ATDC-5 cells. We next examined the effect of metformin on the expression of chondrogenic marker genes by qPCR analyses. In ATDC-5 cells, induction of acp and col2a1 mRNA was inhibited by 2 mM metformin (Fig. 3C). Since Sox family transcription factors are master regulators of chondrogenic differentiation controlling acp and col2a1 mRNA expression, we next analyzed the effect of metformin on the gene expression of sox9 and sox6. In the absence of metformin, the expression level of sox9 mRNA increased approximately 10 fold in the first 2 weeks of differentiation, and started to decrease from week 3 (Fig. 3C). On the other hand, the expression of sox6 mRNA gradually increased and remained maximally elevated after week 2 (Fig. 3C). The addition of metformin significantly decreased the levels of both sox9 and sox6 mRNA from week 1 to week 4 (Fig. 3C). We further analyzed the effect of metformin on the expression of four hypertrophic chondrocyte marker genes: col9a1, col10a1, alkaline phosphatase (alp), and matrix metalloproteinase 13 (mmp13). The addition of metformin significantly repressed the induction processes for these four hypertrophic markers (Fig. 3D).
Fig. 1. Expression and phosphorylation of AMPK subunits during chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST. A: Cytoplasmic lysates were isolated at the indicated weeks. The same amount of cell lysate (15 μg protein/lane) was separated in each lane by SDS-PAGE. Expression and phosphorylation of AMPK were detected with the indicated specific antibodies. B: Total RNA was isolated at the indicated week and the expression level of the indicated mRNA was analyzed by qPCR. Results of qPCR are shown for prkaa1, a2, b1, b2, g1, g2, and g3. Fold increase represents an experimental value divided by the control (prkaa1, b1, or g2 at week 0) value for each. Vertical bars indicate mean S.D. of at least three different experiments.
128
K. Bandow et al. / Bone 74 (2015) 125–133
Fig. 2. Effects of metformin and AICAR on AMPKα phosphorylation in ATDC-5 cells. A: Cells were cultured with varying concentrations of metformin in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. B: The viabilities of ATDC-5 cells after 48 h of treatment with various concentrations of metformin were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. C: Cells were cultured with varying concentrations of AICAR or 2 mM metformin in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. D: The viabilities of ATDC-5 cells after 48 h of treatment with various concentrations of AICAR or 2 mM metformin were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
Metformin suppressed chondrogenic differentiation of mouse embryonic limb cells We next explored the effects of metformin on chondrogenic differentiation of mouse embryonic limb cells. The matrix synthesis induced
by IST was moderately inhibited by the addition of 2 mM metformin (Figs. 4A and B). Consistently, induction of acp and col2a1 mRNA in mouse embryonic limb cells was suppressed by about 30% by 2 mM metformin (Fig. 4C). We subsequently analyzed the effect of metformin on the gene expression of sox9 and sox6. In the absence of metformin,
Fig. 3. Metformin inhibited chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST for the indicated weeks and were stained by Alcian blue (A). Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown (B). ATDC-5 cells were induced to differentiate by IST with or without 2 mM metformin. Total RNA was isolated at the indicated time and reverse-transcripted. Quantitative PCR analyses were performed for sox9, sox6, col2a1, acp (C), and col9a1, col10a1, alp, and mmp13 (D). Fold increase represents an experimental value divided by the control (week 0: sox9, sox6, col2a1, acp, alp, and mmp13; week 1: col9a1; week 2: col10a1) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
K. Bandow et al. / Bone 74 (2015) 125–133
the expression levels of sox9 and sox6 mRNA temporarily increased at day 2 and decreased by day 5 (Fig. 4C). The addition of metformin moderately decreased the expression level of sox9 and sox6 mRNA from day 2 to day 5 (Fig. 4C). High glucose concentration promoted differentiation of ATDC-5 cells In order to investigate how decreased AMPK activity affects chondrocyte differentiation, we examined the effects of a high-glucose culture condition on the ATDC-5 differentiation. We first examined the dose-dependent effect of glucose on the phosphorylation level of AMPKα in ATDC-5 cells. As expected, phosphorylation of AMPKα was inhibited by glucose in a concentration-dependent manner, reaching the maximum inhibition at 15 mM of glucose (Fig. 5A). We thus evaluated the effects of 15 mM glucose treatment on the cartilage matrix synthesis by ATDC-5 cells, and found that the matrix synthesis induced by IST was significantly promoted by 15 mM glucose (Figs. 5B and C). This result has indicated that forced AMPKα inhibition by high glucose concentration may promote cartilage matrix synthesis of ATDC-5 cells. We next examined the effect of high-concentration glucose on the expression of chondrogenic marker genes by qPCR analyses. In ATDC-5 cells, induction of sox9, sox6, acp and col2a1 mRNA was promoted by 15 mM glucose (Fig. 5D). We further analyzed the effect of high
129
glucose concentration on the gene expression of col9a1 and col10a1, chondrogenic marker genes at the late differentiation stage. The mRNA of col9a1, and col10a1 became detectable from week 2 and week 3 of differentiation, respectively (Fig. 5D). The high glucose condition significantly promoted the induction processes of these two late differentiation markers (Fig. 5D).
AMPKα knock-down by siRNA promoted chondrogenic differentiation of ATDC-5 cells For another experiment to examine the effect of reduced AMPK activity on chondrogenic differentiation, we introduced prkaa1 siRNA into ATDC-5 cells. The protein expression level of AMPKα was partially but significantly inhibited by prkaa1 siRNA compared with control siRNA at day 7 (Fig. 6A). Modest mRNA upregulation of sox9, col2a1, and acp, but not sox6, was observed by the transduction of prkaa1 siRNA at day 7 of chondrogenic differentiation (Fig. 6B). The reason for the failed increase of sox6 mRNA is not clear. It is possible that the required threshold level is relatively low for sox6 and the partial inhibitory effect of prkaa1 siRNA (Fig. 6A) was not sufficient. It was confirmed that the partial AMPKα knock-down by siRNA did not affect cell viability and cellular metabolism (Fig. 6C).
Fig. 4. Mouse primary chondrocyte precursor cells from E10.5 embryo limb buds were cultured in micro-mass method with or without 2 mM metformin for 5 days. The cells were stained with Alcian blue (A). Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown (B). Total RNA was collected and qPCR analyses were performed for sox9, sox6, col2a1, and acp (C). Fold increase represents an experimental value divided by the control (day 0) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
130
K. Bandow et al. / Bone 74 (2015) 125–133
Fig. 5. High-concentration glucose promoted chondrogenic differentiation. A: ATDC-5 cells were cultured with varying concentrations of glucose in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. B: ATDC-5 cells were induced to differentiate by IST for the indicated weeks in the presence of 5 mM (Control) or 15 mM glucose (High Glc) and were stained by Alcian blue. C: Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown. D: ATDC-5 cells were induced to differentiate by IST with 5 mM (black bar) or 15 mM glucose (gray bar). Total RNA was isolated at the indicated time and qPCR analyses were performed for sox9, sox6, col2a1, acp, col9a1 and col10a1. Fold increase represents an experimental value divided by the control (week 0: sox9, sox6, col2a1, and acp; week 1: col9a1; week 2: col10a1) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
Inhibitory effects of metformin on the IST-induced transcriptional activation of sox9 The 5′ upstream regions of the Sox9 and Sox6 genes are highly conserved between human and mouse, and the crucial regions for chondrogenic differentiation were found in −202/−128 and −517/−1 upstream of the transcriptional initiation sites in human Sox9 and Sox6 genes, respectively [28,29]. To examine the effect of metformin on the transcriptional activities of sox9 and sox6, we made the luciferase reporter plasmids containing different lengths of the 5′ upstream regions of mouse sox9 and sox6 genes. IST-treatment significantly promoted
transcriptional activities of three (− 953, − 457, and − 207) sox9 and two (−842 and −447) sox6 reporter constructs (Fig. 7A). These results indicated that the essential regulatory regions for sox9 and sox6 genes during chondrogenic differentiation were located within the 207 and 447 nucleotide lengths upstream of their transcriptional initiation sites, respectively. As for sox9, treatment with metformin significantly inhibited the IST-induced transcriptional activations of the −953 and −457 promoter constructs, but failed to reduce the promoter activity of the − 207 construct (Fig. 7A), indicating that a negative effect of metformin was mediated by the 5′-upstream region between − 457 and − 207. On
K. Bandow et al. / Bone 74 (2015) 125–133
131
Fig. 6. AMPKα knock-down by siRNA promoted chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST with 200 μM of control siRNA or prkaa1 siRNA. A: The inhibitory effect of prkaa1 siRNA on AMPKα protein expression was confirmed by Western blotting on day 7. B: Quantitative PCR analyses were performed for sox9, sox6, col2a1, and acp mRNAs on day 7. Fold increase represents an experimental value divided by the control (control siRNA-treatment) value for each. Vertical bars indicate mean SD of at least three different experiments. C: The viabilities of ATDC-5 cells after 48 h of treatment with siRNA were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
the other hand, metformin did not affect the IST-induced transcriptional activation of the sox6 promoter constructs (Fig. 7A). When we analyzed the nucleotide sequences between −457 and −207 in the sox9 gene, we
found the GC-rich putative binding motif for Early growth response-1 (Egr-1) transcriptional repressor protein [30,31] in − 228/− 219. We then constructed pGL-Sox9-953 + 19mt plasmid with three nucleotide
Fig. 7. Metformin inhibits IST-induced transcriptional activation of sox9 gene in ATDC-5 cells. A: ATDC-5 cells, transiently transfected with pGL-Sox9-953 + 19, pGL-Sox9-457 + 19, pGLSox9-207 + 19, pGL-Sox6-842 + 66, pGL-Sox6-447 + 66, or pGL4.19, were stimulated with IST for 48 h. Cells were lysed for luciferase activity measurements as described in Material and methods. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the IST-treated group by Bonferroni test (p b 0.01). B: Nucleotide sequence of the 20 bp element (−233/−214) containing the putative Egr-1 biding motif (−228/−219, underlined). Three-base mutation was introduced at −225/−223 (black boxes) within the Egr-1 motif (mt). ATDC-5 cells, transiently transfected with pGL-Sox9-953 + 19, pGL-Sox9-953 + 19mt, or pGL4.19, were lysed for luciferase activity measurements after IST treatment with 2 mM metformin, 0.5 mM AICAR, 10 mM SA, or vehicle for 48 h. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the IST-treated group by Bonferroni test (p b 0.01). C: Cells were cultured with 10 mM SA in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. D: ATDC-5 cells were cultured with or without 2 mM metformin. Cytoplasmic lysates were isolated at the indicated days. The same amount of cell lysate (15 μg protein/lane) was separated in each lane by SDS-PAGE. Expressions of Egr-1 and β-actin were detected with the indicated specific antibodies.
