Physiol. Genomics Watch the video to learn how APS reaches out to developing nations.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 30: 61-68, 2007. First published February 27, 2007; doi:10.1152/physiolgenomics.00277.2006 Free Article
1094-8341/07 $8.00
This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
30/1/61    most recent
00277.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
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 Google Scholar
Google Scholar
Right arrow Articles by Shirvani, S. M.
Right arrow Articles by Chin, M. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shirvani, S. M.
Right arrow Articles by Chin, M. T.
Received 14 December 2006; accepted in final form 23 February 2007.
Physiological Genomics 30:61-68 (2007)
1094-8341/07 $8.00 © 2007 American Physiological Society

Transcription factor CHF1/Hey2 regulates the global transcriptional response to platelet-derived growth factor in vascular smooth muscle cells

Shervin M. Shirvani 1, Linette Mookanamparambil 2, Marco F. Ramoni 2,3 and Michael T. Chin 1

1 Vascular Medicine Research, Cardiovascular Division, Brigham and Women's Hospital and Harvard Medical School
2 Harvard Partners Center for Genetics and Genomics, Harvard Medical School
3 Children's Hospital Informatics Program, Division of Health Sciences and Technology, Harvard Medical School and Massachusetts Institute of Technology, Boston, Massachusetts


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
The cardiovascular restricted transcription factor CHF1/Hey2 has been previously shown to regulate the smooth muscle response to growth factors. To determine how CHF1/Hey2 affects the smooth muscle response to growth factors, we performed a genomic screen for transcripts that are differentially expressed in wild-type and knockout smooth muscle cells after stimulation with platelet-derived growth factor. We screened 45,101 probes representing >39,000 transcripts derived from at least 34,000 genes, at eight different time points. We analyzed the expression data utilizing an algorithm based on Bayesian statistics to derive the best polynomial clustering model to fit the expression data. We found that in a total of 9,827 transcripts the normalized ratio of knockout to wild-type expression diverged more than threefold from baseline in at least one time point, and these transcripts separated into 17 distinct clusters. Further analysis of each cluster revealed distinct alterations in gene expression patterns for immediate early genes, transcription factors, matrix metalloproteinases, signaling molecules, and other molecules important in vascular biology. Our findings demonstrate that CHF1/Hey2 profoundly affects vascular smooth muscle phenotype by altering both the absolute expression level of a variety of genes and the kinetics of growth factor-induced gene expression.

microarray; genomics; vascular biology; gene knockout


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
ARTERIOSCLEROSIS IS THE LEADING cause of death in developed countries. The pathogenesis of arteriosclerosis is complex, involving the interaction of multiple cell types, genetic factors, soluble and matrix components, and derangements in blood flow (for review, see Ref. 17). The vascular smooth muscle cell (VSMC) is a major component of the arterial wall and is thought to play a critical role in the development of arteriosclerosis (7, 20, 27). In normal vessels, VSMCs are quiescent, differentiated, and contractile and function to maintain vascular tone and blood pressure. In response to injury, whether mechanical, chemical, or immunologic, they develop a dedifferentiated, proliferative, and secretory phenotype as they migrate from the media to the neointima (7, 20, 27). The upstream regulators that affect expression of these various genes in this context are largely unknown.

We have previously cloned the basic helix-loop-helix (bHLH) protein CHF1/Hey2 (2). Studies in the adult vasculature have suggested that CHF1/Hey2 gene expression varies after vascular injury and may be important for smooth muscle cell (SMC) growth and protection against apoptosis (34, 35). Mice with deletion of CHF1/Hey2 have thin-walled arteries (29) and manifest a functional vascular phenotype in which neointimal formation after vascular injury is decreased and smooth muscle cells (SMCs) from these mice are defective in their responses to the growth factors platelet-derived growth factor (PDGF) and heparin-binding epidermal growth factor-like growth factor (30). We and others have also shown that CHF1/Hey2 can inhibit expression of smooth muscle-specific genes (5, 23, 31). A comprehensive analysis of the CHF1/Hey2-dependent genome in the adult vasculature as a method to examine its potential role in the pathogenesis of occlusive vascular diseases has not yet been performed.

To determine the potential transcriptional mechanisms by which CHF1/Hey2 regulates the growth factor response in SMCs, we performed a genome-wide dynamic microarray screen to characterize and compare changes in the transcriptomes of wild-type and CHF1–/– VSMCs following treatment with PDGF, as a function of time. Following cluster analysis using Bayesian dynamics, we have identified numerous potential effector genes whose dynamic gene transcription patterns are profoundly altered after PDGF treatment.

