CG&D
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Beck, G. R.
Right arrow Articles by Moran, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Beck, G. R., Jr.
Right arrow Articles by Moran, E.
Cell Growth & Differentiation Vol. 12, 61-83, February 2001
© 2001 American Association for Cancer Research

Gene Array Analysis of Osteoblast Differentiation1

George R. Beck, Jr.2,3, Brad Zerler and Elizabeth Moran4

Fels Institute for Cancer Research and Molecular Biology, Temple University School of Medicine, Philadelphia, Pennsylvania 19140 [G. R. B., E. M.], and CollaGenex Pharmaceuticals, Inc., Newtown, Pennsylvania 18940 [B. Z.]


   Abstract
TOP
 Abstract
Introduction
 Results and Discussion
 Materials and Methods
 References
 
We have used gene array technology to chart changes in gene expression during differentiation of the mouse calvarial-derived MC3T3-E1 cell line to an osteoblast-like phenotype. Expression was analyzed on a mouse gene array panel of 588 cDNAs representing tightly regulated genes with key roles in various biological processes. When compared with NIH3T3 fibroblasts, MC3T3-E1 cells showed generally higher expression of cyclins and Bcl-2 family members, as well as specific expression of products such as the CD44 antigen, which is consistent with their calvarial origin. MC3T3-E1 cells also showed a surprisingly high level of p53. Differentiation in MC3T3-E1 cells involves withdrawal from the cell cycle by day 7, accompanied by matrix accumulation and, ultimately, mineralization. Gene expression patterns in induced MC3T3-E1 cells generally reflected these stages. Cyclins were sharply down-regulated, and expression of certain antiproliferative factors and tissue-restricted genes was induced. Many of the observed changes, such as the induction of follistatin, bone morphogenetic protein receptor 1A, transforming growth factor ß, and matrix remodeling factors, reflect expected patterns and support the physiological relevance of the results. Other observed changes were not anticipated and offer new insight into the osteoblast differentiation process. An example is the sharp induction of the Tob antiproliferative factor, which has previously been associated specifically with terminal differentiation in muscles. Another example is the induction of the DNA damage-associated proteins EI24 and Gadd45, apparently as a normal aspect of osteoblast differentiation. The oxidative stress-induced protein A170 and the transcription factor Nrf2, which regulates metabolic responses to oxidative stress, were also induced. This response may reflect the in vivo requirement for vascularization during bone growth and fracture repair. Other induced factors include tumor necrosis factor receptor-associated factor-1 (1-TRAF), which is a nuclear factor {kappa}B activator, cellular retinoic acid-binding protein II (CRABP-II), and the transcription factors S-II, SP2, and SEF2 (ITF2/E2:2). SEF2 is the first basic helix-loop-helix protein found to be up-regulated during osteoblast differentiation. Northern blots confirm the induction of SEF2.


   Introduction
TOP
 Abstract
 Introduction
Results and Discussion
 Materials and Methods
 References
 
The ability to design effective treatments for disorders of bone formation such as osteoporosis depends on an understanding of the mechanisms regulating bone remodeling including differentiation of bone-forming cells. Much attention has been given to the expression of specific gene products in osteoblasts, but obtaining a single comprehensive profile of the changes in gene expression as cells differentiate to the osteoblast phenotype was not feasible before the introduction of gene array technology. In this report we have used a gene array approach to monitor the expression of various classes of genes in an osteoblast differentiation system. A frequently used agent of osteoblast differentiation in cell culture studies is ascorbic acid. Exposure to ascorbic acid stimulates suitable mammalian calvarial cells to deposit a collagenous extracellular matrix. Formation of the matrix is accompanied by the induction of specific genes associated with the osteoblast phenotype, such as alkaline phosphatase, osteopontin, and osteocalcin. If a source of organic phosphate is present, discrete zones of hydroxyapatite-containing mineral deposits form within the matrix. The sequence from induction to the formation of mineralized foci proceeds in a tightly regulated order over a span of 2–3 weeks. One obstacle to a broad-range study of gene expression in differentiating osteoblasts is the time span of the differentiation process. Presumably, successive waves of changes in gene expression occur throughout the 2–3 week span required for the differentiation of preosteoblasts to a mineralization phenotype. To address this question, we chose to use a membrane-localized array. Membrane-localized arrays can be probed several times, which makes it feasible to analyze multiple time points effectively.

We used a mouse cDNA expression array containing 588 cDNAs to probe the changes induced in ascorbic acid-treated MC3T3-E1 cells. These cDNAs were selected on the basis that they represent genes reported to play key roles in many different biological processes, and they are each characterized by tight transcriptional regulation. The array includes numerous growth factors, cytokines, interleukins, and their receptors, as well as key genes involved in different stages of embryonic development. MC3T3-E1 cells (1) are a newborn mouse calvarial-derived cell line capable of differentiating along the osteoblast lineage (2, 3, 4) . MC3T3-E1 cells are an established clonal line, but they maintain much of the tightly linked controls between proliferation and differentiation usually seen only in primary cells (2) . These cells thus provide an excellent and frequently used model for studying patterns of gene expression in differentiating osteoblasts.


   Results and Discussion
TOP
 Abstract
 Introduction
 Results and Discussion
Materials and Methods
 References
 
A template showing the organization of cDNA samples in the array is shown in Fig. 1Citation . The array is organized into six grids. In general, the cDNAs in grids A, B, and C are designed to analyze the expression of genes related to cell growth, whereas the cDNAs in grids D, E, and F analyze tissue-specific gene expression. A complete list of the genes assayed is shown in Table 1Citation . Noninduced MC3T3-E1 cells were originally compared with NIH3T3 cells to obtain a baseline picture of gene expression in the calvarial line. Changes in gene expression during differentiation were then probed in MC3T3-E1 cells differentiating in response to induction with ascorbic acid. To get a profile of gene expression during all phases of differentiation (proliferation, matrix accumulation, and terminal differentiation indicated by mineralization), RNA was harvested for analysis at days 0 (cells at 90% confluence), 3, 7, 14, and 21 after the addition of differentiation medium. These points cover the various phases of osteoblast differentiation and correlate with the time points at which the expression patterns of osteoblast markers have typically been studied. A time line illustrating the major phases of osteoblast differentiation is shown in Fig. 2Citation A. Northern blot surveys of genes whose expression is known to be regulated during osteoblast differentiation are shown in Fig. 2Citation B. The changing gene expression profiles indicate the differentiation state of the cells at the approximate time points chosen for the gene array analysis. The cells are actively replicating at the 0 time point, as indicated by expression of histone H4. Histone expression declines by day 4 and is almost undetectable by day 9. Collagen type I is induced early, and its expression declines by day 14 as the mineralization phase nears completion. Osteocalcin is very tightly regulated; its appearance correlates with transition to the mineralization phase. Additional data on the state of the cells under the conditions analyzed here can be seen in Refs. 5 and 6 .



View larger version (63K):
[in this window]
[in a new window]
  		Fig. 1. Organization of cDNA samples in the Atlas mouse gene array. The array presents cDNAs corresponding to 588 genes spotted in duplicate on a nylon membrane. The DNA samples are organized into six grids, each representing a different class of regulated genes. Several plasmid and bacteriophage DNAs are included as negative controls to confirm hybridization specificity, and nine housekeeping genes are included to normalize mRNA abundance (light gray dots). Genomic DNA spots (dark gray dots) serve as orientation marks to help determine the coordinates of hybridization signals. A complete list of the genes is found in Table 1Citation .