132
K. Bandow et al. / Bone 74 (2015) 125–133
substitutions (TTT for CCG) in −225/−223 and found that the inhibitory effect of metformin on the IST-induced transcriptional activity was abrogated when the Egr-1 binding site was mutated (Fig. 7B). To confirm that the inhibitory effect of metformin is mediated through AMPK activation, we further examined the effects of AICAR and salicylate (SA) on the transcriptional activity of the sox9 promoter. SA was reported to directly phosphorylate AMPK [32], and we confirmed that SA could induce AMPK phosphorylation in ATDC-5 cells (Fig. 7C). We found that both AICAR and SA inhibited the transcriptional activity of sox9 promoter in a manner dependent on the Egr-1 binding site, similar to metformin (Fig. 7B). Consistently, the protein expression level of Egr-1 was increased by metformin in ATDC-5 cells (Fig. 7D). Taken together, these results have indicated that AMPK activity inhibits transcriptional activation of sox9 during chondrogenic differentiation by inducing a transcriptional repressor, Egr-1. Discussion Previous reports have demonstrated that cellular AMPK activity is functionally associated with differentiation of several cell types. More specifically, activation of AMPK has been reported to be inhibitory to the differentiation of preadipocytes [7,8], myoblasts [9], and osteoblasts [10]. In this present study, metformin, an established activator of AMPK, clearly inhibited the synthesis of cartilage matrix (Figs. 3A and 4A) and the expression of chondrogenic differentiation marker genes (Figs. 3C and 4C) during the chondrogenic differentiation of both ATDC-5 cell line and primary mouse embryonic limb cells. AICAR, another widely used activator of AMPK [7,8], showed cytotoxicity (Figs. 2C, D) and was not suitable for the chondrogenic differentiation experiments which need long-time cell survival. Conversely, the suppression of AMPK by high glucose condition and prkaa1 siRNA promoted the chondrogenic differentiation process (Figs. 5 and 6B). These finding have indicated that AMPK activity negatively influences chondrogenic differentiation. Cartilage matrix maturation is largely regulated by the gene expression of a series of cartilage matrix proteins in chondrocytes. The expression of col2a1 and acp mRNA was suppressed by metformin in both ATDC-5 cells and embryonic limb cells (Figs. 3C and 4C). Previous reports have shown that the expression of both col2a1 and acp is under the control of two Sox transcription factors, Sox9 and Sox6 [18,20]. In ATDC-5 cells, mRNA levels of sox9 and sox6 gradually increased during the first 2 weeks of differentiation (Fig. 3C). After reaching the peaks at week 2, sox9 mRNA decreased, whereas the sox6 mRNA remained elevated at least until week 4 (Fig. 3C). On the other hand, in embryonic limb cells, the expressions of sox9 and sox6 were temporarily increased at 2 days (Fig. 4C). We presume that the time discrepancy was caused by the polyclonal nature of the embryonic limb cells, which presumably include more differentiated chondrocyte progenitors than ATDC-5 cells. Our present data have revealed that AMPK activity directly decreases the gene expression of sox9 through the inhibition of its transcriptional activity (Fig. 7). The inhibition seems to be mediated by a transcriptional repressor, Egr-1, as the mutation of the putative Egr-1 binding site abrogated the inhibitory effects of AMPK activators (Fig. 7B). Furthermore, metformin significantly increased the protein expression level of Egr-1 in ATDC-5 cells (Fig. 7D). These findings are consistent with several previous reports showing the critical roles of Egr-1 in chondrogenic marker gene expressions in rabbit [33], rat [34], and human [35] chondrocytes. Activation of Egr1 by AMPK has also been reported in mouse pituitary cells [36] and human monocytes [37]. It should be noted that a previous report has demonstrated that Egr-1 suppresses col2a1 promoter activity in rabbit chondrocytes [33] and col2a1 and acp promoter activity in rat chondrocytes [34]. We have also identified putative Egr-1 binding sites at − 32, and − 16 in the promoters of mouse col2a1 and acp genes respectively using AliBaba 2.1, a predicting program of transcription factor binding sites [38]. Therefore, it is also reasonable to assume that AMPK-activated Egr-1
directly represses col2a1 and acp promoter activities in mouse chondrocytes independent of Sox9. Unlike sox9, our promoter assay data have indicated that metformin has no effect on the IST-induced transcriptional activity of sox6 (Fig. 7A), indicating that the inhibitory effect of metformin on sox6 is other than the direct transcriptional downregulation. Consistently, no putative Egr-1 binding site was found in the 5′ upstream regulatory region of sox6 gene using AliBaba 2.1 program. Although there is currently no direct experimental evidence, we would like to propose three hypotheses. As the induction of sox6 mRNA is relatively slower than that of sox9 during chondrogenesis (Fig. 3C), it seems possible that sox6 mRNA level is indirectly decreased by metformin through the downregulation of Sox9 transcriptional factor. This hypothesis is consistent with a previous finding that the nucleotide sequence of −180/−100 in the sox6 gene promoter is conserved between mouse and human and is transcriptionally responsive to Sox9 [29]. Another possible explanation is that metformin suppresses sox6 mRNA expression via a pathway other than transcriptional inhibition, such as decreasing mRNA stability. In fact, activated AMPK is known to decrease Heat Shock Protein (HSP) 70 mRNA in human hepatocellular liver carcinoma cells [39] and Cox2 mRNA in human keratinocytes [40] through mRNA instability. Alternatively, it is possible that metformin negatively regulates sox6 transcription through regulatory regions further upstream or downstream of the sox6 gene. It should be mentioned, however, that no putative Egr-1 binding site was found using AliBaba 2.1 program within the 5000 bp region upstream of the transcriptional initiation site of sox6. Several previous reports have indicated that low AMPKα activity is related to the degradation of articular cartilage in osteoarthritis (OA). The number of chondrocytes constitutively positive for AMPKα phosphorylation was less in severe OA patients [41,42]. In an experimental OA mouse model, AMPK activity was decreased in chondrocytes of knee joint [41]. One possible explanation is that AMPK activity may be protective against the degradation of articular cartilage. AMPK activation suppressed matrix degradation responses to IL-1β and TNFα in human chondrocytes [42], and the knockdown of LKB1, an upstream activator of AMPK, promoted matrix catabolic responses in human chondrocytes [41]. On the other hand, our present data have indicated that AMPK activity exerts suppressive effects on chondrocyte differentiation including the hypertrophic marker gene expression (Fig. 3D). Notably, focal premature chondrocyte differentiation to hypertrophic cells is frequently found in OA cartilage and is considered to be associated with disease progression [13,14]. Thus, deregulated chondrocyte differentiation may be an alternative explanation of how low AMPK activity is associated. Conclusion In summary, our present study has suggested a negative regulatory role of AMPK in chondrogenic differentiation through the suppression of sox9, sox6, col2a1, and acp expressions. This study may provide insights into a new list of target molecules for regenerative therapies of cartilage lesions, such as OA and articular cartilage injuries. Acknowledgments This work was supported by Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (Grant 23592739, Grant 23592740, Grant 23592741, and Grant 25462895). We thank Ms. Momoko Uemura for the secretarial assistance. References [1] Hardie DG. AMP-activated/SNF1 protein kinases: conserved guardians of cellular energy. Nat Rev Mol Cell Biol 2007;8:774–85.
