October 2011

Regular Article

Biol. Pharm. Bull. 34(10) 15891595 (2011)

1589

Overexpression of Hematopoietically Expressed Homeoprotein Induces Nonapoptotic Cell Death in Mouse Prechondrogenic ATDC5 Cells
Riyo MORIMOTO* and Akiko OBINATA
Laboratory of Pharmaceutical Science, Faculty of Physiological Chemistry II, Teikyo University; 10911 Suwarashi, Midori-ku, Sagamihara 2525195, Japan. Received April 21, 2011; accepted June 28, 2011; published online July 7, 2011 Physiological cell death is an essential event in normal development and maintenance of homeostasis. Recently, the morphological and pharmacological characteristics of programmed cell death, which are distinct from those of apoptosis under physiological and pathological conditions, have been reported. However, the molecular mechanism and executioner of this type of cell death are unknown. We show that overexpression of hematopoietically expressed homeoprotein (Hex), a homeoprotein of divergent type, and enhanced green uorescent protein (EGFP) fusion protein (Hex-EGFP) induces cell death in mouse chondrogenic cell line ATDC5. The expression rate of Hex-EGFP decreased more rapidly than that of EGFP 96 h after transfection. The time-lapse image of living cells revealed the Hex-EGFP-positive cells rapidly died in a necrosis-like fashion. The nuclei of Hex-EGFPexpressing cells were rarely fragmented; however, these cells were negative for terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) staining. The expression rate of HexEGFP clearly increased by treatment with radical scavengers, propyl gallate and butylated hydroxyanisole, slightly increased with a caspase inhibitor, zVAD-fmk, and was not affected by N-acetyl cysteine in ATDC5 cells. A uorescent probe indicated that reactive oxygen species (ROS) were localized near the nuclei in Hex-EGFPpositive cells. In differentiated ATDC5 cells, as hypertrophic chondrocyte-like cells, the expression rate of HexEGFP increased above that in uninduced ATDC5 cells. These results suggest that Hex induces nonapoptotic cell death through local accumulation of reactive oxygen species, and mature chondrocytes, which express Hex, might be able to escape cell death induced by Hex in cartilage.
Key words hematopoietically expressed homeoprotein; cell death; reactive oxygen species; chondrocyte

Programmed cell death plays an important role in embryonic development, immune regulation, and maintenance of homeostasis.13) Apoptosis, well-known physiological cell death, is dened by dependence on the caspase family of proteases.3) Apoptosis shows characteristic morphologies such as chromatin condensation, nuclear fragmentation, and breakdown of the cell into apoptotic bodies. Recent studies have shown that physiological cell death is clearly distinct from apoptosis in terms of morphological and pharmacological features.47) A type of nonapoptotic cell death is autophagic cell death. In autophagic cell death, massive autophagosomes and autolysosomes appear and degrade intracellular organelles.8,9) Another type of nonapoptotic programmed cell death, which is morphologically classied as neither apoptosis nor autophagic cell death (for example, mammalian developing chondrocytes after hypertrophy), is probably strictly regulated, yet the underlying molecular mechanisms are unclear.4,10,11) Hematopoietically expressed homeobox protein (Hex), also known as proline-rich homeodomain protein (Prh), is a transcription factor belonging to the group of divergent-type homeoproteins containing a homeodomain and is a key molecule for normal morphogenesis through the regulation of cell proliferation and differentiation in hematopoietic cells, vascular system, liver, thyroid, heart, pancreas, and skin during development in a tissue- and differentiation-stagedependent manner.1217) In hematopoiesis, Hex is required for the differentiation of hemangioblasts into hematopoietic progenitors and the repression of hemangioblast proliferation.18,19) Hex is essential for normal liver morphogenesis by promoting the differentiation of hepatic endoderm into hepatoblasts through the transition to a pseudostratied epithelium, and by the activation of several liver-specic genes
To whom correspondence should be addressed.