 

View this table:
[in this window]
[in a new window]
  		Table 1 Gene expression array analysis

 


View larger version (62K):
[in this window]
[in a new window]
  		Fig. 2. Time course of osteoblast differentiation. A, the major phases of osteoblast differentiation are shown relative to a time line indicating days 0, 3, 7, and 21 after induction. Differentiating osteoblasts progress through three general phases: (a) an early proliferative period that continues for several days after confluence, with the cells exiting the cell cycle between days 3 and 7; (b) a period of collagen deposition and extracellular matrix maturation; and (c) terminal differentiation characterized by mineralization of the mature matrix. B, a Northern blot analysis of a panel of genes known to be regulated during osteoblast differentiation is shown to indicate the differentiation state of the cells at the approximate time points chosen for the gene array analysis. Expression of histone H4 (H4) indicates active replication; its disappearance signals the end of the proliferative phase. Collagen type I (Col-I) is induced early and declines as the matrix deposition phase nears completion. Osteocalcin (OSC) levels peak at about day 14. The appearance of osteocalcin correlates with the transition to the mineralization phase. Additional data on the state of the cells under the conditions analyzed in this report can be seen in Refs. 5 and 6.

 
The gene array data were quantified using a PhosphorImager. The blots contain nine negative controls including M13 mp18 (+) strand DNA, {lambda}DNA, and pUC18. These were negative for all blots analyzed. The blots also contain nine positive controls, which were averaged to create a normalization value for each blot. Final values were multiplied by a factor of 100 for ease of analysis. The normalized results for both cell lines and all time points are shown in Table 1Citation .

Gene Expression in MC3T3-E1 Preosteoblasts Compared with NIH3T3 Cells.
The MC3T3-E1 line is derived from calvarial cells, hence these cells are expected to display a level of tissue-specific gene expression even before further differentiation. Expression patterns specific to the MC3T3-E1 line compared with NIH3T3 cells can be derived from inspection of columns 1 and 2 in Table 1Citation . To facilitate this comparison, genes that show at least a moderate level of expression and vary by more than 2.5-fold between the two lines are listed in Table 2Citation . For the purposes of this study, a moderate level of expression was considered to be a value greater than 20 in Table 1Citation . These limits are arbitrary, but they allow the data to be reduced to a clearer picture of the most notable differences in gene expression. The gene array is designed to probe expression of genes that are active only under specific conditions. Only about 35–40 of the 588 genes screened are expressed at a moderate or higher level in NIH3T3 cells. Expression of 55–60 genes is apparent in MC3T3-E1 cells at day 0. The general pattern is that additional genes are activated in MC3T3-E1 cells, but there are both positive and negative differences between the two cell lines. Some of these are striking. For example, expression of the adipocyte differentiation-associated protein Pref-1 (Table 2Citation , line 35) is undetectable in MC3T3-E1 cells, although it is expressed abundantly in NIH3T3 cells. In contrast, expression of the CD44 cell surface antigen (Table 2Citation , line 29) is not detectable in normal NIH3T3 fibroblasts but is detectable in MC3T3-E1 cells at all time points. This is consistent with histology studies indicating that CD44 is expressed in osteoblasts at all stages of maturation (7) . Gene array autoradiograms illustrating some of the differences between expression in MC3T3-E1 cells and NIH3T3 cells are shown in Fig. 3Citation A. As an initial check on whether the gene array results are representative of what would be revealed by Northern probes, we used a Northern blot to analyze expression of Pref-1 and CD44 in the two cell lines. The results (Fig. 3B)Citation confirm the reciprocal differences between the two cell lines, both the striking difference in expression level of Pref-1 and the more modest difference in expression level of CD44.


View this table:
[in this window]
[in a new window]
  		Table 2 Comparison of gene expression between MC3T3-E1 cells and NIH3T3 cells

 


View larger version (48K):
[in this window]
[in a new window]
  		Fig. 3. Comparison of NIH3T3 and MC3T3-E1 cells. A, autoradiograms of the expression array signals for several genes listed in Table 2Citation and discussed in the text are shown here. Signals from positive controls glyceraldehyde-3-phosphate dehydrogenase (G3PDH) and murine ornithine decarboxylase (MOD) are also shown. Total cell RNA (5 µg) was used for analysis. The results of quantification by PhosphorImager are included in Table 1Citation . B, a Northern blot analysis of Pref-1 and CD44 is shown for comparison with the gene array results.

 
An inspection of Tables 1Citation and 2Citation reveals differences in particular classes of genes. MC3T3-E1 cells show a higher level of expression of multiple genes associated with cell cycle activity. These include many of the cyclins (grid locations A6a–A6l), and the Golgi 4-transmembrane-spanning transporter protein, which is associated with nucleotide transport. Expression of several intracellular transduction modulators (B2g–B7n) is also specifically active in MC3T3-E1 cells relative to NIH3T3 cells. These include Stat6,5 PKC-{theta}, and the Rab-2 ras-related protein gene (Table 2Citation , lines 11, 14, and 17). Several apoptosis-related proteins (C1a–C5d) are expressed at relatively high levels in MC3T3-E1 cells, particularly Bcl-2 family members (Table 2Citation , lines 19–21). The DNA synthesis, repair, and recombination proteins (C5e–C7n) do not vary much as a class. MC3T3-E1 cells begin to withdraw from the cell cycle at about day 3 after they are induced to differentiate, therefore it is expected that many of the genes from this group will be down-regulated after ascorbic acid treatment. Of the 21 genes from the A, B, and C grids that are expressed at a 2.5-fold or higher level in the MC3T3-E1 cells relative to the NIH3T3 cells, about half are sharply down-regulated when MC3T3-E1 cells are induced with ascorbic acid. These are considered further in Table 3Citation . Most of the others do not vary widely during differentiation. Only two genes from the A, B, and C grids are specifically less active in MC3T3-E1 cells. H-ras (Table 2Citation , line 22) was not detected in the calvarial cells, in contrast to N-ras (A5e), which is slightly more active in MC3T3-E1 cells. The gp130 signal-transducing subunit of the interleukin 6 receptor is also expressed at a lower level in MC3T3-E1 cells compared with NIH3T3 cells (Table 2Citation , line 23).


View this table:
[in this window]
[in a new window]
  		Table 3 Genes down-regulated during osteoblast differentiation

 
The genes from the D, E, and F grids are generally associated with tissue-specific gene expression, and they tend to reflect both positive and negative differences. Among the transcription factors and general DNA-binding proteins (D1a–D7n) included in this array, the strikingly specific expression of the adipocyte-associated differentiation factor Pref-1 in NIH3T3 cells was noted above. Expression of the Y box-binding protein YB1 is severalfold higher in NIH3T3 cells than in MC3T3-E1 cells at day 0 (Table 2Citation , line 39). However, it is up-regulated more than 3-fold in MC3T3-E1 cells at day 3 (Table 1)Citation , consistent with suggestions that it plays a role in collagen synthesis (8) . Other differences are apparent in the table. Among receptors and cell surface antigens (E1a–E7n), the granulocyte macrophage colony-stimulating factor receptor is not expressed in the calvarial cells (Table 2Citation , line 40). Calvarial cell-specific expression of the CD44 antigen was noted above. The entire panel of 21 neurotransmitter receptors (E4e–E5k) is essentially not expressed in either cell line or at any point during differentiation of the MC3T3-E1 cells. Because these genes are very specific to neuronal cells, they are an excellent internal control. Among the growth factors, cytokines, and chemokines, IGFBP-4 is specifically active in the calvarial cells (Table 2Citation , line 31), whereas IGFBP-6 is specifically more active in NIH3T3 cells (Table 2Citation , line 41), although the latter factor is induced in the calvarial cells during differentiation. Some cytoskeletal protein genes (F5e–F6e) appear to be more active in the calvarial cells. The expression patterns of proteases and inhibitors (F6f–F7n) are consistent with the matrix formation and remodeling functions of calvarial cells. For example, expression of the serine protease inhibitor homologue J6 (Hsp47; Table 2Citation , line 34) generally correlates with expression of collagens and was previously reported to be induced in MC3T3-E1 cells when collagen expression is induced (9) . Inspection of the entire data set in Table 1Citation reveals other differences between the cell lines that were below the cut-off points chosen for Table 2Citation . Though relatively small, many of these differences are likely to be real and significant. For example, they include additional differences among the matrix proteases.