K. Bandow et al. / Bone 74 (2015) 125–133 [2] Cheung PC, Salt IP, Davies SP, Hardie DG, Carling D. Characterization of AMPactivated protein kinase gamma-subunit isoforms and their role in AMP binding. Biochem J 2000;346(Pt 3):659–69. [3] Hawley SA, Davison M, Woods A, Davies SP, Beri RK, Carling D, et al. Characterization of the AMP-activated protein kinase kinase from rat liver and identification of threonine 172 as the major site at which it phosphorylates AMP-activated protein kinase. J Biol Chem 1996;271:27879–87. [4] Shaw RJ, Lamia KA, Vasquez D, Koo SH, Bardeesy N, Depinho RA, et al. The kinase LKB1 mediates glucose homeostasis in liver and therapeutic effects of metformin. Science 2005;310:1642–6. [5] Lee JH, Koh H, Kim M, Kim Y, Lee SY, Karess RE, et al. Energy-dependent regulation of cell structure by AMP-activated protein kinase. Nature 2007;447:1017–20. [6] McCarty MF. Chronic activation of AMP-activated kinase as a strategy for slowing aging. Med Hypotheses 2004;63:334–9. [7] Habinowski SA, Witters LA. The effects of AICAR on adipocyte differentiation of 3 T3L1 cells. Biochem Biophys Res Commun 2001;286:852–6. [8] Dagon Y, Avraham Y, Berry EM. AMPK activation regulates apoptosis, adipogenesis, and lipolysis by eIF2alpha in adipocytes. Biochem Biophys Res Commun 2006;340:43–7. [9] Fulco M, Cen Y, Zhao P, Hoffman EP, McBurney MW, Sauve AA, et al. Glucose restriction inhibits skeletal myoblast differentiation by activating SIRT1 through AMPKmediated regulation of Nampt. Dev Cell 2008;14:661–73. [10] Kasai T, Bandow K, Suzuki H, Chiba N, Kakimoto K, Ohnishi T, et al. Osteoblast differentiation is functionally associated with decreased AMP kinase activity. J Cell Physiol 2009;221:740–9. [11] Li X, Han Y, Pang W, Li C, Xie X, Shyy JY, et al. AMP-activated protein kinase promotes the differentiation of endothelial progenitor cells. Arterioscler Thromb Vasc Biol 2008;28:1789–95. [12] Akiyama H. Control of chondrogenesis by the transcription factor Sox9. Mod Rheumatol 2008;18:213–9. [13] von der Mark K, Kirsch T, Nerlich A, Kuss A, Weseloh G, Gluckert K, et al. Type X collagen synthesis in human osteoarthritic cartilage. Indication of chondrocyte hypertrophy. Arthritis Rheum 1992;35:806–11. [14] Boos N, Nerlich AG, Wiest I, von der Mark K, Ganz R, Aebi M. Immunohistochemical analysis of type-X-collagen expression in osteoarthritis of the hip joint. J Orthop Res 1999;17:495–502. [15] Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. [16] Elima K, Vuorio E. Expression of mRNAs for collagens and other matrix components in dedifferentiating and redifferentiating human chondrocytes in culture. FEBS Lett 1989;258:195–8. [17] Bi W, Huang W, Whitworth DJ, Deng JM, Zhang Z, Behringer RR, et al. Haploinsufficiency of Sox9 results in defective cartilage primordia and premature skeletal mineralization. Proc Natl Acad Sci U S A 2001;98:6698–703. [18] Ng LJ, Wheatley S, Muscat GE, Conway-Campbell J, Bowles J, Wright E, et al. SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev Biol 1997;183:108–21. [19] Lefebvre V, Li P, de Crombrugghe B. A new long form of Sox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis and cooperatively activate the type II collagen gene. EMBO J 1998;17:5718–33. [20] Smits P, Li P, Mandel J, Zhang Z, Deng JM, Behringer RR, et al. The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Dev Cell 2001;1:277–90. [21] Akiyama H, Chaboissier MC, Martin JF, Schedl A, de Crombrugghe B. The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 2002;16:2813–28. [22] Atsumi T, Miwa Y, Kimata K, Ikawa Y. A chondrogenic cell line derived from a differentiating culture of AT805 teratocarcinoma cells. Cell Differ Dev 1990;30:109–16.
133
[23] Wroblewski J, Engstrom M, Edwall-Arvidsson C, Sjoberg G, Sejersen T, Lendahl U. Distribution of nestin in the developing mouse limb bud in vivo and in micromass cultures of cells isolated from limb buds. Differentiation 1997;61:151–9. [24] Bohensky J, Leshinsky S, Srinivas V, Shapiro IM. Chondrocyte autophagy is stimulated by HIF-1 dependent AMPK activation and mTOR suppression. Pediatr Nephrol 2010;25:633–42. [25] Asahina I, Sampath TK, Hauschka PV. Human osteogenic protein-1 induces chondroblastic, osteoblastic, and/or adipocytic differentiation of clonal murine target cells. Exp Cell Res 1996;222:38–47. [26] Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 2012;9:671–5. [27] Bandow K, Nishikawa Y, Ohnishi T, Kakimoto K, Soejima K, Iwabuchi S, et al. Lowintensity pulsed ultrasound (LIPUS) induces RANKL, MCP-1, and MIP-1beta expression in osteoblasts through the angiotensin II type 1 receptor. J Cell Physiol 2007; 211:392–8. [28] Ushita M, Saito T, Ikeda T, Yano F, Higashikawa A, Ogata N, et al. Transcriptional induction of SOX9 by NF-kappaB family member RelA in chondrogenic cells. Osteoarthritis Cartilage 2009;17:1065–75. [29] Ikeda T, Saito T, Ushita M, Yano F, Kan A, Itaka K, et al. Identification and characterization of the human SOX6 promoter. Biochem Biophys Res Commun 2007;357: 383–90. [30] Christy B, Nathans D. DNA binding site of the growth factor-inducible protein Zif268. Proc Natl Acad Sci U S A 1989;86:8737–41. [31] Lemaire P, Vesque C, Schmitt J, Stunnenberg H, Frank R, Charnay P. The seruminducible mouse gene Krox-24 encodes a sequence-specific transcriptional activator. Mol Cell Biol 1990;10:3456–67. [32] Hawley SA, Fullerton MD, Ross FA, Schertzer JD, Chevtzoff C, Walker KJ, et al. The ancient drug salicylate directly activates AMP-activated protein kinase. Science 2012; 336:918–22. [33] Tan L, Peng H, Osaki M, Choy BK, Auron PE, Sandell LJ, et al. Egr-1 mediates transcriptional repression of COL2A1 promoter activity by interleukin-1beta. J Biol Chem 2003;278:17688–700. [34] Rockel JS, Bernier SM, Leask A. Egr-1 inhibits the expression of extracellular matrix genes in chondrocytes by TNFalpha-induced MEK/ERK signalling. Arthritis Res Ther 2009;11:R8. [35] Vincenti MP, Brinckerhoff CE. Early response genes induced in chondrocytes stimulated with the inflammatory cytokine interleukin-1beta. Arthritis Res 2001;3:381–8. [36] Andrade J, Quinn J, Becker RZ, Shupnik MA. AMP-activated protein kinase is a key intermediary in GnRH-stimulated LHbeta gene transcription. Mol Endocrinol 2013;27: 828–39. [37] Kubosaki A, Tomaru Y, Tagami M, Arner E, Miura H, Suzuki T, et al. Genome-wide investigation of in vivo EGR-1 binding sites in monocytic differentiation. Genome Biol 2009;10:R41. [38] Grabe N. AliBaba2: context specific identification of transcription factor binding sites. In Silico Biol 2002;2:S1–S15. [39] Wang T, Yu Q, Chen J, Deng B, Qian L, Le Y. PP2A mediated AMPK inhibition promotes HSP70 expression in heat shock response. PLoS One 2010;5. [40] Zhang J, Bowden GT. UVB irradiation regulates Cox-2 mRNA stability through AMPK and HuR in human keratinocytes. Mol Carcinog 2008;47:974–83. [41] Petursson F, Husa M, June R, Lotz M, Terkeltaub R, Liu-Bryan R. Linked decreases in liver kinase B1 and AMP-activated protein kinase activity modulate matrix catabolic responses to biomechanical injury in chondrocytes. Arthritis Res Ther 2013;15:R77. [42] Terkeltaub R, Yang B, Lotz M, Liu-Bryan R. Chondrocyte AMP-activated protein kinase activity suppresses matrix degradation responses to proinflammatory cytokines interleukin-1beta and tumor necrosis factor alpha. Arthritis Rheum 2011;63: 1928–37.