such as hepatocyte nuclear factors.2023) Exogenous Hex expression in avian embryonic skin leads to proliferation, resulting in induction of feather buds via Wnt signal or elongation of scales.2426) We reported that homeoprotein Hex is expressed in chondrocytes of embryonic mice and differentiated chondrogenic cell line ATDC5 cells; however, its function in cartilage development and subsequent bone formation remains unclear.27) Here we show that overexpression of Hex-enhanced green uorescent protein (EGFP) fusion gene induced nonapoptotic cell death in mouse chondroprogenitor cell line ATDC5 cells in an uninduced state. The cell death by HexEGFP overexpression was inhibited by treatment with radical scavengers, and the cells exhibited the accumulation of reactive oxygen species (ROS) near the nucleus. Their cell deaths were more reduced in differentiating ATDC5 cells than in undifferentiated cells. These results suggest that hypertrophic chondrocytes can escape cytotoxicity by Hex in vivo. MATERIALS AND METHODS Reagent zVAD-fmk and propyl gallate (PG), and butylated hydroxyanisole (BHA) were purchased from Calbiochem (Darmstadt, Germany) and Sigma-Aldrich (St. Louis, MO, U.S.A.), respectively. RedoxSensor Red CC-1 was purchased from Life Technologies Corporation (Carlsbad, CA, U.S.A.). N-Acetylcysteine (NAC) was purchased from Sigma. Cell Culture Mouse chondrogenic cell line ATDC5 was obtained from the Riken Cell Bank (Tsukuba, Japan). ATDC5 cells were maintained as described previously.2729) The cells were cultured in maintenance medium consisting of Dulbeccos modied Eagles medium/Hams F-12 (1 : 1) (Life
2011 Pharmaceutical Society of Japan

e-mail: r-morimo@pharm.teikyo-u.ac.jp

1590

Vol. 34, No. 10

Technologies Corporation, Carlsbad, CA, U.S.A.) containing 5% fetal bovine serum (FBS), 10 m g/ml human transferrin (Roche Applied Bioscience, Basel, Switzerland), and 30 nm sodium selenite (Sigma) at 37 C in a humidied atmosphere of 5% CO2 in air. For chondrogenic induction followed by hypertrophy induction, the cells were cultured in the maintenance medium supplemented with 10 m g/ml bovine insulin (Sigma) under 5% CO2 for 3 weeks, and subsequently the culture medium was switched to alpha modied Eagles medium containing 5% FBS plus 10 m g/ml insulin, 10 m g/ml transferrin, and 30 nM sodium selenite (mineralizing medium), and the CO2 concentration was shifted to 3% to facilitate mineralization in culture after 3 weeks. The medium was replaced every other day. Expression Vector pEGFP-N3 and pcDNA3.1-Hyg ( ) plasmids were purchased from Life Technologies. Mouse Hex cDNA was amplied by polymerase chain reaction and the cDNA containing the open reading frame region of Hex gene was subcloned into XhoI and BamHI site and EcoRV site in pEGFP-N3 and pcDNA3.1, respectively. The sequence of Hex cDNA was checked by sequencing using an ABI3130 genetic analyzer (Life Technologies). Transfection Transient overexpression of Hex-EGFP fusion gene and intact Hex protein in uninduced ATDC5 cells was achieved by lipofection using Lipofectamine LTX (Life Technologies) in combination with PLUS reagent (Life Technologies). ATDC5 cells were inoculated 24 h before transfection. Transfection was performed according to the manufacturers instructions. The medium was changed at 5 h posttransfection, and the various inhibitors were added to the medium at the time. The mineralizing ATDC5 cells were transfected by electroporation. Briey, the mineralizing ATDC5 cells were dispersed by incubation with 0.02% trypsin (BD Difco, NJ, U.S.A.) and 0.2% type I collagenase (SIGMA) in Ca2 - and Mg2 -free Tyrodes solution (CMF-T) for 4 h at 37 C. The control ATDC5 cells, which were cultured in the medium without insulin for 42 d, were incubated for 20 min in the digesting solution. After digestion, the cells were washed with CMF-T three times followed by washing with Opti-MEM (Life Technologies) twice. The cells (1 106 cells) and plasmid (10 m g) were mixed and pulsed at 175 V for 10 ms, which was repeated ten times, using an electroporation system CUY21 Pro-Vitro (NEPAGENE, Chiba, Japan). Determination of Gene Expression Rate Ninety-nine hours after transfection, the ATDC5 cells were xed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min. The cells were washed with PBS and stained with 10 m g/ml 4 6-diamidino-2-phenylindole (DAPI) for 15 min. The preparation was randomly photographed, and the numbers of EGFP-positive cells and total cells were counted. The expression rate was determined by observation of more than 600 cells. In the case of treatment with various inhibitors, the inhibitor was added to the culture media 5 h after transfection. Time-Lapse Imaging of Living Cells Cells inoculated on a glass-bottomed dish (Asahi Glass Co., Ltd.) followed by transfection were chronologically observed using a confocal laser microscope with a built-in CO2 incubator FV10i (OLYMPUS, Tokyo, Japan) at 30 min intervals. The incubator was kept at 37 C in a humidied atmosphere of 5% CO2