Gene Expression in MC3T3-E1 Cells during Differentiation.
Differentiating MC3T3-E1 cells were analyzed at days 0, 3, 7, 14, and 21. Analyzing multiple time points gives a more active picture of changes in gene expression over the course of the differentiation process. However, the data are not easily presented as a simple list of genes that are increased or decreased. Sample autoradiographs illustrating results obtained at day 0 and day 21 are shown in Fig. 4Citation . A comprehensive picture can be derived from inspection of the full data set in Table 1Citation . The most significant changes are discussed below.



View larger version (70K):
[in this window]
[in a new window]
  		Fig. 4. Expression array analysis of MC3T3-E1 cells. Cells were harvested for RNA isolation at day 0 and day 21 and processed as described in "Materials and Methods." Total cell RNA (5 µg) was used for analysis. The autoradiograph images are shown here. The results of quantification by PhosphorImager are included in Table 1Citation .

 
Genes That Are Down-Regulated during Osteoblast Differentiation.
To obtain a capsule description of the genes that are down-regulated, genes whose relative expression level at time 0 in Table 1Citation is higher than 20 and show a change of at least 2.5-fold maintained over at least two consecutive assay points are listed in Table 3Citation . An internal control is illustrated through analysis of the cyclins. Ascorbic acid-induced MC3T3-E1 cells exit the cell cycle between days 3 and 7 of the differentiation process. If the expression array analysis is reflective of known results, expression of A- and B-type cyclins should decrease by day 7. Analysis of the cyclins included on the blots confirms this time course. Cyclins B1 and B2 are included in Table 3Citation , but the data in Table 1Citation indicate that cyclins A, B1, B2, and C all decrease between days 0 and 7. The results are presented as a graph in Fig. 5Citation . As shown in Tables 1Citation and 2Citation , several additional cyclins are expressed at relatively high levels in noninduced MC3T3-E1 cells. These include cyclins D1, D2, and G. Expression of these genes is maintained at relatively high levels throughout differentiation. Cyclin G can be induced in neuronal cells in response to BMP-4 and may play a role in the modulation of apoptosis (10) . Presumably, it plays a significant role in osteoblast differentiation as well.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5. Cyclin expression in differentiating MC3T3-E1 preosteoblasts. The results of the hybridization analyses using probes derived from RNA at days 0, 3, 7, 14, and 21 were quantified by PhosphorImager and normalized as described in the text. The graph shows cyclin A ({blacksquare}), B1 ({blacklozenge}), B2 ({blacktriangleup}), and C (•) genes.

 
Aside from the cyclins, approximately 7 other genes were down-regulated 2.5-fold or more in cells differentiating along the osteoblast line. The Bcl-2 family member Bak shows an unusually dynamic pattern. It is not detectable in NIH3T3 cells but is expressed at a fairly high level in the untreated calvarial cells. During differentiation, it is shut down completely, but by day 21, expression is restored to the previous high level. The behavior of Bak illustrates the advantage of analyzing the array at multiple time points. Another gene of particular interest in Table 3Citation is p53. The ability of p53 to control and coordinate many aspects of proliferation and apoptosis in response to various internal and external stimuli has been well studied. Regulation of p53 expression during differentiation is not generally expected. Nevertheless, the expression array indicates that p53 expression is relatively high in noninduced MC3T3-E1 cells and that it decreases markedly by day 7 of differentiation. The hybridization results for p53 at all time points are shown in Fig. 6ACitation and plotted in Fig. 6Citation B. To probe the extent to which the data from the expression array reflect synthesis of the protein product, p53 was analyzed at the level of protein synthesis in differentiating osteoblasts. The results of the immune precipitation (Fig. 6C)Citation were consistent with the pattern of RNA expression indicated in the gene array analysis. These results reveal an unexpected but potentially important role for regulation of p53 expression in calvarial cells.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 6. Comparison of p53 levels by expression array and immunoprecipitation. A, the p53 signals from the expression arrays at days 0, 3, 7, 14, and 21 of differentiation are shown in the first column. Positive controls, actin and the ribosomal protein S29, are shown in the second column. B, p53 values derived by PhosphorImager quantification of the blots and normalized as described in the text are shown graphically. C, p53 protein synthesis was measured by immunoprecipitation. The parental cells were metabolically labeled with [35S]methionine in 2-h pulses at the indicated days during differentiation (shown above the lanes). Cell lysates were prepared as described in "Materials and Methods" and immunoprecipitated with the p53 monoclonal antibody PAb421. The resulting autoradiograph is shown.

 
It is important to note that inspection of the total data set in Table 1Citation reveals many additional changes that may also be significant. These include changes in genes that are expressed at low levels or that vary less than 2.5-fold or are down-regulated at a single time point. Cyclins A and C are examples. Other patterns are also apparent. The STAT genes (B4d–B4g) are generally down-regulated, whereas other genes of this type are activated (B4h–B4j). Various changes are apparent within the intracellular kinase and G protein networks (B5f–B7n). Small changes are apparent in various apoptosis-related proteins (C1a–C5d). In general, as expected for cells withdrawing from the cell cycle, the genes that are down-regulated are from the classes represented by the A, B, and C grids, whereas genes that are up-regulated are from the D, E, and F grids. The major exception among the transcription factors is the repression of the multiple homeodomain protein gene ATBF1 (Table 3Citation , line 9). Another transcription factor, butyrate response factor 1 (D1i), is consistently down-regulated, but not within the parameters chosen for Table 3Citation .

Genes That Are Up-Regulated during Osteoblast Differentiation.
Inspection of the data in Table 1Citation indicates that the time of peak expression for the genes induced during differentiation can vary. Genes that reach a peak expression level of at least 20 and show at least a 2.5-fold increase in expression maintained over at least two consecutive time points are listed in Table 4Citation . In general, genes that are induced tend to be various transcription factor-related genes, growth factor-related genes, intracellular signaling protein genes, and antiapoptosis protein genes. Many other genes are induced at a lower level. These can be identified by inspection of the full data set in Table 1Citation . These genes tend to be from the same classes, with the addition of some cell growth inhibitors such as p21 (A7e) and p27 (A7f). A few factors associated with cell growth are activated. These include VEGF receptor 1 (A4j), the Lfc proto-oncogene (A5d), Shc (A5f), and transcription factor TFIID (B4j). Known targets of p53 are up-regulated as the cells withdraw from the cell cycle, although p53 is down-regulated. However, p53 expression in the calvarial cells is initially high and is never significantly reduced below the level seen, for example, in NIH3T3 cells. Table 4Citation highlights changes that reflect withdrawal from the cell cycle and the subsequent activation of specific growth factors, receptors, and transcription factors. Some of these are expected; others have not previously been reported in osteoblasts.