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Original Full Length Article
AMP-activated protein kinase (AMPK) activity negatively regulates chondrogenic differentiation Kenjiro Bandow, Joji Kusuyama, Kyoko Kakimoto, Tomokazu Ohnishi, Tetsuya Matsuguchi ⁎ Department of Oral Biochemistry, Field of Developmental Medicine, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
a r t i c l e
i n f o
Article history: Received 1 May 2014 Revised 2 December 2014 Accepted 3 December 2014 Available online 10 December 2014 Edited by: Regis O'Keefe Keywords: Chondrocyte AMPK Metformin Sox9 Egr-1
a b s t r a c t Chondrocytes are derived from mesenchymal stem cells, and play an important role in cartilage formation. Sex determining region Y box (Sox) family transcription factors are essential for chondrogenic differentiation, whereas the intracellular signal pathways of Sox activation have not been clearly elucidated. AMP-activated protein kinase (AMPK) is a serine–threonine kinase generally regarded as a key regulator of cellular energy homeostasis. It is known that the catalytic alpha subunit of AMPK is activated by upstream AMPK kinases (AMPKKs) including liver kinase B1 (LKB1). We have previously reported that AMPK is a negative regulator of osteoblastic differentiation. Here, we have explored the role of AMPK in chondrogenic differentiation using in vitro culture models. The phosphorylation level of the catalytic AMPK alpha subunit significantly decreased during chondrogenic differentiation of primary chondrocyte precursors as well as ATDC-5, a well-characterized chondrogenic cell line. Treatment with metformin, an activator of AMPK, significantly reduced cartilage matrix formation and inhibited gene expression of sox6, sox9, col2a1 and aggrecan core protein (acp). Thus, chondrocyte differentiation is functionally associated with decreased AMPK activity. © 2015 Elsevier Inc. All rights reserved.
Introduction AMP-activated protein kinase (AMPK) is a serine–threonine kinase, which maintains the balance between production and consumption of ATP in eukaryotic cells [1]. AMPK is a heterotrimeric enzyme composed of the catalytic α subunit and regulatory β and γ subunits in a 1:1:1 stoichiometric ratio [2]. AMPK is activated by the elevation of intracellular AMP/ATP ratio caused by energy restriction. Subsequently, ATP-generating pathways are activated, while ATP-consuming mechanisms are inhibited, thereby restoring the normal cellular AMP/ATP ratio. Thus, AMPK is generally considered as an essential regulator of energy homeostasis of cells, and termed as a metabolic “energy sensor” of the biological system. The activation of AMPK is induced by phosphorylation of the catalytic α subunit [3]. The most characterized upstream AMPK kinase (AMPKK) is liver kinase B1 (LKB1), which is a reported target of metformin, a type 2 diabetes drug [4]. Recent reports have revealed that AMPK has non-metabolic functions as well. In vivo studies of Drosophila have demonstrated that AMPK is essentially involved in cell polarity and mitosis [5]. On the other hand, activation of AMPK has been proposed as one of the regulatory mechanisms of mammal longevity [6]. Notably, several lines of evidence have indicated the involvement of AMPK in the regulation of cellular differentiation. Activation of AMPK has been suggested to be inhibitory to the differentiation of adipocytes [7,8], myoblasts [9], and ⁎ Corresponding author. Fax: +81 99 275 6138. E-mail address: [email protected] (T. Matsuguchi).
http://dx.doi.org/10.1016/j.bone.2014.12.001 8756-3282/© 2015 Elsevier Inc. All rights reserved.
osteoblasts [10]. In contrast, it has also been reported that AMPK is promotive of the differentiation of endothelial progenitor cells [11]. In the earliest event of chondrogenesis, recruited mesenchymal cells proliferate and aggregate. Subsequently, the cells mature into chondrocytes, producing cartilage matrix proteins. After differentiation into hypertrophic cells, they are finally replaced by bone tissue [12]. Deregulation of chondrocyte differentiation has been described in various chronic skeletal diseases including osteoarthritis (OA) [13–15]. Chondrocytes produce a characteristic cartilage matrix that consists of collagen and proteoglycan [16]. During mouse embryogenesis, a transcription factor, Sox9, is expressed in all chondrocytes and their progenitor cells [12]. Heterozygous sox9 mutants exhibited delayed chondrogenic mesenchymal condensation and enlargement of the hypertrophic zone [17]. Sox9 regulates col2a1 gene, which encodes one component of collagen type II, in the early stage of chondrogenesis [18]. Two other Sox family transcription factors, Sox5 and Sox6, are co-expressed with Sox9 [19] and are required for the expression of aggrecan, which is a cartilage-specific proteoglycan [20]. Inactivation of Sox9 before mesenchymal condensation results in the absence of Sox5 and Sox6, as well as other chondrogenic marker gene expression [21], indicating that Sox9 is the essential master regulator of chondrocyte differentiation. In this present study, we have explored the role of AMPK in chondrogenic differentiation by in vitro chondrocyte differentiation models. It was found that the phosphorylation level of AMPKα was progressively decreased during chondrogenic differentiation. Stimulation with metformin during chondrogenic differentiation significantly
126
K. Bandow et al. / Bone 74 (2015) 125–133
decreased gene expression of sox9 and sox6 along with other chondrogenic differentiation markers including col2a1, and aggrecan core protein (acp). Conversely, knock-down of AMPKα expression by siRNA increased sox9, col2a1, and acp mRNA Thus, our present data indicate that differentiation of chondrocytes is functionally associated with decreased AMPK activity.
Western blot analysis For Western blot analyses, total cellular lysate preparation and immunoblotting procedures were performed as previously described [27]. Quantitative polymerase chain reaction (qPCR) analysis
Materials and methods Reagents and antibodies Sodium selenite, transferrin human, metformin, sodium salicylate, 5-aminoimidazole-4-carboxamide-1-β-D-ribofuranoside (AICAR) and insulin from bovine pancreas were purchased from Sigma-Aldrich (St. Louis, MO). Antibodies specifically recognizing AMPKα1/2, AMPKβ1/2, phospho-AMPKα (Thr172), and EGR1 were purchased from Cell Signaling Technology (Danvers, MA). Antibodies against AMPKγ1/ 2/3 and β-Actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
Monolayer and micro-mass culture ATDC-5, a well-characterized chondrogenic cell line derived from mouse teratocarcinoma [22], was obtained from Riken Cell Bank (Tsukuba, Japan), and maintained in Dulbecco's modified Eagle medium/Ham's F-12 (DMEM/F-12) (Wako Pure Chemical Industries, Osaka, Japan) containing 5% fetal bovine serum (FBS), 50 U/ml penicillin, and 50 mg/ml streptomycin. The cells were induced to differentiate by IST (1.7 mM insulin, 10 nM sodium selenite, and 125 μM transferrin) for 4 weeks. C57BL/6 mice (wild-type) were obtained from Japan Clea (Tokyo, Japan). The animals were maintained in accordance with protocols approved by the Animal Care and Use Committee at Kagoshima University, Japan. Primary chondrogenic progenitor cells were isolated from murine limb buds at embryonic day 10.5 (E10.5) and cultured as previously described [23].
Transfection of small interference RNA (siRNA) Chemically synthesized siRNAs, purified by HPLC, were purchased from Sigma-Aldrich (MO, USA). The following sequences were synthesized as prkaa1 siRNA: sense sequence: 5′-GAU GCA AUA AGC AUG GAU ATT-3′ and anti-sense sequence: 5′-UAU GCA UGC UUA UUG CAU CTT-3′. Non-targeting control siRNA duplexes (Control siRNA-A) were purchased from Santa Cruz Biotechnology. ATDC-5 cells were seeded in a well of a six-well plate at a density of 5.0 × 104 cells, and then cultured for 12 h before each experiment. The duplex siRNA was transfected using HilyMax reagent (Dojindo Laboratories, Kumamoto, Japan), according to the manufacturer's instructions.