in air. Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick-End Labeling (TUNEL) Assay TUNEL staining was performed using an in situ cell death kit (Roche) according to the instruction manual. For positive control of TUNEL staining, we used the sample after xation and subsequent permeabilization treated with 30 U/ml DNase I (Roche) for 10 min at room temperature for partial digestion of genomic DNA. Detection of ROS ROS was visualized using Redox Sensor CC-1 Red (Life Technologies). At 48 h post-transfection, cells were loaded into fresh medium containing 5 m M Redox Sensor CC-1 Red for 10 min at 37 C and were washed twice with PBS. The uorescence image was observed under a uorescent microscope (Olympus Corp., Tokyo, Japan). RESULTS Overexpression of Hex-EGFP Gene Induces Cell Death To investigate the physiological role of Hex in chondrogenesis, we tried to establish the ATDC5 cell line stably expressing Hex gene; however, a resistant colony did not appear after drug selection (data not shown). Uninduced ATDC5 cells do not express Hex protein. Inducing chondrocyte-like cells by addition of 10 m g/ml insulin in the medium, the ATDC5 cells express Hex and the amount of Hex protein increases.27) First, we observed the course of Hex-EGFP expression in uninduced ATDC5 cells. EGFP-expressing cells were a positive control for transfection and expression persistence. After transfection, the rates of Hex-EGFP and EGFP-expressing cells peaked at 48 h post-transfection in uninduced ATDC5 cells (data not shown). The expression of Hex-EGFP rapidly disappeared 96 h after transfection although expression of EGFP gradually decreased (Figs. 1A, B). The rate of Hex-EGFP-expressing cells was about 5fold lower than that of EGFP-expressing cells at 48 h posttransfection. The reason why gene expression rate differed between the individual genes is that protein synthesis is affected by transfection and translation efciency, mRNA and protein stability, and cell viability. The cells expressing exogenous intact Hex protein also exhibited an expression pattern similar to that of EGFP-tagged Hex protein (data not shown). Hex-EGFP protein was localized in the nuclei (Fig. 1C). Subsequently, we investigated the hypothesis that Hex protein is rapidly degraded by proteasome in uninduced ATDC5 cells because Hex was found to interact with proteasome subunit C8 in human leukemia cell line K567.30) Since the expression rate of Hex-EGFP increased similarly to the rate of EGFP cells by treatment with a proteasome inhibitor, MG-132, we concluded that Hex-EGFP was not specically degraded by proteasome (data not shown). Live-cell imaging of Hex-EGFP-expressing cells revealed that the nuclei of Hex-EGFP-positive cells shrank, and then the uorescence of Hex-EFGP protein diffused to the cytoplasmic region, and was thereafter eliminated (Fig. 2A). Seventy percent of HexEGFP-expressing ATDC5 cells died in the 20 h from 48 h post-transfection, although 10% of EGFP-positive cells died (Fig. 2B). Next, we checked whether the cell death induced by HexEGFP is apoptosis because the nuclei of the cells rarely frag-