View this table:
[in this window]
[in a new window]
  		Table 4 Genes up-regulated during osteoblast differentiation

 
The Tob family of antiproliferation proteins has been linked with terminal differentiation of muscle (e.g., Refs. 11 and 12 ). Tob (Table 4Citation , line 1) is not detected in noninduced MC3T3-E1 cells. Its activation in differentiating osteoblasts has not been reported previously, but it is strongly activated with a peak expression at day 3, which suggests that it plays a significant role in the down-regulation of proliferation as the cells proceed to the mineralization phase. Transcription factor A10 (Table 4Citation , line 5), also designated EKLF, is known as an erythroid cell-specific transcription factor (13 , 14) , thus its induction in differentiating osteoblasts is also unexpected. However, the strong induction and peak expression at day 14, when osteogenic proteins are most actively transcribed, suggests that this factor may play a significant role in osteoblast-specific gene expression.

EI24 (Table 4Citation , line 2) was isolated as a gene product induced in a p53-dependent manner in NIH3T3 cells treated with the cytotoxic drug etoposide (15) . It is also induced by ionizing radiation. Its induction during normal differentiation in osteoblasts has not been reported, but it is interesting that its expression in osteoblasts roughly parallels the unusually high expression of p53. Gadd45 (Table 4Citation , line 6) is also strongly induced at later stages of differentiation. Like EI24, this product is usually regulated by p53 in response to DNA damage. An advantage of the gene array approach is its potential to reveal such unexpected patterns as the induction of these specific DNA damage response genes during normal differentiation.

The oxidative stress-induced protein A170 has been studied in macrophages (e.g., Ref. 16 ) and was recently cloned as a gene up-regulated in MC3T3-E1 after treatment with TGF-ß (17) . The gene array analysis reveals that it is also strongly activated in response to ascorbic acid induction (Table 4Citation , line 3). Activation is apparent by day 14 and persists through the mineralization phase. The transcription factor Nrf2 (Table 4Citation , line 9) regulates a wide-ranging metabolic response to oxidative stress, including activation of A170 (see, for example, Ref. 18 ). The activation of Nrf2 in combination with A170 implies that an oxidative stress response is a regulated aspect of osteoblast differentiation in culture, most likely as a consequence of matrix accumulation and mineralization. This may relate to the critical requirement for angiogenesis during bone growth and fracture healing. VEGF (F4j) is strongly linked with angiogenesis and is expressed in mouse osteoblasts at late stages of differentiation (19) . It can also be induced in an osteoblast-like cell line in response to hypoxia (20) . VEGF induction at the RNA level is just barely detectable on the blot at day 21 (F4j). However, induction of the VEGF receptor (A4j) is more clearly apparent in Table 1Citation , although it is below the cut-off chosen for Table 4Citation .

Induction of a member of the TRAF family of NF-{kappa}B activators in differentiating osteoblasts (Table 4Citation , line 4) is interesting in light of a report suggesting that TRAF family members play a role in the regulation of osteoclast function (21) . Expression of stromelysin-3 (Table 4Citation , line 7) has been observed in MC3T3-E1 cells (22) , but its specific induction during differentiation in response to ascorbic acid has not been reported previously.

CRABP-II (Table 4Citation , line 9) is presumed to modulate the level of retinoic acid available to bind to the receptor in cells. Expression of CRABP-II has been observed only in selective tissues and cell types; these include human osteoblasts (23) . The gene is expressed at a very low level in noninduced MC3T3-E1 cells, but it is activated 4–10 fold as the cells progress to a more osteoblast-like phenotype. The usual effector for CRABP-II expression is retinoic acid, and expression of CRABP-II is generally associated with coexpression of the retinoic acid nuclear receptors (retinoic acid receptors and RXRs). Expression of RXR-{gamma} (D6f), ß2-retinoic acid receptor (B3k), or RXR-ß cis-11 (B4c) was not detected on the blots, although other forms of the receptor may be expressed. The retinoid-related orphan receptor (ROR)-{alpha} 1 (D5i) is induced by day 7. This is consistent with recent results indicating a significant role for the ROR-{alpha} receptor in bone metabolism (24) . Another component of this signaling pathway, the RXR interacting protein (RIP-15; D6g), is induced with kinetics similar to CRABP-II, although the expression level is lower. CRABP-II expression can be stimulated by TGF-ß in mouse embryonic palatal cells (25) . Consistent with this finding, CRABP-II expression parallels an increased expression of TGF-ß (Table 4Citation , lines 18 and 19; discussed further below) in the MC3T3-E1 cells.

IGFs have been shown to enhance matrix formation and cell proliferation in bone models (reviewed in Ref. 26 ). The IGFs are modulated by the binding of IGFBPs (27) . The expression of IGFBP-5 and IGFBP-6 (Table 4Citation , lines 17 and 18) in MC3T3-E1 cells during differentiation is expected and is consistent with the activities of these proteins in modulating the effects of IGF-1 (F3a) and IGF-2 (F2n). Expression of many of these factors in MC3T3-E1 cells has been described previously (28) , although these investigators did not analyze expression through each stage of differentiation. The peak expression of IGFBP-6 and IGFBP-5 at about day 14 is consistent with a previous study (29) using MC3T3-E1 cells. In contrast to that report, the expression array did not show a marked increase in expression of IGFBP-2 (A5m). We did, however, observe an increase in expression of IGFBP-4 (F2l), although its expression was fairly high even before exposure to differentiation medium, as discussed under Table 2Citation . IRSs (IRS-1 and IRS-2) are critical for mediating the anabolic effects of insulin and IGF-1 (F3a) on bone metabolism (30) . The blots indicate expression of IRS-1 in MC3T3-E1 cells before differentiation and an increase of more than 5-fold by day 14 of differentiation (Table 4Citation , line 14). As indicated under Table 2Citation , IRS-1 was not detected in NIH3T3 cells. Consistent with its role in IGF signal transduction, we found that the increase in IRS-1 expression in MC3T3-E1 cells parallels that of the IGFBPs, peaking at day 14.

The TGF-ß superfamily consists of TGF-ßs, BMPs, and activins. They are multifunctional proteins that exert their biological effects by signaling through a family of heteromeric serine/threonine kinase receptors (reviewed in Ref. 31 ). TGF-ß1 and TGF-ß2 are associated with the extracellular matrix and are integral players in the processes of bone formation (32) . The increased expression of TGF-ß1 and TGF-ß2 (Table 4Citation , lines 18 and 19) and BMP-2 and BMP-4 (discussed below) during differentiation of the MC3T3-E1 cells is consistent with an important role for the TGF-ß superfamily in bone cell development and remodeling.