Measurement of cell viability Cell viability was measured by 3-(4,5-di-methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, yellow tetrazole (MTT) assay as previously described [24].
Alcian blue staining Chondrogenic matrix was visualized by Alcian blue staining as previously described [25]. The stained area was photographically measured by ImageJ program [26].
The isolation of total RNA and qPCR was conducted as previously described [27]. The primer sequences were described in Table 1. Two housekeeping genes, ribosomal protein L13a (rpl13a) and TATA box binding protein (tbp), were used as endogenous controls in the qPCR. Luciferase reporter gene assay The promoter regions of mouse sox9 (− 953/+ 19) and sox6 (− 842/+66) genes containing the putative transcriptional initiation sites were cloned by PCR from C57BL/6 genomic DNA using pairs of primers, 5′-TCA ACC CCG GAG TAG TTT TG-3′ and 5′-CTC TCC GAC TTC CAG CTC AG-3′ for sox9 and 5′-CTG CTA CAG GCT CCT TTG CT-3′ and 5′-TCA CAT TGG CAA CTG CCT TA-3′ for sox6, and ligated into the luciferase reporter vector pGL4.19 (Promega, WI) to generate pGL-Sox9953 + 19 and pGL-Sox6-842 + 66, respectively. Deletion (pGL-Sox9457 + 19, pGL-Sox9-207 + 19, and pGL-Sox6-447 + 66) and mutation (pGL-Sox9-953 + 19mt) constructs were created by methods using PCR as previously described [28]. The primer sequences were described in Table 2. Each of the reporter plasmid or the vector control, pGL4.19, was transfected into 70% confluent ATDC-5 cells using HilyMax reagent, according the manufacturer's instruction. At 24 h after transfection, the
Table 1 Primers used in this study. Gene symbol prkaa1
Primers [5′–3′]
TGAGAACGTCCTGCTTGATG ATAATTGGGTGAGCCACAGC prkaa2 CCCCTGGTCTCTGTCTTTCT GCTGGATTTCTGTGTTTTCG prkab1 TCCGATGTGTCTGAGCTGTC CCCGTGTCCTTGTTCAAGAT prkab2 GAAGGAGGCAAGGAGGTCTT ACTGATGCTCTCCCTCTGGA prkag1 CTCCGCCTTACCTGTAGTGG AAGTAGTGGGACCGATGCTG prkag2 GTCCATCGGCTGGTTGTAGT TCTCCTTCTGTTTGGCACCT prkag3 CTCCCAATGACAGCCTGTTT GGAACTTGAGTAGCCGCTTG sox6 ATCTCCCACCCAGAACCTCT CAGGGCAGGAGAGTTGAGAC sox9 CCATGTGGCCAGCAGATG TTTTAGCACATGGGATGTCTTG AA col2a1 AGAACAGCATCGCCTACCTG CTTGCCCCACTTACCAGTGT col9a1 CAGGATTGGCCAAGATGACT TTCCCAGCTTGTAAGCCACT col10a1 CATCTCCCAGCACCAGAATC GTGTCTTGGGGCTAGCAAGT acp AACTTCTTTGCCACCGGAGA GGTGCCCTTTTTACACGTGAA alp TCCTG CCAAAAACCTCAAAGG TGCTTCATGCAGAGCCTGC mmp13 GCCATTTCATGCTTCCTGAT CTCTGGTGTTTTGGGATGCT rpl13a GCTTACCTGGGGCGTCTG ACATTCTTTTCTGCCTGTTTCC tbp CAAACCCAGAATTGTTCTCCTT ATGTGGTCTTCCTGAATCCCT
GenBank
Amplication length
NM_001013367 112 bp (467–578 nt) NM_178143
159 bp (1767–1925 nt)
NM_031869
140 bp (699–838 nt)
NM_182997
131 bp (433–553 nt)
NM_016781
113 bp (742–885 nt)
NM_145401
154 bp (2094–2205 nt)
NM_153744
129 bp (968–1096 nt)
NM_011445
145 bp (1492–1636 nt)
NM_011448
99 bp (3049–3147 nt)
NM_031163
161 bp (4412–4572 nt)
NM_007740
161 bp (386–516 nt)
NM_009925
152 bp (48–199 nt)
NM_007424
111 bp (6161–6271 nt)
NM_007431
101 bp (1138–1238 nt)
NM_008607
131 bp (793–890 nt)
NM_009438
149 bp (437–585 nt)
NM_013684
131 bp (876–1006 nt)
K. Bandow et al. / Bone 74 (2015) 125–133 Table 2 Primers for luciferase reporter vectors.
sox6 promoter −842 + 66 −447 + 66 sox9 promoter −953 + 19 −457 + 19 −207 + 19 For mutagenesis −953 + 219mt −228mt + 19
Forward primers [5′–3′]
Reverse primers [5′–3′]
GGTACCTGCTACAGG CTCCTTTGCT GGTACCGCCAGGACA GTTGGTTGTTT
AGATCTCACATTGGCA ACTGCCTTA AGATCTCACATTGGCA ACTGCCTTA
CTCGAGTCAACCCCG GAGTAGTTTTG GGTACCACGGAGAC AGCATC GGTACCCCTCACCCC ACCATCCA C
AGATCTCTCCGACTTC CAGCTCAG AGATCTCTCCGACTTC CAGCTCAG AGATCTCTCCGACTTC CAGCTCAG
CTCGAGTCAACCCCG GAGTAGTTTTG GCCTTTCGCCCCCCT CCAAG
GGCGAAAGGCAAGC GGGGGA AGATCTCTCCGACTT CCAGCTCAG
127
differentiation of ATDC-5 cells (Figs. 1A and B). On the other hand, the phosphorylation level of the AMPKα subunit was significantly decreased during differentiation in ATDC-5 cells (Fig. 1A). Metformin inhibited differentiation of ATDC-5 cells
cells were treated with IST and metformin, followed by further incubation for 48 h. Luciferase activities were measured and normalized by protein contents of the cell lysates.
Results The phosphorylation level of AMPKα decreased during chondrogenic differentiation In order to assess AMPK activity during the course of chondrocyte differentiation, we analyzed phosphorylation of AMPKα and the expression levels of AMPK α/β/γ subunits during chondrogenic differentiation by Western blot and qPCR analyses. ATDC-5 cells were induced to differentiate in the presence of IST, and total cell lysates and RNA were isolated. It was found that the protein levels of AMPKα, β, and γ subunits and mRNA expression levels of prkaa, prkab, and prkag (respectively coding AMPKα, β, and γ subunits) remained constant during the chondrogenic
In order to investigate how forced activation of AMPK affects chondrocyte differentiation, we sought to use metformin and AICAR, known activators of AMPK. We first examined the dose-dependent effect of metformin on the phosphorylation level of AMPKα in ATDC-5 cells. As expected, significant phosphorylation of AMPKα was induced by metformin (Fig. 2A). The maximum phosphorylation was observed at 2 mM of metformin. On the other hand, the cell viability of ATDC-5 cells was not significantly affected by metformin at up to 2 mM, but was decreased at 5 mM metformin (Fig. 2B). Unlike metformin, AICAR failed to induce sufficient phosphorylation of AMPKα in ATDC-5 cells at non-toxic concentrations (Figs. 2C, D). We thus evaluated the effects of 2 mM metformin treatment on the cartilage matrix synthesis by ATDC-5 cells. The matrix synthesis induced by IST was significantly inhibited by 2 mM metformin (Figs. 3A and B), indicating that the forced AMPKα activation is correlated with the inhibition of matrix synthesis in ATDC-5 cells. We next examined the effect of metformin on the expression of chondrogenic marker genes by qPCR analyses. In ATDC-5 cells, induction of acp and col2a1 mRNA was inhibited by 2 mM metformin (Fig. 3C). Since Sox family transcription factors are master regulators of chondrogenic differentiation controlling acp and col2a1 mRNA expression, we next analyzed the effect of metformin on the gene expression of sox9 and sox6. In the absence of metformin, the expression level of sox9 mRNA increased approximately 10 fold in the first 2 weeks of differentiation, and started to decrease from week 3 (Fig. 3C). On the other hand, the expression of sox6 mRNA gradually increased and remained maximally elevated after week 2 (Fig. 3C). The addition of metformin significantly decreased the levels of both sox9 and sox6 mRNA from week 1 to week 4 (Fig. 3C). We further analyzed the effect of metformin on the expression of four hypertrophic chondrocyte marker genes: col9a1, col10a1, alkaline phosphatase (alp), and matrix metalloproteinase 13 (mmp13). The addition of metformin significantly repressed the induction processes for these four hypertrophic markers (Fig. 3D).