October 2011

1591

Fig. 1. Time Course of the ATDC5 Cells Transiently Expressing HexEGFP in Uninduced ATDC5 Cells
(A) Images of population of Hex-EGFP- (upper panels) and EGFP-expressing cells (lower panels). The cells were xed and stained with DAPI every 48 h. Scale bar 50 m m. (B) The expression rates of Hex-EGFP and EGFP. These rates were determined by observation of more than 200 cells in photographs at random. The results are shown as means S.D., n 3. (C) Enlarged image of Hex-EGFP stained with DAPI. Scale bar 10 m m.

mented. The fragmented nuclei of Hex-EGFP-positive cells were negative for TUNEL staining, which is used to detect apoptotic cells (Fig. 3, red). The control of TUNEL staining treated with DNase I is positive (data not shown). DNA staining with DAPI of the cells also indicated that chromatin condensation had not occurred (Fig. 3, blue). These results indicated that overexpression of Hex-EGFP gene caused nonapoptotic cell death in uninduced ATDC5 cells. Overexpression of Hex-EGFP Gene Induces Accumulation of ROS Near the Nucleus To reveal the molecular mechanism of the cell death induced by Hex-EGFP overexpression, we characterized the pharmacological properties of the cell death using various inhibitors. ROS are regulators of redox-sensitive cell signal pathway, and are associated with physiological and pathological cell death, such as ischemia and hypoxia.7,8,31) Propyl gallate (PG), a free radical scavenger, dramatically increased the expression rate of HexEGFP after 48 h post-transfection (Fig. 4). A radical scavenger, butylated hydroxyanisole (BHA), also increased the rate of Hex-EGFP-positive cells, whereas a caspase inhibitor, zVAD-fmk, slightly elevated it and a glutathione precursor as an antioxidant, N-acetyl cysteine (NAC), did not affect it. Because all these drugs allowed the rate of EGFP-positive cells to increase 1.2-fold, the increase by treatment with PG and BHA does not have a nonspecic effect on transfection ef-

Fig. 2. Live-Cell Imaging of Hex-EGFP-Expressing ATDC5 Cells

(A) A series of uorescent images of Hex-EGFP-expressing ATDC5 cells are shown. The cells were monitored using a confocal laser microscope with a built-in CO2 incubator FV10i at 30 min intervals from 48 h post-transfection. Arrows and numbers written in the bottom left corner of each gure indicate dying cells and periods from 48 h posttransfection, respectively. Scale bar 10 m m. (B) Mortality of Hex-EGFP-expressing ATDC5 cells from 48 to 68 h post-transfection. The mortality was calculated by observation of more than 20 EGFP-positive cells for each experiment. The results are shown as means S.D., n 6. Asterisk indicates that a value of Hex-EFGP is signicantly different from the value of EGFP on each day with Students t-test (p 0.00005).

ciency. The nuclei of Hex-EGFP-expressing cells treated with zVAD-fmk shrank; on the other hand, those of HexEGFP-expressing cells treated with PG and BHA exhibited normal morphology (Fig. 4A, inserts). A uorescent indicator for ROS, Redox Sensor CC-1 Red, showed the accumula-

1592

Vol. 34, No. 10

Fig. 3. TUNEL Staining of Hex-EGFP-Expressing ATDC5 Cells

The cells expressing Hex-EGFP were xed at 48 h post-transfection and then subjected to TUNEL staining. Arrows and arrowheads indicate the fragmented and shrinking nuclei, respectively. Lower panels were enlarged to show a Hex-EGFP-expressing cell with a fragmented nucleus. DNA was visualized by DAPI. Scale bar 10 m m.