Previous transcriptional profiling studies of VSMCs have been performed to identify transcripts that distinguish arterial from venous SMCs (15), that distinguish plaque VSMCs from medial SMCs (19), that establish SMC identity in an in vitro model of smooth muscle differentiation (32), and that are differentially expressed in rats susceptible to hypertension (6). In each case, static comparisons were made. Our study is the first to systematically perform a comparative time-series analysis of gene expression in CHF1/Hey2 knockout cells to identify transcriptional alterations that affect vascular smooth response to growth factors. Here we report that knockout of a single gene, CHF1/Hey2, alters the growth factor-induced phenotype of SMCs by dramatically altering the dynamic expression of a large number of genes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Culture of mouse aortic SMCs, PDGF treatment, and RNA harvest.
Preparation of aortic SMCs for culture has been described (30). The purity of the cells was verified by immunostaining for both smooth muscle {alpha}-actin and calponin. Fibroblast contamination was not detectable. Wild-type and knockout cells were prepared at the same time from littermates at 8–10 wk of age. Two independent preparations of each genotype were tested for their differential proliferative and migratory response to PDGF, as previously described (30). Each knockout cell preparation demonstrated defects in the proliferative and migratory response to PDGF, as expected. For these experiments, the mouse aortic smooth muscle cells (MASMCs) were between passages 10 and 20, but RNAs were prepared from cells at the same passage. For PDGF induction, cells were grown to 95% confluence on 150-mm plates, washed four times with phosphate-buffered saline, and serum starved with DMEM supplemented with 0.4% FCS. After 48 h in a 37°/5% CO2 humidified incubator, the cells were washed four times with phosphate-buffered saline and then treated with either DMEM or DMEM supplemented with 5 ng/ml human PDGF-BB. Cells were then incubated in the humidified incubator for the length of time indicated by the experiment. After the appropriate length of treatment, the cells were washed one time with phosphate-buffered saline, and RNA was harvested with the Qiagen RNeasy protocol according to the manufacturer's instructions (Qiagen).

For the microarray experiment, RNA was harvested from knockout and wild-type MASMCs at 0 (no PDGF-BB treatment), 15, 30, 60, 120, 240, and 480 min following stimulation with PDGF-BB. This mRNA was reverse transcribed into cDNA that was then used as a template for in vitro transcription and biotin labeling of a cRNA target population. The resulting mixtures from each time point were hybridized to Affymetrix Mouse Expression Set 430 microarrays consisting of 45,101 probes corresponding to >39,000 transcripts, followed by staining with a streptavidin-phycoerythrin conjugate. The chips were then washed in a fluidics station and scanned with a GeneArray Scanner (Hewlett-Packard) to collect fluorescence intensity data. All microarray data were deposited in the Gene Expression Omnibus database under accession number GSE6526.

Microarray data analysis.
Knockout-to-wild-type ratios of gene expression were calculated for each probe at each time point and then normalized to the expression ratio at the zero time point to generate expression profiles. Cluster analysis of the expression profiles was performed with Bayesian analysis using CAGED software (26). The Bayesian approach assumes that the observed data were generated by a set of unknown processes without assumptions regarding the number of processes. All clustering models were initially assumed to be equally probable (i.e., a uniform prior distribution over these models was assumed). The software then calculated the posterior probability of each model given the observed data and selected the most probable. Genes grouped this way into a single cluster were considered to be generated by the same process. An advantage of this approach is that it automatically determines the number of clusters partitioning the data as the most probable way to partition the observed gene profiles. A more detailed description of the application of this method is available online at http://www.hpcgg.org/biofx/Chin/median/3fold/poly63foldmedian.html.

Following cluster analysis, members of each cluster were sorted by gene ontology classification, biological process, biochemical activitiy, and subcellular localization using an online resource for cardiogenomics (http://db.chip.org/unchip/unchip-top). Members of each cluster that were annotated were examined manually for their relevance to cell growth, vascular biology, transcription, and signal transduction. To validate genes from the microarray analysis, RNA from independently cultured and treated cells was reverse transcribed to cDNA and amplified with candidate gene-specific primers using the Quantitect one-step RT-PCR system (Qiagen) according to the manufacturer's instructions.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Identification of genes dysregulated in the absence of CHF1/Hey2.
The differentially expressed genes included 9,827 genes grouped into 17 clusters. The natural logarithm of the normalized ratio of knockout to wild-type gene expression as a function of time for each cluster is shown in Fig. 1. A complete listing of all differentially expressed genes and the statistical goodness of fit for each cluster is available online (http://www.hpcgg.org/biofx/Chin/median/3fold/poly63foldmedian.html). Gene Ontology annotations for each member of each cluster were downloaded from a publicly available online data resource (http://db.chip.org/unchip/unchip-top). Members of each cluster were sorted by Gene Ontology classification and then reviewed manually. Those that had particular relevance to vascular biology, cell signaling, or transcription were examined in more detail.


Figure 1
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 1. Clusters of dynamic gene expression profiles of vascular smooth muscle cell after platelet-derived growth factor (PDGF) stimulation. The data are plotted as the natural logarithm of the normalized ratio of knockout (KO) cell gene expression to wild-type (WT) cell gene expression for each probe vs. time in minutes. Expression profiles for 17 clusters representing 9,827 genes are shown.