Fig. 1. Expression and phosphorylation of AMPK subunits during chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST. A: Cytoplasmic lysates were isolated at the indicated weeks. The same amount of cell lysate (15 μg protein/lane) was separated in each lane by SDS-PAGE. Expression and phosphorylation of AMPK were detected with the indicated specific antibodies. B: Total RNA was isolated at the indicated week and the expression level of the indicated mRNA was analyzed by qPCR. Results of qPCR are shown for prkaa1, a2, b1, b2, g1, g2, and g3. Fold increase represents an experimental value divided by the control (prkaa1, b1, or g2 at week 0) value for each. Vertical bars indicate mean S.D. of at least three different experiments.
128
K. Bandow et al. / Bone 74 (2015) 125–133
Fig. 2. Effects of metformin and AICAR on AMPKα phosphorylation in ATDC-5 cells. A: Cells were cultured with varying concentrations of metformin in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. B: The viabilities of ATDC-5 cells after 48 h of treatment with various concentrations of metformin were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. C: Cells were cultured with varying concentrations of AICAR or 2 mM metformin in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. D: The viabilities of ATDC-5 cells after 48 h of treatment with various concentrations of AICAR or 2 mM metformin were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
Metformin suppressed chondrogenic differentiation of mouse embryonic limb cells We next explored the effects of metformin on chondrogenic differentiation of mouse embryonic limb cells. The matrix synthesis induced
by IST was moderately inhibited by the addition of 2 mM metformin (Figs. 4A and B). Consistently, induction of acp and col2a1 mRNA in mouse embryonic limb cells was suppressed by about 30% by 2 mM metformin (Fig. 4C). We subsequently analyzed the effect of metformin on the gene expression of sox9 and sox6. In the absence of metformin,
Fig. 3. Metformin inhibited chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST for the indicated weeks and were stained by Alcian blue (A). Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown (B). ATDC-5 cells were induced to differentiate by IST with or without 2 mM metformin. Total RNA was isolated at the indicated time and reverse-transcripted. Quantitative PCR analyses were performed for sox9, sox6, col2a1, acp (C), and col9a1, col10a1, alp, and mmp13 (D). Fold increase represents an experimental value divided by the control (week 0: sox9, sox6, col2a1, acp, alp, and mmp13; week 1: col9a1; week 2: col10a1) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
K. Bandow et al. / Bone 74 (2015) 125–133
the expression levels of sox9 and sox6 mRNA temporarily increased at day 2 and decreased by day 5 (Fig. 4C). The addition of metformin moderately decreased the expression level of sox9 and sox6 mRNA from day 2 to day 5 (Fig. 4C). High glucose concentration promoted differentiation of ATDC-5 cells In order to investigate how decreased AMPK activity affects chondrocyte differentiation, we examined the effects of a high-glucose culture condition on the ATDC-5 differentiation. We first examined the dose-dependent effect of glucose on the phosphorylation level of AMPKα in ATDC-5 cells. As expected, phosphorylation of AMPKα was inhibited by glucose in a concentration-dependent manner, reaching the maximum inhibition at 15 mM of glucose (Fig. 5A). We thus evaluated the effects of 15 mM glucose treatment on the cartilage matrix synthesis by ATDC-5 cells, and found that the matrix synthesis induced by IST was significantly promoted by 15 mM glucose (Figs. 5B and C). This result has indicated that forced AMPKα inhibition by high glucose concentration may promote cartilage matrix synthesis of ATDC-5 cells. We next examined the effect of high-concentration glucose on the expression of chondrogenic marker genes by qPCR analyses. In ATDC-5 cells, induction of sox9, sox6, acp and col2a1 mRNA was promoted by 15 mM glucose (Fig. 5D). We further analyzed the effect of high
129
glucose concentration on the gene expression of col9a1 and col10a1, chondrogenic marker genes at the late differentiation stage. The mRNA of col9a1, and col10a1 became detectable from week 2 and week 3 of differentiation, respectively (Fig. 5D). The high glucose condition significantly promoted the induction processes of these two late differentiation markers (Fig. 5D).
AMPKα knock-down by siRNA promoted chondrogenic differentiation of ATDC-5 cells For another experiment to examine the effect of reduced AMPK activity on chondrogenic differentiation, we introduced prkaa1 siRNA into ATDC-5 cells. The protein expression level of AMPKα was partially but significantly inhibited by prkaa1 siRNA compared with control siRNA at day 7 (Fig. 6A). Modest mRNA upregulation of sox9, col2a1, and acp, but not sox6, was observed by the transduction of prkaa1 siRNA at day 7 of chondrogenic differentiation (Fig. 6B). The reason for the failed increase of sox6 mRNA is not clear. It is possible that the required threshold level is relatively low for sox6 and the partial inhibitory effect of prkaa1 siRNA (Fig. 6A) was not sufficient. It was confirmed that the partial AMPKα knock-down by siRNA did not affect cell viability and cellular metabolism (Fig. 6C).
Fig. 4. Mouse primary chondrocyte precursor cells from E10.5 embryo limb buds were cultured in micro-mass method with or without 2 mM metformin for 5 days. The cells were stained with Alcian blue (A). Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown (B). Total RNA was collected and qPCR analyses were performed for sox9, sox6, col2a1, and acp (C). Fold increase represents an experimental value divided by the control (day 0) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
130
K. Bandow et al. / Bone 74 (2015) 125–133
Fig. 5. High-concentration glucose promoted chondrogenic differentiation. A: ATDC-5 cells were cultured with varying concentrations of glucose in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. B: ATDC-5 cells were induced to differentiate by IST for the indicated weeks in the presence of 5 mM (Control) or 15 mM glucose (High Glc) and were stained by Alcian blue. C: Relative ratios of Alcian blue-stained area after metformin-treatment in comparison with the control are shown. D: ATDC-5 cells were induced to differentiate by IST with 5 mM (black bar) or 15 mM glucose (gray bar). Total RNA was isolated at the indicated time and qPCR analyses were performed for sox9, sox6, col2a1, acp, col9a1 and col10a1. Fold increase represents an experimental value divided by the control (week 0: sox9, sox6, col2a1, and acp; week 1: col9a1; week 2: col10a1) value for each. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
Inhibitory effects of metformin on the IST-induced transcriptional activation of sox9 The 5′ upstream regions of the Sox9 and Sox6 genes are highly conserved between human and mouse, and the crucial regions for chondrogenic differentiation were found in −202/−128 and −517/−1 upstream of the transcriptional initiation sites in human Sox9 and Sox6 genes, respectively [28,29]. To examine the effect of metformin on the transcriptional activities of sox9 and sox6, we made the luciferase reporter plasmids containing different lengths of the 5′ upstream regions of mouse sox9 and sox6 genes. IST-treatment significantly promoted
transcriptional activities of three (− 953, − 457, and − 207) sox9 and two (−842 and −447) sox6 reporter constructs (Fig. 7A). These results indicated that the essential regulatory regions for sox9 and sox6 genes during chondrogenic differentiation were located within the 207 and 447 nucleotide lengths upstream of their transcriptional initiation sites, respectively. As for sox9, treatment with metformin significantly inhibited the IST-induced transcriptional activations of the −953 and −457 promoter constructs, but failed to reduce the promoter activity of the − 207 construct (Fig. 7A), indicating that a negative effect of metformin was mediated by the 5′-upstream region between − 457 and − 207. On
K. Bandow et al. / Bone 74 (2015) 125–133
131
Fig. 6. AMPKα knock-down by siRNA promoted chondrogenic differentiation. ATDC-5 cells were induced to differentiate by IST with 200 μM of control siRNA or prkaa1 siRNA. A: The inhibitory effect of prkaa1 siRNA on AMPKα protein expression was confirmed by Western blotting on day 7. B: Quantitative PCR analyses were performed for sox9, sox6, col2a1, and acp mRNAs on day 7. Fold increase represents an experimental value divided by the control (control siRNA-treatment) value for each. Vertical bars indicate mean SD of at least three different experiments. C: The viabilities of ATDC-5 cells after 48 h of treatment with siRNA were analyzed by MTT assay. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the control group by Bonferroni test (p b 0.01).