tion of ROS near the nucleus in the Hex-EGFP-expressing cells (Fig. 5). The nuclei of Hex-EGFP-expressing cells were exed and appeared to surround ROS. Differentiation of ATDC5 Cells into Hypertrophic Chondrocytes Reduces the Cell Death by Hex-EGFP Overexpression We are most interested in whether and how Hex is associated with chondrogenesis, followed by mineralization. The expression level of Hex protein was found to be elevated during chondrocyte maturation in embryonic mouse and differentiating ATDC5 cells, and Hex protein was localized in the nuclei in prehypertrophic chondrocytes.27) To examine whether Hex induces cell death in vivo, we transfected Hex gene into chondrocyte-like cells differentiated from ATDC5 cells. ATDC5 cells can differentiate into early chondrocyte-like cells by treatment with 10 m g/ml insulin in the maintenance medium under 5% CO2 and subsequently deposit mineral by treatment with 10 m g/ml insulin in mineralizing medium under 3% CO2; therefore, these cells are a good model of chondrogenesis and subsequent mineralization in vitro.28,29) The rate of Hex-EGFP-expressing cells among the cells cultured with insulin was about ten times more than that in the cells cultured without insulin at 96 h after transfection (Fig. 6). It is noteworthy that the expression rate of EGFP in mineralized ATDC5 cells elevated 1.3 times more than that in uninduced cells; thus, differentiation to mineralizing ATDC5 cells causes an increase in transfection efciency. We conclude that Hex-EGFP overexpression induced nonapoptotic cell death in uninduced ATDC5 cells, but not in the differentiated cells. DISCUSSION We reported that Hex is expressed in chondrocytes of murine embryos; however, the physiological role of Hex is still unknown. Here we show that Hex-EGFP overexpression causes nonapoptotic cell death in uninduced ATDC5 cells, and that this cell death is inhibited by treatment with radical scavengers PG and BHA. Hex-EGFP overexpression causes the accumulation of ROS near the nucleus. The cell death induced by Hex-EGFP is reduced in the differentiated ATDC5 cells compared with that in uninduced cells. These results suggest that hypertrophic chondrocytes can escape cytotoxicity by Hex in cartilage in vivo. This is the rst evidence of the function and the molecular pathway of Hex-EGFP in car-

tilage. Nevertheless, data that Hex-EGFP overexpression in embryonic chick skin promotes cell proliferation have been reported.2426) We suspect that this inconsistency results from species- and differentiation-stage-specic effects. The cell death induced by Hex-EGFP overexpression exhibited neither chromatin condensation nor TUNEL positivity, although the cell death was associated with nuclear shrinkage but rarely with fragmentation of the nucleus. Livecell imaging showed that the cell death induced by HexEGFP occurs in a necrosis-like fashion because of disruption of the nuclear and plasma membrane, which is indicated by efux of the uorescence of Hex-EGFP to cytoplasm followed by that to the extracellular region. Hence, we conclude that the cell death by Hex-EGFP is nonapoptotic cell death. Our results are supported by the previous report by Noy et al. that human Hex-EGFP overexpression increased the number of annexin V /propidium iodide (PI) cells but did not affect the number of annexin V /PI cells among human leukemia K562 cells.32) Annexin V a protein that binds to phos, phatidylserine with high afnity, is usually used as a probe for the early phase of apoptosis.33,34) PI is a membrane-impermeable uorescent dye that is generally excluded in viable cells. The annexin V /PI cells are dened as apoptotic cells. The population of annexin V /PI includes not only late apoptotic cells but also necrotic cells; thus, the preceding data suggest that Hex induces programmed cell death excluding apoptosis.35) Noy and colleagues also indicated that knockdown of Hex gene increased cell proliferation, and the increase was enhanced by stimulation with VEGF in K562 cells.32) However, the precise molecular mechanism and the executioner of the cell death by Hex were not described in the report. The cell death induced by Hex-EGFP overexpression was inhibited by radical scavengers PG and BHA, and the accumulation of ROS near the nucleus in Hex-EGFP-expressing ATDC5 cells indicates the strong correlation between the cell death by Hex-EGFP overexpression and ROS accumulation. ROS are associated with several cell death processes in the pathogenesis of diseases, including ischemic injury, neurodegeneration, and viral infection.7,8,29,36) The morphological characteristics of these cell deaths were distinct from those of apoptosis, and these cell deaths were not inhibited by caspase inhibitors.5,7,8) ROS are enzymatically generated by NADPH oxidase, the mitochondrial electron transport system, xanthine oxidase, cytochrome p450, and nitric oxide synthase.37) NADPH oxidase consists of multiple subunits and their expression levels are dependent on cell types and environmental context. NADPH oxidase 2 (NOX2) and 4, which are catalytic subunits of NADPH oxidase, are expressed in primary mouse chondrocytes and ATDC5 cells, and NOX2 and NOX4 are required for chondrogenic differentiation.38) Indeed, ROS are elevated in the hypertrophic zone in cartilage, and articially increasing ROS level induces chondrocyte hypertrophy.39) These reports indicate that ROS are essential for chondrogenesis and subsequent hypertrophy. Hex overexpression inuenced the cellular localization of ROS rather than the amount of ROS in the cells. The accumulation of ROS near the nucleus may cause oxidization of the nuclear membrane and subsequent cytotoxicity in Hex-EGFP-overexpressing cells. In fact, the nuclear membrane of Hex-EGFP-expressing cell