BMP receptor 1A (Table 4Citation , line 13) is a serine/threonine kinase receptor that mediates the osteogenic effects of the BMPs. It is specifically expressed in the calvarial cells relative to NIH3T3 cells and is activated as much as 10-fold during differentiation of MC3T3-E1 cells. The kinetics of its expression, peaking at day 14 and declining by day 21, when mineralization is apparent, correlate with the expression pattern of many of the other osteogenic proteins discussed here. The constitutive expression of the BMP-2 and BMP-4 ligands in MC3T3-E1 cells has been reported previously (33) . These authors found that expression of functional BMP receptor 1A is required for progression of these cells to a mineralization phenotype during long-term culture. Expression of several members of the BMP family (F1a–F1f) is detectable, as shown in Table 1Citation . The temporal expression pattern of BMP receptor 1A is a potential mechanism to coordinate the osteogenic effects of the BMPs with other differentiation mediators.

Follistatin (Table 4Citation , line 15) is an activin-A-binding protein. Activin-A, a member of the TGF-ß superfamily, is a pluripotent growth factor with important roles in development, erythropoiesis, and the local regulation of many tissues. Hashimoto et al. (34) reported that MC3T3-E1 cells contain a high number of activin-binding sites on their surface and that activin promotes mitogenesis and suppresses alkaline phosphatase activity in these cells. Activin activity is suppressed through the formation of complexes with inhibin and follistatin. The inhibin {alpha} subunit (F2g) is detectable only at very low levels on the blots. The inhibin ßA subunit (F2h) is expressed at higher levels and is induced with the same kinetics as follistatin, but just below the cut-off point chosen for Table 4Citation . The kinetics of follistatin and inhibin ßA expression observed here are consistent with the timing of alkaline phosphatase induction in MC3T3-E1 cells after exposure to ascorbic acid (5) . The blots do not show consistent expression of the activin type I receptor (E1a), although other activin-binding sites may be present on these cells.

Cathepsins are secreted lysosomal proteases that mediate the degradation of certain cellular and extracellular proteins, inside and outside of the lysosome. Cathepsin K has been demonstrated to play an important role in osteoclast function because of its ability to efficiently degrade type-1 collagen (reviewed in Ref. 35 ), but induction of cathepsins has not been studied previously during ascorbic acid-mediated differentiation in osteoblasts. This family also proteolytically activates proenzymes, prohormones, and growth factors. Activation of TGF-ß in human osteoblasts in response to glucocorticoid treatment correlates with a dose-dependent increase in the mRNA levels of cathepsin B and D (36) . Inspection of Table 1Citation indicates that cathepsin B (F6g) and cathepsin D (F6h; Table 4Citation , line 20) are also induced at certain times in ascorbic acid-treated MC3T3-E1 cells. The expression arrays support the suggestion that increased expression of the cathepsins is an important factor in the proteolytic activation of TGF-ß in differentiating osteoblasts and possibly in remodeling of the matrix.

The D grid of the expression array presents cDNAs for almost 100 tightly regulated transcription factors. The Cbfa1 transcription factor, which is not included on the array, is clearly associated with osteoblast-specific gene expression (reviewed in Refs. 37 –39). However, the transcription factor profile in differentiating cells is complex, and the expression array affords an opportunity to compare the relative expression of numerous factors of interest. In addition to Nrf2, discussed above, three other transcription factors are activated above the limits set for Table 4Citation . These are transcription factors S-II, SEF-2, and SP2 (Table 4Citation , lines 10–12). None of these has been studied in osteoblasts before. The expression pattern of all three factors mimics that of the typical differentiation factors, starting out low, increasing until day 14, and then decreasing after the onset of mineralization. Transcription factor S-II (TFIIS) is a member of a set of general transcription elongation factors that permit RNA polymerase II to transcribe faster and more efficiently (reviewed in Ref. 40 ). In vertebrates, a family of related genes has been identified, of which some members are expressed in a tissue-specific manner. The arrays indicate that expression of TFIIS is restricted to a point in osteoblast differentiation at which proliferation has ceased, and the expression of genes associated with differentiation has been initiated. Transcription factor SP2 is part of the SP1 multigene family. It has been reported to play a role in the regulated expression of the T-cell receptor and the CT{alpha} genes (41 , 42) .

Transcription factor SEF2 (also known as ITF2 or E2-2) is a member of the E factor class of the bHLH transcription factor family. Almost every member of this family has been implicated in the regulation of transcription during cell type determination and differentiation. Several laboratories have examined the role of bHLH proteins in osteoblasts. Functional E boxes have been demonstrated in osteocalcin promoter assays in MC3T3-E1 cells and other osteoblast-like cell lines (43, 44, 45) . Two recent studies report a negative role for the bHLH proteins TWIST and DERMO-1 in osteoblast differentiation. Overexpression of TWIST causes dedifferentiation in the human osteosarcoma line SaOS2, whereas underexpression leads to up-regulation of osteoblast markers including alkaline phosphatase, type I collagen, and osteopontin (46) . Tamura and Noda (47) demonstrated expression of DERMO-1 mRNA in undifferentiated MC3T3-E1 cells and found that it is expressed at a lower level in these cells by day 21 of differentiation. The results presented here represent the first report of SEF2 expression in osteoblasts. The time course of SEF2 expression suggests that this bHLH protein plays a positive role in transcriptional regulation of the osteoblast phenotype.

Inspection of the raw data in Table 1Citation reveals many additional changes that may also be significant, although they involve lower expression levels or induction at only a single assay point. The activation of the Y box-binding protein (D7j) at day 3 was discussed under Table 2Citation . IFN-regulatory factor 2 (D4l) is activated, whereas IFN-regulatory factor 1 is repressed (B7k in Table 3Citation ). IFN-inducible protein 1 (D4k) is induced. The glucocorticoid receptor form A (E3m), growth hormone receptor (E3n), and insulin receptors (E4a) are induced at low but fairly consistent levels. In contrast, the androgen (E3j), calcitonin (E3k), and estrogen (E3l) receptors show no expression at any assay point.

Many of the observed changes, such as the induction of follistatin, BMP receptor 1A, TGF-ß, and matrix remodeling factors, reflect expected patterns and support the physiological relevance of the results. Other observed changes were not anticipated and offer new insight into the osteoblast differentiation process. To evaluate the reliability of the data indicating unexpected gene inductions, we performed Northern blot analyses on selected genes with distinct expression patterns. The genes we chose include SEF2, because this is the first bHLH protein found to be up-regulated during osteoblast differentiation. We also examined the oxidative stress-associated proteins A170 and Nrf2. The results of the Northern blots are shown in Fig. 7Citation , where they are compared with the gene array results depicted in bar graphs. The gene array analysis indicated a steady rise in expression of SEF2, with a peak at about day 14. The Northern blots show an almost identical pattern. A170 was essentially undetectable in the gene array analysis until it rose sharply at about day 14. The gene arrays suggest that A170 expression falls off sharply as well. The Northern blots confirm this dramatic rise and fall in expression. The gene arrays indicate that Nrf2 is induced 4–5 fold by day 7 and remains fairly high thereafter. The Northern blot is mostly consistent with this pattern. The Northern blot suggests that induction may begin by day 4. The Nrf2 signal on the Northern blots does not stay consistently elevated, but this may represent normal variation in assays. Overall, the Northern blots give a high degree of confidence that the results from the gene array are a fair representation of actual changes in gene expression during osteoblast differentiation.



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 7. Comparison of gene expression by Northern blots and gene array analysis. Northern blot signals for selected genes as described in the text are shown on the left. Bar graphs depicting the gene array results are shown on the right for comparison. Genes were analyzed at a similar set of time points as indicated in the figure.