the other hand, metformin did not affect the IST-induced transcriptional activation of the sox6 promoter constructs (Fig. 7A). When we analyzed the nucleotide sequences between −457 and −207 in the sox9 gene, we
found the GC-rich putative binding motif for Early growth response-1 (Egr-1) transcriptional repressor protein [30,31] in − 228/− 219. We then constructed pGL-Sox9-953 + 19mt plasmid with three nucleotide
Fig. 7. Metformin inhibits IST-induced transcriptional activation of sox9 gene in ATDC-5 cells. A: ATDC-5 cells, transiently transfected with pGL-Sox9-953 + 19, pGL-Sox9-457 + 19, pGLSox9-207 + 19, pGL-Sox6-842 + 66, pGL-Sox6-447 + 66, or pGL4.19, were stimulated with IST for 48 h. Cells were lysed for luciferase activity measurements as described in Material and methods. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the IST-treated group by Bonferroni test (p b 0.01). B: Nucleotide sequence of the 20 bp element (−233/−214) containing the putative Egr-1 biding motif (−228/−219, underlined). Three-base mutation was introduced at −225/−223 (black boxes) within the Egr-1 motif (mt). ATDC-5 cells, transiently transfected with pGL-Sox9-953 + 19, pGL-Sox9-953 + 19mt, or pGL4.19, were lysed for luciferase activity measurements after IST treatment with 2 mM metformin, 0.5 mM AICAR, 10 mM SA, or vehicle for 48 h. Vertical bars indicate mean ± S.D. of at least three different experiments. *Significant difference from the value of the IST-treated group by Bonferroni test (p b 0.01). C: Cells were cultured with 10 mM SA in culture medium for 2 days. Cell lysates were separated by SDS-PAGE, and Western blotting was performed with the indicated antibodies. D: ATDC-5 cells were cultured with or without 2 mM metformin. Cytoplasmic lysates were isolated at the indicated days. The same amount of cell lysate (15 μg protein/lane) was separated in each lane by SDS-PAGE. Expressions of Egr-1 and β-actin were detected with the indicated specific antibodies.
132
K. Bandow et al. / Bone 74 (2015) 125–133
substitutions (TTT for CCG) in −225/−223 and found that the inhibitory effect of metformin on the IST-induced transcriptional activity was abrogated when the Egr-1 binding site was mutated (Fig. 7B). To confirm that the inhibitory effect of metformin is mediated through AMPK activation, we further examined the effects of AICAR and salicylate (SA) on the transcriptional activity of the sox9 promoter. SA was reported to directly phosphorylate AMPK [32], and we confirmed that SA could induce AMPK phosphorylation in ATDC-5 cells (Fig. 7C). We found that both AICAR and SA inhibited the transcriptional activity of sox9 promoter in a manner dependent on the Egr-1 binding site, similar to metformin (Fig. 7B). Consistently, the protein expression level of Egr-1 was increased by metformin in ATDC-5 cells (Fig. 7D). Taken together, these results have indicated that AMPK activity inhibits transcriptional activation of sox9 during chondrogenic differentiation by inducing a transcriptional repressor, Egr-1. Discussion Previous reports have demonstrated that cellular AMPK activity is functionally associated with differentiation of several cell types. More specifically, activation of AMPK has been reported to be inhibitory to the differentiation of preadipocytes [7,8], myoblasts [9], and osteoblasts [10]. In this present study, metformin, an established activator of AMPK, clearly inhibited the synthesis of cartilage matrix (Figs. 3A and 4A) and the expression of chondrogenic differentiation marker genes (Figs. 3C and 4C) during the chondrogenic differentiation of both ATDC-5 cell line and primary mouse embryonic limb cells. AICAR, another widely used activator of AMPK [7,8], showed cytotoxicity (Figs. 2C, D) and was not suitable for the chondrogenic differentiation experiments which need long-time cell survival. Conversely, the suppression of AMPK by high glucose condition and prkaa1 siRNA promoted the chondrogenic differentiation process (Figs. 5 and 6B). These finding have indicated that AMPK activity negatively influences chondrogenic differentiation. Cartilage matrix maturation is largely regulated by the gene expression of a series of cartilage matrix proteins in chondrocytes. The expression of col2a1 and acp mRNA was suppressed by metformin in both ATDC-5 cells and embryonic limb cells (Figs. 3C and 4C). Previous reports have shown that the expression of both col2a1 and acp is under the control of two Sox transcription factors, Sox9 and Sox6 [18,20]. In ATDC-5 cells, mRNA levels of sox9 and sox6 gradually increased during the first 2 weeks of differentiation (Fig. 3C). After reaching the peaks at week 2, sox9 mRNA decreased, whereas the sox6 mRNA remained elevated at least until week 4 (Fig. 3C). On the other hand, in embryonic limb cells, the expressions of sox9 and sox6 were temporarily increased at 2 days (Fig. 4C). We presume that the time discrepancy was caused by the polyclonal nature of the embryonic limb cells, which presumably include more differentiated chondrocyte progenitors than ATDC-5 cells. Our present data have revealed that AMPK activity directly decreases the gene expression of sox9 through the inhibition of its transcriptional activity (Fig. 7). The inhibition seems to be mediated by a transcriptional repressor, Egr-1, as the mutation of the putative Egr-1 binding site abrogated the inhibitory effects of AMPK activators (Fig. 7B). Furthermore, metformin significantly increased the protein expression level of Egr-1 in ATDC-5 cells (Fig. 7D). These findings are consistent with several previous reports showing the critical roles of Egr-1 in chondrogenic marker gene expressions in rabbit [33], rat [34], and human [35] chondrocytes. Activation of Egr1 by AMPK has also been reported in mouse pituitary cells [36] and human monocytes [37]. It should be noted that a previous report has demonstrated that Egr-1 suppresses col2a1 promoter activity in rabbit chondrocytes [33] and col2a1 and acp promoter activity in rat chondrocytes [34]. We have also identified putative Egr-1 binding sites at − 32, and − 16 in the promoters of mouse col2a1 and acp genes respectively using AliBaba 2.1, a predicting program of transcription factor binding sites [38]. Therefore, it is also reasonable to assume that AMPK-activated Egr-1
directly represses col2a1 and acp promoter activities in mouse chondrocytes independent of Sox9. Unlike sox9, our promoter assay data have indicated that metformin has no effect on the IST-induced transcriptional activity of sox6 (Fig. 7A), indicating that the inhibitory effect of metformin on sox6 is other than the direct transcriptional downregulation. Consistently, no putative Egr-1 binding site was found in the 5′ upstream regulatory region of sox6 gene using AliBaba 2.1 program. Although there is currently no direct experimental evidence, we would like to propose three hypotheses. As the induction of sox6 mRNA is relatively slower than that of sox9 during chondrogenesis (Fig. 3C), it seems possible that sox6 mRNA level is indirectly decreased by metformin through the downregulation of Sox9 transcriptional factor. This hypothesis is consistent with a previous finding that the nucleotide sequence of −180/−100 in the sox6 gene promoter is conserved between mouse and human and is transcriptionally responsive to Sox9 [29]. Another possible explanation is that metformin suppresses sox6 mRNA expression via a pathway other than transcriptional inhibition, such as decreasing mRNA stability. In fact, activated AMPK is known to decrease Heat Shock Protein (HSP) 70 mRNA in human hepatocellular liver carcinoma cells [39] and Cox2 mRNA in human keratinocytes [40] through mRNA instability. Alternatively, it is possible that metformin negatively regulates sox6 transcription through regulatory regions further upstream or downstream of the sox6 gene. It should be mentioned, however, that no putative Egr-1 binding site was found using AliBaba 2.1 program within the 5000 bp region upstream of the transcriptional initiation site of sox6. Several previous reports have indicated that low AMPKα activity is related to the degradation of articular cartilage in osteoarthritis (OA). The number of chondrocytes constitutively positive for AMPKα phosphorylation was less in severe OA patients [41,42]. In an experimental OA mouse model, AMPK activity was decreased in chondrocytes of knee joint [41]. One possible explanation is that AMPK activity may be protective against the degradation of articular cartilage. AMPK activation suppressed matrix degradation responses to IL-1β and TNFα in human chondrocytes [42], and the knockdown of LKB1, an upstream activator of AMPK, promoted matrix catabolic responses in human chondrocytes [41]. On the other hand, our present data have indicated that AMPK activity exerts suppressive effects on chondrocyte differentiation including the hypertrophic marker gene expression (Fig. 3D). Notably, focal premature chondrocyte differentiation to hypertrophic cells is frequently found in OA cartilage and is considered to be associated with disease progression [13,14]. Thus, deregulated chondrocyte differentiation may be an alternative explanation of how low AMPK activity is associated. Conclusion In summary, our present study has suggested a negative regulatory role of AMPK in chondrogenic differentiation through the suppression of sox9, sox6, col2a1, and acp expressions. This study may provide insights into a new list of target molecules for regenerative therapies of cartilage lesions, such as OA and articular cartilage injuries. Acknowledgments This work was supported by Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (Grant 23592739, Grant 23592740, Grant 23592741, and Grant 25462895). We thank Ms. Momoko Uemura for the secretarial assistance. References [1] Hardie DG. AMP-activated/SNF1 protein kinases: conserved guardians of cellular energy. Nat Rev Mol Cell Biol 2007;8:774–85.