October 2011

1593

Fig. 4. The Effect of Various ROS Inhibitors on the Expression Rate of Hex-EGFP in Uninduced ATDC5 Cells
(A) Image of Hex-EGFP- or EGFP-expressing cells treated with various inhibitors. The ATDC5 cells transfected with Hex-EGFP were cultured with PG (50 m M), BHA (100 m M), NAC (1 mM), and zVAD-fmk (5 m M) from 5 to 96 h after transfection. The inserts show a typical pattern of Hex-EGFP-expressing cell treated with the inhibitor. Scale bar 20 m m. (B) The rates of Hex-EGFP-expressing cells at 48 and 96 h post-transfection. The rate of EGFP-positive cells was determined by observation of more than 400 cells. These results upon treatment with drugs indicate the % rates of Hex-EGFP- or EGFP-positive cells treated with vehicle. These rates were determined by observation of more than 2000 cells in photographs at random. The values are shown as means S.D., n 35.

was disrupted before breakdown of plasma membrane in live-cell imaging. We are most interested in how Hex is involved in chondrogenesis, followed by mineralization. In cartilage of embryonic mouse and differentiating ATDC5 cells, the expression

level of Hex protein is elevated and Hex protein is imported into the nucleus in the prehypertrophic stage.27) Our data suggest that Hex can function as a transcription factor in prehypertrophic chondrocytes without cytotoxicity. In this study, the molecular mechanism avoiding cytotoxicity by ROS in

1594

Vol. 34, No. 10

Fig. 5. Visualization of ROS in Hex-EGFP-Expressing ATDC5 Cells

The Hex-EGFP- or EGFP-expressing ATDC5 cells were incubated with 5 m M Redox Sensor Red CC-1 at 48 h after transfection. The uorescence as a consequence of reaction of Redox Sensor Red CC-1 with ROS was observed by uorescent microscopy. Arrows indicate that the nuclei exed and appeared to surround ROS of Hex-EGFP-expressing cells. Scale bar 20 m m.