 
Each cluster demonstrated distinct characteristics. Cluster 1, for example, was notable for an dynamic changes in the knockout to wild-type expression ratio over time, where the ratio of knockout to wild-type gene expression initially increased at early time points but ultimately decreased to below baseline at the end of the time course. Many members of this cluster fall into the category of "immediate early genes" that were originally described as genes that are rapidly induced after serum stimulation and do not require protein synthesis for induction (14). Examples include c-fos, c-jun, egr2, ier2, KLF10, ier3, ier5, and egr3. Other members relevant to vascular biology include thrombospondin-1, factor III (tissue factor), osteomodulin, and Itag5 (integrin {alpha}V). Transcription factors include KLF2, KLF9, NFIL3E4BP4-like, Bhlhb2, INAP, Copeb, Tieg, egr2/Krox20, c-jun, Nr4a1, c-fos, ATF3, and junB. A representative transcriptional ratio over time is shown for c-jun, which shows a much greater induction in the knockout cells initially, but a delayed increase in the wild-type cells at the end of the time course (Fig. 2A). These findings indicate that both the magnitude and the timing of the cellular response to PDGF is altered. While the majority of the cluster shows a similar profile, the degree of induction and relative expression at each time point varies considerably, which probably reflects the sensitivity of the clustering method. Another representative example is egr2, which also shows earlier induction in the knockout cells and delayed induction in the wild-type cells (Fig. 2B). Alteration in the response of immediate early genes suggests that absence of CHF/Hey2 alters the ability of cells to respond to a variety of growth factor stimuli.


Figure 2
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 2. Expression profiles for representative genes from clusters 1 and 2. Data are presented showing the raw expression profile for KO and WT cells over time at top and the natural logarithm of the normalized individual expression ratios over time at bottom for cluster 1 members c-jun (A) and egr2 (B). A similar analysis is shown for the cluster 2 members vasodilator-stimulated phosphoprotein (VASP, C) and serum response factor (SRF, D). {triangleup}, KO cell gene expression; {circ}, WT cell gene expression; solid line in the bottom panels, normalized expression ratio.

 
Cluster 2 is also notable for alteration in the expression dynamics of gene expression. Knockout gene expression again shows an early relative increase, a later relative decrease, and then converges with wild-type gene expression at later time points. The members of this cluster include a variety of structural genes, such as smoothelin, tropomodulin, tubulin 4, syndecan 4, and vasodilator-stimulated phosphoprotein (VASP) (1). VASP is of particular interest because it has previously been shown to play a role in vascular biology, as an integrator of the signaling response. Its expression pattern is shown (Fig. 2C). Other potentially important members of this cluster include the matrix metalloproteinase inhibitor tissue inhibitor of metalloproteinase (TIMP) 3, the integrin signaling molecule Bcar1, and the transcriptional regulators serum response factor (SRF), Foxc2, and CARP. SRF has previously been shown to play important roles in both the growth factor response and muscle specific gene expression (36). Analysis of the SRF expression pattern is shown in Fig. 2D and demonstrates that the overall induction of SRF is blunted in the knockout compared with wild-type cells.

Cluster 3 is most notable for differential expression of the transcriptional regulators ATF5ß, MafB, HSAz, Zik1, zfp161, nucleolin, zfp97, and zfp235. Cluster 4 also is notable for differential regulation of primarily transcriptional regulators, including Hoxb8, Hes5, Tead3, Ipas, Bcl11a, and Rel. Cluster 5 is most striking for including the architectural transcription factors hmga1 and hmga2 and the extracellular matrix regulators TIMP, matrix metalloproteinase (mmp) 10, and mmp13. These architectural transcription factors have been implicated in cell growth and the development of benign mesenchymal tumors and are induced in vascular smooth muscle cells after wire injury in vivo and serum stimulation in vitro (3). Interestingly, they have also been described as "delayed early genes" induced several hours after growth factor stimulation, requiring protein synthesis (13). Analysis of their transcriptional expression kinetics demonstrates that their induction is delayed in the knockout cells (Fig. 3, A and B), raising the possibility that the delayed early response to growth factors is altered in these cells. Induction of mmp10 and mmp13 is also delayed in the knockout cells (Fig. 4, A and B), which would suggest a lack of invasiveness and may partially explain the lack of migration through a collagen-coated matrix and diminished neointimal formation observed in vitro and in vivo (30).


Figure 3
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 3. Expression profiles for "delayed early" genes from cluster 5. Data are presented as described in the legend to Fig. 2. The delayed early genes Hmga1 (A) and Hmga2 (B) show normal induction 1 h after PDGF stimulation in WT cells. In knockout cells, there is delayed induction peaking 4 h after stimulation. These findings suggest a general alteration in the delayed early gene response to PDGF.

 

Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 4. Expression profiles for representative genes from clusters 5 and 6. Data are presented as described in the legend to Fig. 2. The extracellular matrix regulators mmp10 (A) and mmp13 (B) show significant induction in WT cells 60 min after PDGF stimulation. The response of each of these genes is severely blunted in the KO cells, suggesting a potential mechanism by which KO cells show decreased migration through a collagen substrate in vitro. The biosynthetic enzyme Has1 (C) and the calcitonin signaling molecule Ramp2 (D) show predominantly increased expression in the KO cells. Has1 is a biosynthetic enzyme for the extracellular matrix component hyaluronan, which has been shown to inhibit smooth muscle proliferation.

 
Cluster 6 contains a significant contingent of transcription factors, such as Zfp288, Sp1, Foxg1, Tip5, mtf2, pbx2, mafg, copeb, sox4, ssbp2, CBP, zfp30, zfp161, Gli5. It also contains a number of signaling molecules, including Ramp2, Dkk1, sntb2, Gna13, il6ra, Has1, Pip5k1b, Ierepo4, Gpr83, fsc1, prk A anchor protein 4, PNQLARE, EMK, v1rb4, and ArhGAP5. The dynamic expression profile for cluster 6 is also interesting, because it is primarily positive, reflecting a general increase in expression in the knockout cells compared with wild-type. Since CHF1/Hey2 is known to be a transcriptional repressor, members of this cluster may be direct targets of CHF1/Hey2. The dynamic expression profiles of Has1, a biosynthetic enzyme for the extracellular matrix component hyaluronan, and Ramp2, a modulator of calcitonin signaling, are shown (Fig. 4, C and D).