 
During preparation of this study, Seth et al. (48) reported an analysis of differentiation in ascorbic acid-induced MC3T3-E1 cells at days 6 and 14 using a SAGE approach. This method relies on the cloning, sequencing, and database identification of short stretches of expressed messages to generate a profile of expressed sequences. Compared with the gene array methodology, the SAGE protocol has the advantage of being unbiased at the outset, but it does not have the advantage of focusing on a specific array of regulatory genes. The two approaches are thus complementary. The SAGE analysis detected expression of numerous structural genes as well as various markers of osteoblast differentiation including a series of products associated with extracellular matrix formation. The major unanticipated and potentially important developmental gene identified in induced cells was Lefty-1. This is a TGF-ß superfamily member that was not included in our array. Lefty-1 appears able to modulate expression of alkaline phosphatase in response to ascorbic acid.

Summary.
The data generated from this type of approach are largely descriptive. However, the overview that is obtained is a very useful baseline for more analytical probes of functional pathways. Differentiation in any system involves a cascade of signals from receptor-mediated events to transcription factor-mediated changes in gene expression. The analysis presented here offers a dynamic picture of these events during osteoblast differentiation. In this context, it is important to note not just the genes that change but also those that are not expressed or induced. Where comparisons are possible, the expression array results presented here are generally consistent with previous studies. However, many genes previously studied in osteoblasts were analyzed under different conditions, in different cell lines, or under induction with different agents. The ability to compare data collected on hundreds of genes under one set of conditions with data from other systems strengthens our overall understanding of the molecular basis of osteogenesis. The analysis has also produced several novel insights, as discussed throughout the text. These data are valuable not just for a better understanding of osteogenesis but also for the comparisons they permit with other tissue types. This information should aid in the development of effective treatments for bone disorders. It may also help predict side effects on bone metabolism from drugs that target the same factors for intervention in other diseases.


   Materials and Methods
TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
References
 
Cell Culture.
Low-passage MC3T3-E1 cells were a gift from Roland Baron (Yale University, New Haven, CT). Cells were maintained in {alpha}-MEM (Irvine Scientific) plus 10% fetal bovine serum (Summit Biotechnologies), supplemented with 50 units/ml penicillin and 50 µg/ml streptomycin (Mediatech). For differentiation assays, cells were plated at an approximate initial density of 5 x 104 cells/cm2. Differentiation was induced by the addition of 50 µg/ml (final concentration) ascorbic acid (Sigma) and 10 mM (final concentration) ß-glycerol phosphate (Sigma) to standard growth medium. The medium was changed every 3–4 days, and the inducing agents were replaced with each media change. NIH3T3 cells were a gift from Scott Shore (Temple University, Philadelphia, PA). These cells were grown in DMEM (CellGro) plus 10% fetal bovine supplemented with 50 units/ml penicillin and 50 µg/ml streptomycin, as described above.

cDNA Array.
We used the Atlas mouse cDNA Expression Array (Clontech 7741-1) and the recommended protocol. The internet support site6 maintained by the manufacturer includes a complete list of genes in the array with a short description of gene function and a list of relevant publications.

RNA and Gene Array Probe Preparation.
Total cell RNA was prepared from appropriate cell cultures at various time points using Trizol reagent (Life Technologies, Inc.) according to the manufacturer’s recommendations. Samples were treated with DNase (Boehringer Mannheim) before probe preparation. Labeled probes were prepared by reverse transcription of 5 µg of RNA in the presence of [32P]dATP (New England Nuclear). Probes were hybridized to the membrane overnight, the membrane was washed thoroughly according to the manufacturer’s recommended protocol, and the results were visualized by either a PhosphorImager for quantification or by autoradiography. Signals were quantitated by exposure on a Fujix BAS 200 (Fuji) and analyzed with MacBAS 2.0 software. The blots contain nine negative controls and nine positive controls. The positive controls can be used with phosphorimaging to normalize the signal from blot to blot.

cDNA Northern Probes.
cDNA probes, type-I collagen, osteocalcin, and actin have been described previously (5) . The cDNA for histone H4 was purchased from American Type Culture Collection (clone ID, 775450). cDNAs used for PREF-1, SEF-2, A170, CD44, and Nrf2 were created by reverse transcription-PCR. Primer sequences were purchased from Clontech.

Northern Blots.
Total cell RNA was prepared from appropriate cell cultures at various time points using Trizol reagent (Life Technologies, Inc.) according to the manufacturer’s recommendations. Twenty µg of RNA were loaded per lane and fractionated by electrophoresis through a 1% formaldehyde-agarose gel. The RNA was transferred to a Hybond-N nylon membrane (Amersham Life Science Inc.) and cross-linked by UV irradiation and baking at 80°C. 32P-labeled probes were prepared using a random primed labeling kit (Boehringer Mannheim). Between successive probes, blots were stripped by treatment with boiling in 0.1% SDS.

Radiolabeling Cellular Proteins and Immunoprecipitation.
MC3T3-E1 cells were incubated with [35S]methionine/cysteine (New England Nuclear) in methionine/cysteine-free DMEM (Life Technologies, Inc.) for 2 h. Cells were washed in PBS and collected by centrifugation (2000 rpm for 5 min). Cell pellets were lysed in p300 lysis buffer (49) supplemented with the following protease inhibitors at a final concentration as indicated: aprotinin (2.0 µg/ml); leupeptin (2.0 µg/ml); and pepstatin (1.0 µg/ml). One mg of total cell lysate (Bradford assay) was immunoprecipitated with the p53 monoclonal antibody PAb421 (50) . The immunoprecipitation was separated by 10% SDS-PAGE and visualized by autoradiography.


   Acknowledgments
 
We thank Roland Baron and Scott Shore and Ed Harlow for gifts of cell lines and antibodies. We thank Peter Dallas, Xavier Graña, Scott Shore, Steve Popoff, and Gil Morris for helpful discussions. We also thank Nerissa Ngati and Valery Audige for excellent technical assistance.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported by National Cancer Institute Grant CA53592 (to E. M.), a biomedical science grant from the Arthritis Foundation (to E. M.), and the Temple University Research Enterprise Program (E. M.). Back

2 G. R. B. was supported in part by NIH Training Grant T30 CA09214, by a Daniel Swern Fellowship from Temple University, and by National Cancer Institute Scholar Award CA84573. Back

3 Present address: National Cancer Institute, Gene Regulation Section, Building 560, Room 21-21, Frederick, MD 21702-1201. Back

4 To whom requests for reprints should be addressed at Fels Institute for Cancer Research and Molecular Biology, Temple University School of Medicine, Philadelphia, PA 19140. Phone: (215) 707-7313; Fax: (215) 707-6989; E-mail: betty{at}unix.temple.edu Back

5 The abbreviations used are: STAT, signal transducers and activators of transcription; IGFBP, insulin-like growth factor-binding protein; VEGF, vascular endothelial growth factor; TGF, transforming growth factor; Nrf2, NFE2-related factor 2; RXR, retinoid X receptor; IGF, insulin-like growth factor; IRS, insulin receptor substrate; BMP, bone morphogenetic protein; bHLH, basic helix-loop-helix; SAGE, serial analysis of gene expression; PKC, protein kinase C; NF-{kappa}B, nuclear factor {kappa}B; CRABP, cellular retinoic acid-binding protein; TFII, transcription factor II; TRAF, tumor necrosis factor receptor-associated factor. Back

6 Internet address: http://www.clontech.com/archive/JAN98UPD/Atlaslist.html Back

Received for publication 9/ 6/00. Revision received 12/15/00. Accepted for publication 12/18/00.