tive cells for biochemical analysis, we are constructing an inducible expression system of intact Hex protein by treatment with doxycycline as an analog of tetracycline. We have preliminary data using this inducible Hex expression system that support our idea that chondrogenic differentiation avoids the cytotoxicity caused by Hex. Hex-protein-expressing cells induced by treatment with doxycycline remained at 5 d after cultivation in differentiating conditions. On the other hand, Hex-expressing cells were not detected at the same time when the cells were cultured in undifferentiating conditions (data not shown). Further precise analysis of the mechanism must be undertaken. Recent reports suggest that nonapoptotic cell death including autophagic cell death is a backup system of apoptosis. Inhibition of apoptosis by caspase inhibitor or by RNA interference inhibition of the executing molecule of apoptosis induces autophagic cell death or necrosis-like cell death instead of apoptosis.7) Our research might not only contribute to elucidation of the physiological roles of Hex in cartilage but also lead to new ndings in association with the molecular mechanism of nonapoptotic cell death. Acknowledgment We thank Yusuke Yabuta for technical support in the construction of expression vectors. This work was supported by a Scientic Research Grant [19790073 to R.M.] from the Ministry of Education, Culture, Sports, Science and Technology of Japan. REFERENCES
1) 2) 3) 4) 5) 6) 7) 8) 9) Degterev A., Boyce M., Yuan J., Oncogene, 22, 85438567 (2003). Jacobson M. D., Weil M., Raff M. C., Cell, 88, 347354 (1997). Danial N. N., Korsmeyer S. J., Cell, 116, 205219 (2004). Clarke P. G., Anat. Embryol. (Berlin), 181, 195213 (1990). Leist M., Jttel M., Nat. Rev. Mol. Cell Biol., 2, 589598 (2001). Bialik S., Zalckvar E., Ber Y., Rubinstein A. D., Kimchi A., Trends Biochem. Sci., 35, 556564 (2010). Vandenabeele P., Galluzzi L., Vanden Berghe T., Kroemer G., Nat. Rev. Mol. Cell Biol., 11, 700714 (2010). Tsujimoto Y., Shimizu S., Cell Death Differ., 12 (Suppl. 2), 1528 1534 (2005). Bursch W., Karwan A., Mayer M., Dornetshuber J., Frhwein U., Schulte-Hermann R., Fazi B., Di Sano F., Piredda L., Piacentini M., Petrovski G., Fss L., Gerner C., Toxicology, 254, 147157 (2008). Schweichel J. U., Merker H. J., Teratology, 7, 253266 (1973). Prez H. E., Luna M. J., Rojas M. L., Kouri J. B., Apoptosis, 29, 265 277 (2004). Crompton M. R., Bartlett T. J., MacGregor A. D., Manoletti G., Buratti E., Giancotti V Goodwin G. H., Nucleic Acids Res., 20, 5661 ., 5667 (1992). Bedford F. K., Ashworth A., Enver T., Wiedemann L. M., Nucleic Acids Res., 21, 12451249 (1993). Hromas R., Radich J., Collins S., Biochem. Biophys. Res. Commun., 195, 976983 (1993). Ho C. Y., Houart C., Wilson S. W., Stainier D. Y., Curr. Biol., 9, 11311134 (1999). Tanaka T., Inazu T., Yamada K., Myint Z., Keng V W., Inoue Y., . Taniguchi N., Noguchi T., Biochem. J., 339, 111117 (1999). Sou A., Jayaraman P. S., Biochem. J., 412, 399413 (2008). Guo Y., Chan R., Ramsey H., Li W., Xie X., Shelley W. C., MartinezBarbera J. P., Bort B., Zaret K., Yoder M., Hromas R., Blood, 102, 24282435 (2003). Kubo A., Chen V Kennedy M., Zahradka E., Daley G. Q., Keller G., ., Blood, 105, 45904597 (2005). Keng V W., Yagi H., Ikawa M., Nagano T., Myint Z., Yamada K., . Tanaka T., Sato A., Muramatsu I., Okabe M., Sato M., Noguchi T., Biochem. Biophys. Res. Commun., 276, 11551161 (2000).