Cluster 7 includes the transcription factors Runx1, Tfdp1, Elk3, Tcfl4, jerky, gcdp, and zfp207, the cell cycle or cell growth-related genes cyclin H, Vegf-c, Rgs2, c-mer, Fyn, cdc25a, and the extracellular matrix regulators Timp3 and mmp3. Cluster 8 also includes numerous transcription factors such as Id2, CRTR-1, SMAD6, Id3, Id4, hmgb2, hmg2, tle2, rbl2, hmgb3, Hoxa7, Hoxb6, Klf9, prrx1, Dbp, Tead2, Klf7, Irx3, Klf4, zfp95, Crem, GATA6, pbx1b, Mef2c, per1, per2, nr4a2, TCF4, RPL13a, and MITF-2A. Timp3 and mmp3 are directly relevant to vascular biology for their roles in protection against vascular inflammation (9) and plaque stabilization (12). Their dynamic expression patterns are shown (Fig. 5, A and B). GATA6 is also relevant, as it has been postulated to maintain the differentiated, contractile phenotype of vascular smooth muscle and prevent neointimal formation after injury (18, 22). We have previously shown that CHF1/Hey2 can directly interact with GATA6 and inhibit smooth muscle myosin heavy chain gene expression (31). Analysis of the GATA6 dynamic expression pattern shows that there is a delay in GATA6 downregulation in the knockout cells (Fig. 5C), which would be consistent with our previously reported phenotype in which mice lacking CHF1/Hey2 demonstrate decreased neointima formation after injury (30).


Figure 5
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 5. Expression profiles for extracellular matrix regulators from cluster 7 and the transcription factor GATA6 from cluster 8. Data are presented as described in the legend to Fig. 4. Both Timp3 (A) and mmp3 (B) demonstrate predominantly increased expression in the KO cells, which is consistent with their role as vasculoprotective molecules. GATA6 (C) demonstrated delayed downregulation, consistent with its role as a promoter of the differentiated, quiescent phenotype.

 
Cluster 9 is by far the largest cluster, containing 5,357 members. The individual members in each category are too numerous to list, but their distribution by functional class is shown (Fig. 6A). One large subgroup consists of transcription factors and molecules involved in replication and translation. Some interesting transcription factor genes are the forkhead proteins Foxa2, Foxo1, and Foxp1, which are known to play some roles in regulating cellular responsiveness to growth factor signaling (24, 25). Other molecules relevant to vascular biology include thrombomodulin (16) and vcam1 (4), and their dynamic expression profiles are shown (Fig. 6, B and C). It is of interest to note that although these genes are in the same cluster by virtue of similar alterations in expression ratios over time, their individual dynamic expression values diverge significantly.


Figure 6
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 6. Functional diversity among annotated genes in cluster 9 (A). Cluster 9 is the largest cluster, containing 5,357 members. A significant number are involved in transcription and translation, which may explain the large number of differentially expressed genes overall. Expression profiles for vascular molecules in cluster 9 (B, C). Data are presented as described in the legend to Fig. 2. Thrombomodulin (B) and vcam1 (C) show a general increase in expression in the KO cells. Thrombomodulin is known to promote anticoagulation, while vcam1 is known to promote inflammatory cell adhesion.

 
Cluster 10 is notable for the inclusion of multiple molecules that play roles in signal transduction and receptor signaling, such as Hax1, Bcap37, Ssr4, Dnaj3, Eif4ebp1, Itgb4bp, and Dok1. Cluster 11 is notable for numerous transcription factors such as Taf10, Rbbp7, Ncor2, Gli5, Etv6, Srebf1, Nfe2l2, Sal2, Nfya, Nfkb2, Hdac1, Hdac3, Sap18, zfp263, Nr2f6, Sra1, hdac7a, Alrp, Relb, and Tcerg1. This cluster also includes a number of proapoptotic molecules such as Bnip3, Bnip2, Pdcd6ip, Rnp7, and Sh3glb1 and several tyrosine kinases, including Yes, Ilk, Map2k1, Chuk, and Prpk. Cluster 12 is notable for the cell cycle molecules cyclin D1, E2f6, mcmd7, Rad1, Rgs2, Pin1, calm1, and Ppm1g, the transcription factors Gtf2h2, Gabpa, Hmx3, Mirf3, Preb, Klf5, Rpo2-3, Rpo1-3, Sdccag33, Miz1, Pparg, Dr1, Gtf2h1, and the signaling molecules Mif, Pdgfa, Il1r, Map3k7 (TAK1), Map2k2, Jak2, and Hras1. Clusters 13 and14 do not appear to be overrepresented by any particular gene ontology classification. Cluster 15 is notable for the differential expression of the transcription factors Fkhr, Prrx2, Atf2, Ahr, Hif1a, Fosl2, Bcl3, Rnpc2, Nrbf2, Taf7, Atf7ip, Pias1, and the tyrosine kinases Limk2, stk25, Clk4, Pak1, Pak3, Bmpr1a, Prkx, Cdk5, Tgfbr1, Mapkapk2. Cluster 16 contains the transcription factors Snai2, Hoxb5, Foxp1, Klf15, and Nr2f2. Cluster 17 is notable for differential expression of the cell adhesion molecules Vcam1, col15a1, Fn14, and Cspg2. The other members of these clusters are functionally diverse.