K. Bandow et al. / Bone 74 (2015) 125–133 [2] Cheung PC, Salt IP, Davies SP, Hardie DG, Carling D. Characterization of AMPactivated protein kinase gamma-subunit isoforms and their role in AMP binding. Biochem J 2000;346(Pt 3):659–69. [3] Hawley SA, Davison M, Woods A, Davies SP, Beri RK, Carling D, et al. Characterization of the AMP-activated protein kinase kinase from rat liver and identification of threonine 172 as the major site at which it phosphorylates AMP-activated protein kinase. J Biol Chem 1996;271:27879–87. [4] Shaw RJ, Lamia KA, Vasquez D, Koo SH, Bardeesy N, Depinho RA, et al. The kinase LKB1 mediates glucose homeostasis in liver and therapeutic effects of metformin. Science 2005;310:1642–6. [5] Lee JH, Koh H, Kim M, Kim Y, Lee SY, Karess RE, et al. Energy-dependent regulation of cell structure by AMP-activated protein kinase. Nature 2007;447:1017–20. [6] McCarty MF. Chronic activation of AMP-activated kinase as a strategy for slowing aging. Med Hypotheses 2004;63:334–9. [7] Habinowski SA, Witters LA. The effects of AICAR on adipocyte differentiation of 3 T3L1 cells. Biochem Biophys Res Commun 2001;286:852–6. [8] Dagon Y, Avraham Y, Berry EM. AMPK activation regulates apoptosis, adipogenesis, and lipolysis by eIF2alpha in adipocytes. Biochem Biophys Res Commun 2006;340:43–7. [9] Fulco M, Cen Y, Zhao P, Hoffman EP, McBurney MW, Sauve AA, et al. Glucose restriction inhibits skeletal myoblast differentiation by activating SIRT1 through AMPKmediated regulation of Nampt. Dev Cell 2008;14:661–73. [10] Kasai T, Bandow K, Suzuki H, Chiba N, Kakimoto K, Ohnishi T, et al. Osteoblast differentiation is functionally associated with decreased AMP kinase activity. J Cell Physiol 2009;221:740–9. [11] Li X, Han Y, Pang W, Li C, Xie X, Shyy JY, et al. AMP-activated protein kinase promotes the differentiation of endothelial progenitor cells. Arterioscler Thromb Vasc Biol 2008;28:1789–95. [12] Akiyama H. Control of chondrogenesis by the transcription factor Sox9. Mod Rheumatol 2008;18:213–9. [13] von der Mark K, Kirsch T, Nerlich A, Kuss A, Weseloh G, Gluckert K, et al. Type X collagen synthesis in human osteoarthritic cartilage. Indication of chondrocyte hypertrophy. Arthritis Rheum 1992;35:806–11. [14] Boos N, Nerlich AG, Wiest I, von der Mark K, Ganz R, Aebi M. Immunohistochemical analysis of type-X-collagen expression in osteoarthritis of the hip joint. J Orthop Res 1999;17:495–502. [15] Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. [16] Elima K, Vuorio E. Expression of mRNAs for collagens and other matrix components in dedifferentiating and redifferentiating human chondrocytes in culture. FEBS Lett 1989;258:195–8. [17] Bi W, Huang W, Whitworth DJ, Deng JM, Zhang Z, Behringer RR, et al. Haploinsufficiency of Sox9 results in defective cartilage primordia and premature skeletal mineralization. Proc Natl Acad Sci U S A 2001;98:6698–703. [18] Ng LJ, Wheatley S, Muscat GE, Conway-Campbell J, Bowles J, Wright E, et al. SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev Biol 1997;183:108–21. [19] Lefebvre V, Li P, de Crombrugghe B. A new long form of Sox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis and cooperatively activate the type II collagen gene. EMBO J 1998;17:5718–33. [20] Smits P, Li P, Mandel J, Zhang Z, Deng JM, Behringer RR, et al. The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Dev Cell 2001;1:277–90. [21] Akiyama H, Chaboissier MC, Martin JF, Schedl A, de Crombrugghe B. The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 2002;16:2813–28. [22] Atsumi T, Miwa Y, Kimata K, Ikawa Y. A chondrogenic cell line derived from a differentiating culture of AT805 teratocarcinoma cells. Cell Differ Dev 1990;30:109–16.
133
[23] Wroblewski J, Engstrom M, Edwall-Arvidsson C, Sjoberg G, Sejersen T, Lendahl U. Distribution of nestin in the developing mouse limb bud in vivo and in micromass cultures of cells isolated from limb buds. Differentiation 1997;61:151–9. [24] Bohensky J, Leshinsky S, Srinivas V, Shapiro IM. Chondrocyte autophagy is stimulated by HIF-1 dependent AMPK activation and mTOR suppression. Pediatr Nephrol 2010;25:633–42. [25] Asahina I, Sampath TK, Hauschka PV. Human osteogenic protein-1 induces chondroblastic, osteoblastic, and/or adipocytic differentiation of clonal murine target cells. Exp Cell Res 1996;222:38–47. [26] Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 2012;9:671–5. [27] Bandow K, Nishikawa Y, Ohnishi T, Kakimoto K, Soejima K, Iwabuchi S, et al. Lowintensity pulsed ultrasound (LIPUS) induces RANKL, MCP-1, and MIP-1beta expression in osteoblasts through the angiotensin II type 1 receptor. J Cell Physiol 2007; 211:392–8. [28] Ushita M, Saito T, Ikeda T, Yano F, Higashikawa A, Ogata N, et al. Transcriptional induction of SOX9 by NF-kappaB family member RelA in chondrogenic cells. Osteoarthritis Cartilage 2009;17:1065–75. [29] Ikeda T, Saito T, Ushita M, Yano F, Kan A, Itaka K, et al. Identification and characterization of the human SOX6 promoter. Biochem Biophys Res Commun 2007;357: 383–90. [30] Christy B, Nathans D. DNA binding site of the growth factor-inducible protein Zif268. Proc Natl Acad Sci U S A 1989;86:8737–41. [31] Lemaire P, Vesque C, Schmitt J, Stunnenberg H, Frank R, Charnay P. The seruminducible mouse gene Krox-24 encodes a sequence-specific transcriptional activator. Mol Cell Biol 1990;10:3456–67. [32] Hawley SA, Fullerton MD, Ross FA, Schertzer JD, Chevtzoff C, Walker KJ, et al. The ancient drug salicylate directly activates AMP-activated protein kinase. Science 2012; 336:918–22. [33] Tan L, Peng H, Osaki M, Choy BK, Auron PE, Sandell LJ, et al. Egr-1 mediates transcriptional repression of COL2A1 promoter activity by interleukin-1beta. J Biol Chem 2003;278:17688–700. [34] Rockel JS, Bernier SM, Leask A. Egr-1 inhibits the expression of extracellular matrix genes in chondrocytes by TNFalpha-induced MEK/ERK signalling. Arthritis Res Ther 2009;11:R8. [35] Vincenti MP, Brinckerhoff CE. Early response genes induced in chondrocytes stimulated with the inflammatory cytokine interleukin-1beta. Arthritis Res 2001;3:381–8. [36] Andrade J, Quinn J, Becker RZ, Shupnik MA. AMP-activated protein kinase is a key intermediary in GnRH-stimulated LHbeta gene transcription. Mol Endocrinol 2013;27: 828–39. [37] Kubosaki A, Tomaru Y, Tagami M, Arner E, Miura H, Suzuki T, et al. Genome-wide investigation of in vivo EGR-1 binding sites in monocytic differentiation. Genome Biol 2009;10:R41. [38] Grabe N. AliBaba2: context specific identification of transcription factor binding sites. In Silico Biol 2002;2:S1–S15. [39] Wang T, Yu Q, Chen J, Deng B, Qian L, Le Y. PP2A mediated AMPK inhibition promotes HSP70 expression in heat shock response. PLoS One 2010;5. [40] Zhang J, Bowden GT. UVB irradiation regulates Cox-2 mRNA stability through AMPK and HuR in human keratinocytes. Mol Carcinog 2008;47:974–83. [41] Petursson F, Husa M, June R, Lotz M, Terkeltaub R, Liu-Bryan R. Linked decreases in liver kinase B1 and AMP-activated protein kinase activity modulate matrix catabolic responses to biomechanical injury in chondrocytes. Arthritis Res Ther 2013;15:R77. [42] Terkeltaub R, Yang B, Lotz M, Liu-Bryan R. Chondrocyte AMP-activated protein kinase activity suppresses matrix degradation responses to proinflammatory cytokines interleukin-1beta and tumor necrosis factor alpha. Arthritis Rheum 2011;63: 1928–37.