10) 11) Fig. 6. Differentiation of ATDC5 Cells into Mineralized ChondrocyteLike Cells Increases the Expression Rate of Hex-EGFP
ATDC5 cells were differentiated as described in Materials and Methods. Undifferentiated cells were cultured in the same conditions for differentiation without insulin. The cells were transfected with Hex-EGFP or EGFP by electroporation, with subsequent observation at 96 h after transfection. The expression rate was determined by observation of more than 500 cells in photographs at random. The results are shown as means S.D., n 3. Asterisk indicates that a value of Hex-EFGP is signicantly different from the value of EGFP on each day with Students t-test ( p 0.05, p 0.005). Scale bar 20 m m.

12)

13) 14) 15) 16)

differentiated ATDC5 cells is still unclear. We consider two possibilities for this mechanism: one is the alteration of localization of ROS to prevent the perinuclear accumulation during differentiation. The other is acquisition of tolerance to ROS in differentiated ATDC5 cells. We have used a transient transfection system for exogenous gene expression and monitored EGFP protein as a tag. To separate the tag and to increase the number of Hex-posi-

17) 18)

19) 20)

October 2011 21) Denson L. A., Karpen S. J., Bogue C. W., Jacobs H. C., Am. J. Physiol. Gastrointest. Liver Physiol., 279, G347G355 (2000). 22) Tanaka H., Yamamoto T., Ban T., Satoh S., Tanaka T., Shimoda M., Miyazaki J., Noguchi T., Arch. Biochem. Biophys., 442, 117124 (2005). 23) Bort R., Signore M., Tremblay K., Martinez Barbera J. P., Zaret K. S., Dev. Biol., 290, 4456 (2006). 24) Obinata A., Akimoto Y., Omoto Y., Hirano H., Dev. Growth Differ., 44, 281292 (2002). 25) Obinata A., Akimoto Y., Int. J. Dev. Biol., 49, 885890 (2005). 26) Obinata A., Akimoto Y., Int. J. Dev. Biol., 49, 953960 (2005). 27) Morimoto R., Yamamoto A., Akimoto A., Obinata A., J. Biochem., 50, 6171 (2011). 28) Shukunami C., Shigeno C., Atsumi T., Ishizeki K., Suzuki F., Hiraki Y., J. Cell Biol., 133, 457468 (1996). 29) Shukunami C., Ishizeki K., Atsumi T., Ohta Y., Suzuki F., Hiraki Y., J. Bone Miner. Res., 12, 11741188 (1997). 30) Bess K. L., Swingler T. E., Rivett A. J., Gaston K., Jayaraman P. S., Biochem. J., 374, 667675 (2003).

1595 31) Kubota C., Torii S., Hou N., Saito N., Yoshimoto Y., Imai H., Takeuchi T., J. Biol. Chem., 285, 667674 (2010). 32) Noy P., Williams H., Sawasdichai A., Gaston K., Jayaraman P. S., Mol. Cell. Biol., 30, 21202134 (2010). 33) Homburg C. H., de Haas M., von dem Borne A. E., Verhoeven A. J., Reutelingsperger C. P., Roos D., Blood, 85, 532540 (1995). 34) Verhoven B., Schlegel R. A., Williamson P., J. Exp. Med., 182, 1597 1601 (1995). 35) Vinken M., Decrock E., De Vuyst E., Leybaert L., Vanhaecke T., Rogiers V Altern. Lab. Anim., 37, 209218 (2009). ., 36) Niizuma K., Endo H., Chan P. H., J. Neurochem., 109 (Suppl. 1), 133138 (2009). 37) Griendling K. K., Sorescu D., Ushio-Fukai M., Circ. Res., 86, 494 501 (2000). 38) Kim K. S., Choi H. W., Yoon H. E., Kim I. Y., J. Biol. Chem., 285, 4029440302 (2010). 39) Morita K., Miyamoto T., Fujita N., Kubota Y., Ito K., Takubo K., Miyamoto K., Ninomiya K., Suzuki T., Iwasaki R., Yagi M., Takaishi H., Toyama Y., Suda T., J. Exp. Med., 204, 16131623 (2007).