Confirmation of candidate genes.
Several candidate genes were examined with qRT-PCR to confirm microarray results: Itga5, Has1, MBD1, PGHS-B, SRF, and VASP. RNA was harvested from independently replicated samples at the initial time point and 2 h after treatment with PDGF-BB. Levels of mRNA for the candidate genes and an internal control, {alpha}-tubulin, were measured by a qRT-PCR assay with expression normalized to tubulin mRNA. For all of these genes, the trend observed in the microarray data was reiterated in the qRT-PCR experiment (data not shown).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
In this experiment, we sought to identify genes in VSMCs whose expression in response to growth factor stimulation is altered in the absence of CHF1/Hey2. To accomplish this objective, we compared the transcriptomes of CHF1/Hey2 knockout and wild-type VSMCs at eight time points following PDGF stimulation and mined the resultant data for genes whose expression varied significantly from the baseline ratios between the two groups of cells. Data mining from our experiment yielded a wide variety of candidate genes whose expression was significantly different between CHF1/Hey2 knockout and wild-type cells. The number and diversity of genes that are differentially expressed were initially surprising. These findings, however, reflect both the sensitivity and power of the CAGED algorithm (26) for determining differential expression in time series analysis through Bayesian statistical techniques. They also demonstrate that the absence of CHF1/Hey2 leads to profound alterations in growth factor-induced gene expression and consequently is an important regulator of the cellular response to growth factors. The tremendous divergence in gene expression patterns is likely due in part to the early temporal disruption of gene expression coupled with the large numbers of transcription factors that are affected. Disruption of the immediate early gene response to growth factors would be expected to have profound effects on genetic cascades that are dependent upon these early factors. There are many transcription factors that show altered expression, and since each transcription factor is expected to regulate multiple target genes, the extensive changes in gene expression observed over the course of the experiment are not unexpected. Of note also is that many genes did not show significant differences at the beginning or end of the time course but showed alterations in the kinetics and dynamics of gene expression profiles over time. These findings underscore the importance of time-series analysis of gene expression to identify and dissect important transcriptional cascades in regulating cellular responses to various stimuli. These data also provide insight into the transcriptional changes that underlie the smooth muscle response to a potent cellular mitogen and provide a unique and valuable dataset to explore the connection between extracellular binding of PDGF and intracellular gene regulation by CHF1/Hey2.

Many of the identified genes have roles in vascular smooth muscle function, and their expression profiles are consistent with our previously observed effects of CHF1/Hey2 loss of function on vascular smooth muscle phenotype (30). For instance, the hyaluronan synthase 1 (Has1) gene was especially sensitive to the presence of CHF1/Hey2. There was significant upregulation of Has1 in response to PDGF in the knockout cells that was absent in the wild-type cells. This suggests that CHF1/Hey2 functions to keep Has1 expression suppressed when the cell undergoes growth factor stimulation. It has been reported that PDGF induces the production of hyaluronan by VSMCs, which in turn exerts a potent suppressor effect on VSMC proliferation (21). Our data suggest that one mechanism by which absence of CHF1/Hey2 affects smooth muscle proliferation in vivo is by promoting the synthesis of excess free hyaluronan compared with wild-type cells, which in turn leads to decreased relative proliferation. Further study would be required to confirm the hypothesis, but nonetheless it stands as an exciting potential therapeutic pathway for modulating arteriosclerosis. Many other molecules important in the vessel wall (e.g., VASP, mmp10, mmp13, TIMP, thrombomodulin, etc.) have also been identified in our dataset, as described above.

The dataset derived from this experiment facilitates the generation of many testable hypotheses that will have important ramifications for understanding vascular smooth muscle function and CHF1/Hey2-dependent gene regulation. Thus far CHF family members have only been shown to influence gene transcription indirectly by binding and regulating other transcription factors in the nucleus such as GATA1 (8), GATA4 (10, 37), GATA6 (10, 31), ARNT (2), MyoD (33), and Hes1 (11). Though CHF family members have DNA binding domains and bind E boxes (11), their direct target genes in vivo remain to be elucidated. Our analysis provides a list of candidate genes whose expression is significantly altered by CHF1 within 2 h of growth factor stimulation and whose promoter elements could be investigated further for direct regulation by CHF1/Hey2. Our dataset provides a significant resource for understanding CHF functioning and vascular smooth muscle phenotype regulation.

One limitation of our study is that some genes appear in multiple clusters. Zfp161, for example, appears in clusters 3 and 6. The probes used on the chip represent different segments of the mRNA, and it is possible that the assignments of the "same" transcript to multiple clusters may reflect the presence of alternatively spliced transcripts. Another limitation to our method is that the many divergent transcripts detected are likely to reflect indirect effects of CHF1/Hey2 loss of function, rather than direct effects. Other methods, such as CHF1/Hey2 knockdown by siRNA or CHF1/Hey2 overexpression by viral gene transfer, may provide further insight into the direct targets of CHF1/Hey2. An additional caveat to our experimental design is that the VSMCs used are of intermediate passage number and may have undergone some degree of dedifferentiation and divergence, which could make our observations less relevant to smooth muscle biology in vivo. One alternative approach would be to examine early passage or freshly isolated VSMCs; however, we have been limited in the availability of knockout mice due to the profound neonatal lethality due to congenital heart defects in this model, as previously described (28).