   References
TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 

  1. Sudo H., Kodama H., Amagi Y., Yamamoto S., Kasai S. In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria.. J. Cell Biol., 96: 191-198, 1983.[Abstract/Free Full Text]
  2. Quarles D., Yohay D. A., Lever L. W., Caton R., Wenstrup R. J. Distinct proliferative and differentiated stages of murine MC3T3–E1 cells in culture: an in vitro model of osteoblast development.. J. Bone Miner. Res., 7: 683-692, 1992.[Medline]
  3. Franceschi R. T., Iyer B. S., Cui Y. Effects of ascorbic acid on collagen matrix formation and osteoblast differentiation in murine MC3T3–E1 cells.. J. Bone Miner. Res., 9: 843-854, 1994.[Medline]
  4. Towler D. A. Use of cultured osteoblastic cells to identify and characterize transcriptional regulatory elements. Bilezikian J. Raisz L. G. Rodan G. A. eds. . Principles of Bone Biology, : 1173-1188, Academic Press New York 1996.
  5. Beck G. R., Jr., Sullivan E. C., Moran E., Zerler B. Relationship between alkaline phosphatase levels, osteopontin expression and mineralization in differentiating MC3T3–E1 osteoblasts. J. Cell. Biochem., 68: 269-280, 1998.[Medline]
  6. Beck G. R., Jr., Zerler B., Moran E. Phosphate is a specific signal for induction of osteopontin gene expression.. Proc. Natl. Acad. Sci. USA, 97: 8352-8357, 2000.[Abstract/Free Full Text]
  7. Jamal H. H., Aubin J. E. CD44 expression in fetal rat bone: in vivo and in vitro.. Exp. Cell Res., 223: 467-477, 1996.[Medline]
  8. Dhalla A. K., Ririe S. S., Swamynathan S. K., Weber K. T., Guntaka R. V. chk-YB-1b, a Y-box binding protein, activates transcription from rat {alpha}1(I) procollagen gene promoter.. Biochem. J., 336: 373-379, 1998.
  9. Yamamura I., Hirata H., Hosokawa N., Nagata K. Transcriptional activation of the mouse HSP47 gene in mouse osteoblast MC3T3–E1 cells by TGF-ß.. Biochem. Biophys. Res. Commun., 244: 68-74, 1998.[Medline]
  10. Okamoto K., Prives C. A role of cyclin G in the process of apoptosis.. Oncogene, 18: 4606-4615, 1999.[Medline]
  11. Holland N. D., Zhang S. C., Clark M., Panopoulou G., Lehrach H., Holland L. Z. Sequence and developmental expression of AmphiTob, an amphioxus homolog of vertebrate Tob in the PC3/BTG1/Tob family of tumor suppressor genes.. Dev. Dyn., 210: 11-18, 1997.[Medline]
  12. Rodier A., Marchal-Victorion S., Rochard P., Casas F., Cassar-Malek I., Rouault J. P., Magaud J. P., Mason D. Y., Wrutniak C., Cabello G. BTG1: a triiodothyronine target involved in the myogenic influence of the hormone.. Exp. Cell Res., 249: 337-348, 1999.[Medline]
  13. Bieker J. J. Isolation, genomic structure, and expression of human erythroid Kruppel-like factor (EKLF).. DNA Cell Biol., 15: 347-352, 1996.[Medline]
  14. Miller I. J., Bieker J. J. A novel, erythroid cell-specific murine transcription factor that binds to the CACCC element and is related to the Kruppel family of nuclear proteins.. Mol. Cell. Biol., 13: 2776-2786, 1993.[Abstract/Free Full Text]
  15. Lehar S. M., Nacht M., Jacks T., Vater C. A., Chittenden T., Guild B. C. Identification and cloning of EI24, a gene induced by p53 in etoposide-treated cells.. Oncogene, 12: 1181-1187, 1996.[Medline]
  16. Yanagawa T., Yuki K., Yoshida H., Bannai S., Ishii T. Phosphorylation of A170 stress protein by casein kinase II-like activity in macrophages.. Biochem. Biophys. Res. Commun., 241: 157-163, 1997.[Medline]
  17. Okazaki M., Ito S., Kawakita K., Takeshita S., Kawai S., Makishima F., Oda H., Kakinuma A. Cloning, expression profile, and genomic organization of the mouse STAP/A170 gene.. Genomics, 60: 87-95, 1999.[Medline]
  18. Ishii T., Itoh K., Takahashi S., Sato H., Yanagawa T., Katoh Y., Bannai S., Yamamoto M. Transcription factor Nrf2 coordinately regulates a group of oxidative stress-inducible genes in macrophages.. J. Biol. Chem., 275: 16023-16029, 2000.[Abstract/Free Full Text]
  19. Deckers M. M., Karperien M., van der Bent C., Yamashita T., Papapoulos S. E., Lowik C. W. Expression of vascular endothelial growth factors and their receptors during osteoblast differentiation.. Endocrinology, 141: 1667-1674, 2000.[Abstract/Free Full Text]
  20. Steinbrech D. S., Mehrara B. J., Saadeh P. B., Chin G., Dudziak M. E., Gerrets R. P., Gittes G. K., Longaker M. T. Hypoxia regulates VEGF expression and cellular proliferation by osteoblasts in vitro.. Plast. Reconstr. Surg., 104: 738-747, 1999.[Medline]
  21. Wong B. R., Josien R., Lee S. Y., Vologodskaia M., Steinman R. M., Choi Y. The TRAF family of signal transducers mediates NF-{kappa}B activation by the TRANCE.. J. Biol. Chem., 273: 28355-28359, 1998.[Abstract/Free Full Text]
  22. Delany A. M., Canalis E. Dual regulation of stromelysin-3 by fibroblast growth factor-2 in murine osteoblasts.. J. Biol. Chem., 273: 16595-16600, 1998.[Abstract/Free Full Text]
  23. Kindmark A., Torma H., Johansson A. G., Ljunghall S., Melhus H. Cellular retinoic acid-binding protein type II is expressed in adult human osteoblasts and in adult liver.. Biochem. Biophys. Res. Commun., 189: 1397-1403, 1992.[Medline]
  24. Meyer T., Kneissel M., Mariani J., Fournier B. In vitro and in vivo evidence for orphan nuclear receptor ROR{alpha} function in bone metabolism.. Proc. Natl. Acad. Sci. USA, 97: 9197-9202, 2000.[Abstract/Free Full Text]
  25. Nugent P., Greene R. M. Interactions between the transforming growth factor ß (TGFß) and retinoic acid signal transduction pathways in murine embryonic palatal cells.. Differentiation (Camb.), 58: 149-155, 1994.[Medline]
  26. Linkhart T. A., Mohan S., Baylink D. J. Growth factors for bone growth and repair: IGF, TGFß and BMP.. Bone, 19(Suppl.1): 1S-12S, 1996.[Medline]
  27. Baxter R. C. Insulin-like growth factor (IGF)-binding proteins: interactions with IGFs and intrinsic bioactivities.. Am. J. Physiol. Endocrinol. Metab., 278: E967-E976, 2000.[Abstract/Free Full Text]
  28. Amarnani S., Merriman H. L., Linkhart T. A., Baylink D. J., Mohan S. Autocrine regulators of MC3T3–E1 cell proliferation.. J. Bone Miner. Res., 8: 157-165, 1993.[Medline]