As expected, numerous genes implicated in VSMC growth and proliferation including c-fos and jun were markedly dysregulated in the knockout cells, which again is consistent with the in vivo behavior of cells lacking CHF1/Hey2. In addition, numerous transcription factors and signaling molecules also showed significant dysregulation. The effect of transcription factor dysregulation on the divergence of the transcriptional response is discussed above. The dysregulation of numerous signaling molecules such as tyrosine kinases, cell surface receptors and serine-threonine kinases would also be expected to contribute to the functional and transcriptional divergence between the wild-type and knockout cells. Although many studies have examined transcriptional responses to cell signaling, our study emphasizes the potential role of a single transcription factor, CHF1/Hey2, to regulate profoundly the transcriptional and signaling response of a single cell type to a single defined stimulus. Consequently, based on our current and previous study (30), we propose that CHF1/Hey2 regulates the transcriptional and signaling "repertoire" of the VSMC.

Another limitation of our study is that it does not assess differences in baseline gene expression between the two cell populations. The method instead takes the initial baseline expression ratio between each knockout and wild-type transcript and then normalizes each subsequent ratio at ensuing time points to the initial ratio to determine how this expression ratio varies during the time course. Accordingly, there may be large differences in basal gene expression, but if the ratio remains fixed during the time course, then the transcript will not be detected by the Bayesian statistical algorithm used. To address the question of baseline differences in gene expression, we previously performed a static comparison of baseline gene expression in untreated cells, using three independent RNA samples for each cell type, prior to the initiation of the current study. Although we were able to identify numerous transcripts that are differentially expressed, none were directly related to PDGF signaling, other than SOS1 (M. T. Chin, unpublished results). The SOS1 expression ratio, interestingly, does not vary significantly over the time course of the study, and it does not appear in any of the clusters. The current study, in conjunction with our unpublished results, clearly demonstrates that dynamic differences in gene expression may be at least as important as baseline changes in defining the transcriptional response to growth factors.

An essential question raised by our study is one of how the various changes in dynamic gene expression observed aid in our understanding of the vascular phenotype of CHF1/Hey2-deficient mice. We have previously demonstrated that these mice show defects in PDGF-induced proliferation and migration (30). Our current data demonstrate gene expression alterations in a number of transcription factors associated with immediate early responses to growth factors (c-jun, egr2, SRF), delayed early responses to growth factors (HMGA1, HMGA2), and various molecules associated with the development of occlusive vascular lesions (VASP, mmp10, mmp13, TIMP3, mmp3, GATA6, thrombomodulin, and VCAM1). These findings suggest that the knockout phenotype ensues from perturbation of a complex transcriptional network controlling essential aspects of the smooth muscle cellular response to growth factors and vascular remodeling.

Our experiment was conceived to identify the transcriptional pathways altered in SMCs lacking CHF1/Hey2 that affect smooth muscle behavior during growth factor stimulation. Through dynamic microarray profiling, data mining, and qRT-PCR validation, we have successfully identified numerous genes that may be important in these pathways. The top candidates, along with their cluster number and Gene Ontology classification, are listed in Table 1. Further experimentation to explore the many hypotheses generated will be facilitated by this dataset. The number of genes differentially expressed in CHF1/Hey2 knockout cells and their broad functional and structural diversity demonstrate a significant role for CHF1/Hey2 in VSMC function. Additional characterization of these transcriptional pathways will prove useful for understanding vascular smooth muscle function in health and disease.


View this table:
[in this window]
[in a new window]

  		Table 1. Top candidate genes dysregulated in the absence of CHF1/Hey2

 

   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by National Institutes of Health Grants HL-076232 to M. T. Chin and HG-003354 to M. F. Ramoni. S. M. Shirvani was supported by a fellowship from the Sarnoff Foundation.


   FOOTNOTES
 
Address for reprint requests and other correspondence: M. T. Chin, Center for Cardiovascular Biology, Univ. of Washington School of Medicine, Box 358050, 815 Mercer St., Rm. 353, Seattle, WA 98109 (e-mail: mtchin{at}u.washington.edu).

Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Chen L, Daum G, Chitaley K, Coats SA, Bowen-Pope DF, Eigenthaler M, Thumati NR, Walter U, Clowes AW. Vasodilator-stimulated phosphoprotein regulates proliferation and growth inhibition by nitric oxide in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24: 1403–1408, 2004.[Abstract/Free Full Text]
  2. Chin MT, Maemura K, Fukumoto S, Jain MK, Layne MD, Watanabe M, Hsieh CM, Lee ME. Cardiovascular basic helix loop helix factor 1, a novel transcriptional repressor expressed preferentially in the developing and adult cardiovascular system. J Biol Chem 275: 6381–6387, 2000.[Abstract/Free Full Text]
  3. Chin MT, Pellacani A, Hsieh CM, Lin SS, Jain MK, Patel A, Huggins GS, Reeves R, Perrella MA, Lee ME. Induction of high mobility group I architectural transcription factors in proliferating vascular smooth muscle in vivo and in vitro. J Mol Cell Cardiol 31: 2199–2205, 1999.[CrossRef][Web of Science][Medline]
  4. Cybulsky MI, Iiyama K, Li H, Zhu S, Chen M, Iiyama M, Davis V, Gutierrez-Ramos JC, Connelly PW, Milstone DS. A major role for VCAM-1, but not ICAM-1, in early atherosclerosis. J Clin Invest 107: 1255–1262, 2001.[Web of Science][Medline]
  5. Doi H, Iso T, Yamazaki M, Akiyama H, Kanai H, Sato H, Kawai-Kowase K, Tanaka T, Maeno T, Okamoto E, Arai M, Kedes L, Kurabayashi M. HERP1 inhibits myocardin-induced vascular smooth muscle cell differentiation by interfering with SRF binding to CArG box. Arterioscler Thromb Vasc Biol 25: 2328–2334, 2005.[Abstract/Free Full Text]
  6. Dupuis M, Soubrier F, Brocheriou I, Raoux S, Haloui M, Louedec L, Michel JB, Nadaud S. Profiling of aortic smooth muscle cell gene expression in response to chronic inhibition of nitric oxide synthase in rats. Circulation 110: 867–873, 2004.[Abstract/Free Full Text]
  7. Dzau VJ, Braun-Dullaeus RC, Sedding DG. Vascular proliferation and atherosclerosis: new perspectives and therapeutic strategies. Nat Med 8: 1249–1256, 2002.[CrossRef][Web of Science][Medline]
  8. Elagib KE, Xiao M, Hussaini IM, Delehanty LL, Palmer LA, Racke FK, Birrer MJ, Shanmugasundaram G, McDevitt MA, Goldfarb AN. Jun blockade of erythropoiesis: role for repression of GATA-1 by HERP2. Mol Cell Biol 24: 7779–7794, 2004.[Abstract/Free Full Text]
  9. Federici M, Hribal ML, Menghini R, Kanno H, Marchetti V, Porzio O, Sunnarborg SW, Rizza S, Serino M, Cunsolo V, Lauro D, Mauriello A, Smookler DS, Sbraccia P, Sesti G, Lee DC, Khokha R, Accili D, Lauro R. Timp3 deficiency in insulin receptor-haploinsufficient mice promotes diabetes and vascular inflammation via increased TNF-alpha. J Clin Invest 115: 3494–3505, 2005.[CrossRef][Web of Science][Medline]
  10. Fischer A, Klattig J, Kneitz B, Diez H, Maier M, Holtmann B, Englert C, Gessler M. Hey basic helix-loop-helix transcription factors are repressors of GATA4 and GATA6 and restrict expression of the GATA target gene ANF in fetal hearts. Mol Cell Biol 25: 8960–8970, 2005.[Abstract/Free Full Text]
  11. Iso T, Sartorelli V, Poizat C, Iezzi S, Wu HY, Chung G, Kedes L, Hamamori Y. HERP, a novel heterodimer partner of HES/E(spl) in Notch signaling. Mol Cell Biol 21: 6080–6089, 2001.[Abstract/Free Full Text]
  12. Johnson JL, George SJ, Newby AC, Jackson CL. Divergent effects of matrix metalloproteinases 3, 7, 9, and 12 on atherosclerotic plaque stability in mouse brachiocephalic arteries. Proc Natl Acad Sci USA 102: 15575–15580, 2005.[Abstract/Free Full Text]
  13. Lanahan A, Williams JB, Sanders LK, Nathans D. Growth factor-induced delayed early response genes. Mol Cell Biol 12: 3919–3929, 1992.[Abstract/Free Full Text]
  14. Lau LF, Nathans D. Expression of a set of growth-related immediate early genes in BALB/c 3T3 cells: coordinate regulation with c-fos or c-myc. Proc Natl Acad Sci USA 84: 1182–1186, 1987.[Abstract/Free Full Text]
  15. Li J, Adams LD, Wang X, Pabon L, Schwartz SM, Sane DC, Geary RL. Regulator of G protein signaling 5 marks peripheral arterial smooth muscle cells and is downregulated in atherosclerotic plaque. J Vasc Surg 40: 519–528, 2004.[CrossRef][Web of Science][Medline]
  16. Li YH, Liu SL, Shi GY, Tseng GH, Liu PY, Wu HL. Thrombomodulin plays an important role in arterial remodeling and neointima formation in mouse carotid ligation model. Thromb Haemost 95: 128–133, 2006.[Web of Science][Medline]
  17. Libby P. The vascular biology of atherosclerosis. In: Heart Disease: A Textbook of Cardiovascular Medicine, edited by Braunwald E, Zipes DP, Libby P. Philadelphia, PA: W. B. Saunders, 2000, p. 995–1009.
  18. Mano T, Luo Z, Malendowicz SL, Evans T, Walsh K. Reversal of GATA-6 downregulation promotes smooth muscle differentiation and inhibits intimal hyperplasia in balloon-injured rat carotid artery. Circ Res 84: 647–654, 1999.[Abstract/Free Full Text]