  29. Thrailkill K. M., Siddhanti S. R., Fowlkes J. L., Quarles L. D. Differentiation of MC3T3–E1 osteoblasts is associated with temporal changes in the expression of IGF-land IGFBPs.. Bone, 17: 307-313, 1995.[Medline]
  30. Ogata N., Chikazu D., Kubota N., Terauchi Y., Tobe K., Azuma Y., Ohta T., Kadowaki T., Nakamura K., Kawaguchi H. Insulin receptor substrate-1 in osteoblast is indispensable for maintaining bone turnover.. J. Clin. Investig., 105: 935-943, 2000.[Medline]
  31. Piek E., Heldin C. H., Ten Dijke P. Specificity, diversity, and regulation in TGF-ß superfamily signaling.. FASEB J., 13: 2105-2124, 1999.[Abstract/Free Full Text]
  32. Bonewald L. F., Mundy G. R. Role of transforming growth factor-ß in bone remodeling.. Clin. Orthop., 250: 261-276, 1990.
  33. Suzawa M., Takeuchi Y., Fukumoto S., Kato S., Ueno N., Miyazono K., Matsumoto T., Fujita T. Extracellular matrix-associated bone morphogenetic proteins are essential for differentiation of murine osteoblastic cells in vitro.. Endocrinology, 140: 2125-2133, 1999.[Abstract/Free Full Text]
  34. Hashimoto M., Shoda A., Inoue S., Yamada R., Kondo T., Sakurai T., Ueno N., Muramatsu M. Functional regulation of osteoblastic cells by the interaction of activin-A with follistatin.. J. Biol. Chem., 267: 4999-5004, 1992.[Abstract/Free Full Text]
  35. Turk B., Turk D., Turk V. Lysosomal cysteine proteases: more than scavengers.. Biochim. Biophys. Acta, 1477: 98-111, 2000.[Medline]
  36. Oursler M. J., Riggs B. L., Spelsberg T. C. Glucocorticoid-induced activation of latent transforming growth factor-ß by normal human osteoblast-like cells.. Endocrinology, 133: 2187-2196, 1993.[Abstract]
  37. Ducy P., Karsenty G. Genetic control of cell differentiation in the skeleton.. Curr. Opin. Cell Biol., 10: 614-619, 1998.[Medline]
  38. Franceschi R. T. The developmental control of osteoblast-specific gene expression: role of specific transcription factors and the extracellular matrix environment.. Crit. Rev. Oral Biol. Med., 10: 40-57, 1999.[Abstract]
  39. Lian J. B., Stein G. S., Stein J. L., van Wijnen A. J. Osteocalcin gene promoter: unlocking the secrets for regulation of osteoblast growth and differentiation.. J. Cell. Biochem. Suppl., 30–31: 62-72, 1998.
  40. Wind M., Reines D. Transcription elongation factor SII.. Bioessays, 22: 327-336, 2000.[Medline]
  41. Bakovic M., Waite K. A., Vance D. E. Functional significance of Sp1, Sp2, and Sp3 transcription factors in regulation of the murine CTP: phosphocholine cytidylyltransferase {alpha} promoter.. J. Lipid Res., 41: 583-594, 2000.[Abstract/Free Full Text]
  42. Kingsley C., Winoto A. Cloning of GT box-binding proteins: a novel Sp1 multigene family regulating T-cell receptor gene expression.. Mol. Cell. Biol., 12: 4251-4261, 1992.[Abstract/Free Full Text]
  43. Kazhdan I., Rickard D., Leboy P. S. HLH transcription factor activity in osteogenic cells.. J. Cell. Biochem., 65: 1-10, 1997.[Medline]
  44. Quarles L. D., Siddhanti S. R., Medda S. Developmental regulation of osteocalcin expression in MC3T3–E1 osteoblasts: minimal role of the proximal E-box cis-acting promoter elements.. J. Cell. Biochem., 65: 11-24, 1997.[Medline]
  45. Tamura M., Noda M. Identification of a DNA sequence involved in osteoblast-specific gene expression via interaction with helix-loop-helix (HLH)-type transcription factors.. J. Cell Biol., 126: 773-782, 1994.[Abstract/Free Full Text]
  46. Lee M. S., Lowe G. N., Strong D. D., Wergedal J. E., Glackin C. A. TWIST, a basic helix-loop-helix transcription factor, can regulate the human osteogenic lineage.. J. Cell. Biochem., 75: 566-577, 1999.[Medline]
  47. Tamura M., Noda M. Identification of DERMO-1 as a member of helix-loop-helix type transcription factors expressed in osteoblastic cells.. J. Cell. Biochem., 72: 167-176, 1999.[Medline]
  48. Seth A., Lee B. K., Qi S., Vary C. P. H. Coordinate expression of novel genes during osteoblast differentiation.. J. Bone Miner. Res., 15: 1683-1696, 2000.[Medline]
  49. Yaciuk P., Moran E. Analysis with specific polyclonal antiserum indicates that the E1A-associated 300-kDa product is a stable nuclear phosphoprotein that undergoes cell cycle phase-specific modification.. Mol. Cell. Biol., 11: 5389-5397, 1991.[Abstract/Free Full Text]
  50. Harlow E., Crawford L. V., Pim D. C., Williamson N. M. Monoclonal antibodies specific for simian virus 40 tumor antigens.. J. Virol., 39: 861-869, 1981.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 CarcinogenesisHome page
S.-K. Kim, H.-R. Jang, J.-H. Kim, M. Kim, S.-M. Noh, K.-S. Song, G. H. Kang, H. J. Kim, S.-Y. Kim, H.-S. Yoo, et al.
CpG methylation in exon 1 of transcription factor 4 increases with age in normal gastric mucosa and is associated with gene silencing in intestinal-type gastric cancers
Carcinogenesis, August 1, 2008; 29(8): 1623 - 1631.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
N. Nabavi, Y. Urukova, M. Cardelli, J. E. Aubin, and R. E. Harrison
Lysosome Dispersion in Osteoblasts Accommodates Enhanced Collagen Production during Differentiation
J. Biol. Chem., July 11, 2008; 283(28): 19678 - 19690.
[Abstract] [Full Text] [PDF]

Home page
 haematolHome page
J. Ma, M. Shi, J. Li, B. Chen, H. Wang, B. Li, J. Hu, Y. Cao, B. Fang, and R. C. Zhao
Senescence-unrelated impediment of osteogenesis from Flk1+ bone marrow mesenchymal stem cells induced by total body irradiation and its contribution to long-term bone and hematopoietic injury
Haematologica, July 1, 2007; 92(7): 889 - 896.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
N. G. Nagl Jr., D. R. Zweitzig, B. Thimmapaya, G. R. Beck Jr., and E. Moran
The c-myc Gene Is a Direct Target of Mammalian SWI/SNF-Related Complexes during Differentiation-Associated Cell Cycle Arrest
Cancer Res., February 1, 2006; 66(3): 1289 - 1293.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
N. G. Nagl Jr., A. Patsialou, D. S. Haines, P. B. Dallas, G. R. Beck Jr., and E. Moran
The p270 (ARID1A/SMARCF1) Subunit of Mammalian SWI/SNF-Related Complexes Is Essential for Normal Cell Cycle Arrest
Cancer Res., October 15, 2005; 65(20): 9236 - 9244.
[Abstract] [Full Text] [