  19. Mulvihill ER, Jaeger J, Sengupta R, Ruzzo WL, Reimer C, Lukito S, Schwartz SM. Atherosclerotic plaque smooth muscle cells have a distinct phenotype. Arterioscler Thromb Vasc Biol 24: 1283–1289, 2004.[Abstract/Free Full Text]
  20. Owens GK. Regulation of differentiation of vascular smooth muscle cells. Physiol Rev 75: 487–517, 1995.[Abstract/Free Full Text]
  21. Papakonstantinou E, Karakiulakis G, Roth M, Block LH. Platelet-derived growth factor stimulates the secretion of hyaluronic acid by proliferating human vascular smooth muscle cells. Proc Natl Acad Sci USA 92: 9881–9885, 1995.[Abstract/Free Full Text]
  22. Perlman H, Suzuki E, Simonson M, Smith RC, Walsh K. GATA-6 induces p21(Cip1) expression and G1 cell cycle arrest. J Biol Chem 273: 13713–13718, 1998.[Abstract/Free Full Text]
  23. Proweller A, Pear WS, Parmacek MS. Notch signaling represses myocardin-induced smooth muscle cell differentiation. J Biol Chem 280: 8994–9004, 2005.[Abstract/Free Full Text]
  24. Puig O, Tjian R. Transcriptional feedback control of insulin receptor by dFOXO/FOXO1. Genes Dev 19: 2435–2446, 2005.[Abstract/Free Full Text]
  25. Puigserver P, Rodgers JT. Foxa2, a novel transcriptional regulator of insulin sensitivity. Nat Med 12: 38–39, 2006.[CrossRef][Web of Science][Medline]
  26. Ramoni MF, Sebastiani P, Kohane IS. Cluster analysis of gene expression dynamics. Proc Natl Acad Sci USA 99: 9121–9126, 2002.[Abstract/Free Full Text]
  27. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 362: 801–809, 1993.[CrossRef][Medline]
  28. Sakata Y, Kamei CN, Nakagami H, Bronson R, Liao JK, Chin MT. Ventricular septal defect and cardiomyopathy in mice lacking the transcription factor CHF1/Hey2. Proc Natl Acad Sci USA 99: 16197–16202, 2002.[Abstract/Free Full Text]
  29. Sakata Y, Koibuchi N, Xiang F, Youngblood JM, Kamei CN, Chin MT. The spectrum of cardiovascular anomalies in CHF1/Hey2 deficient mice reveals roles in endocardial cushion, myocardial and vascular maturation. J Mol Cell Cardiol 40: 267–273, 2006.[CrossRef][Web of Science][Medline]
  30. Sakata Y, Xiang F, Chen Z, Kiriyama Y, Kamei CN, Simon DI, Chin MT. Transcription factor CHF1/Hey2 regulates neointimal formation in vivo and vascular smooth muscle proliferation and migration in vitro. Arterioscler Thromb Vasc Biol 24: 2069–2074, 2004.[Abstract/Free Full Text]
  31. Shirvani S, Xiang F, Koibuchi N, Chin MT. CHF1/Hey2 suppresses SM-MHC promoter activity through an interaction with GATA-6. Biochem Biophys Res Commun 339: 151–156, 2006.[CrossRef][Web of Science][Medline]
  32. Spin JM, Nallamshetty S, Tabibiazar R, Ashley EA, King JY, Chen M, Tsao PS, Quertermous T. Transcriptional profiling of in vitro smooth muscle cell differentiation identifies specific patterns of gene and pathway activation. Physiol Genomics 19: 292–302, 2004.[Abstract/Free Full Text]
  33. Sun JX, Kamei CN, Layne MD, Jain MK, Liao JK, Lee ME, Chin MT. Regulation of myogenic terminal differentiation by the hairy-related transcription factor CHF2. J Biol Chem 276: 18591–18596, 2001.[Abstract/Free Full Text]
  34. Wang W, Campos AH, Prince CZ, Mou Y, Pollman MJ. Coordinate Notch3-hairy-related transcription factor pathway regulation in response to arterial injury. Mediator role of platelet-derived growth factor and ERK. J Biol Chem 277: 23165–23171, 2002.[Abstract/Free Full Text]
  35. Wang W, Prince CZ, Mou Y, Pollman MJ. Notch3 signaling in vascular smooth muscle cells induces c-FLIP expression via ERK/MAPK activation. Resistance to Fas ligand-induced apoptosis. J Biol Chem 277: 21723–21729, 2002.[Abstract/Free Full Text]
  36. Wang Z, Wang DZ, Hockemeyer D, McAnally J, Nordheim A, Olson EN. Myocardin and ternary complex factors compete for SRF to control smooth muscle gene expression. Nature 428: 185–189, 2004.[CrossRef][Medline]
  37. Xiang F, Sakata Y, Cui L, Youngblood JM, Nakagami H, Liao JK, Liao R, Chin MT. Transcription factor CHF1/Hey2 suppresses cardiac hypertrophy through an inhibitory interaction with GATA4. Am J Physiol Heart Circ Physiol 290: H1997–H2006, 2006.[Abstract/Free Full Text]




This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
30/1/61    most recent
00277.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
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 Google Scholar
Google Scholar
Right arrow Articles by Shirvani, S. M.
Right arrow Articles by Chin, M. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shirvani, S. M.
Right arrow Articles by Chin, M. T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2007 by the American Physiological Society.