Mechanisms Responsible for the Promoter-Specific Effects of Myocardin.

Understanding the mechanism of smooth muscle cell (SMC) differentiation will provide the foundation for elucidating SMC-related diseases such as atherosclerosis, restenosis and asthma. Recent studies have demonstrated that the interaction of SRF with the co-activator myocardin is a critical determinant of smooth muscle development. It has been proposed that the specific transcriptional activation of smooth muscle-restricted genes, as opposed to other SRF-dependent genes, by myocardin, results from the presence of multiple CArG boxes in smooth muscle genes that facilitate myocardin homodimer formation. This proposal was further tested in the current study. Our results show that the SMC-specific telokin promoter, which contains only a single CArG box, is strongly activated by myocardin. Furthermore, myocardin and a dimerization defective mutant myocardin, induce expression of endogenous telokin but not c-fos in 10T1/2 fibroblast cells. Knocking down myocardin by siRNA deceased telokin promoter activity and expression in A10 SMCs. A series of telokin and c-fos promoter chimeric and mutant reporter genes were generated to determine the mechanisms responsible for the promoter specific effects of myocardin. Data from these experiments demonstrated that the Ets binding site (EBS) in the c-fos promoter partially blocks the activation of this promoter by myocardin. However, the binding of Ets factors alone was not sufficient to explain the promoter specific effects of myocardin. Elements 3’ of the CArG box in the c-fos promoter act in concert with the EBS to block the ability of myocardin to activate the promoter. Conversely, elements 5’ and 3’ of the CArG box in the telokin promoter act in concert with the CArG box to facilitate myocardin stimulation of the promoter. Together these data suggest that the promoter specificity of myocardin is dependent on complex combinatorial interactions of multiple cis elements and their trans binding factors. following order to further investigate the mechanisms underlying the promoter-specific effects of myocardin we have compared the ability of myocardin to activate two single CArG box-containing genes, the smooth muscle-specific telokin gene and the widely expressed c-fos gene. Results demonstrate that myocardin and its dimerization deficient LZ (leucine zipper) mutant is capable of strongly trans-activating single CArG box containing, smooth muscle-specific, telokin promoter and to induce telokin expression in 10T1/2 cells although the myocardin LZ mutant is less effective than the wild type myocardin. In contrast, myocardin had no effect on c-fos promoter activity or c-fos gene expression in 10T1/2 cells. Knocking down endogenous myocardin in SMC cells by siRNA decreased telokin promoter activity and endogenous telokin expression. Analysis of a series of chimeric and mutant telokin and c-fos reporter genes demonstrated that in the c-fos promoter the Ets binding site (EBS), which binds ets factors, partially blocks the activation of this promoter by myocardin, however, an additional region between –300 and +39 is required to prevent myocardin activation of the c-fos promoter. Conversely, multiple cis-elements in telokin promoter are required for maximal myocardin activation. We propose that the gene specificity of myocardin is dependent on combinations of multiple positive and negative cis elements and their trans binding factors.


Introduction
There is extensive evidence showing that altered control of the differentiated state of smooth muscle cells contributes to the development and/or progression of a variety of diseases, including atherosclerosis, hypertension and asthma. These diseases are all associated with decreased expression of proteins required for the differentiated function of smooth muscle cells. An understanding of the mechanisms that control smooth muscle cell differentiation is required before it will be possible to determine how these control processes are altered in pathological conditions. 4 upregulated following myocardin infection of rat aortic smooth muscle cells (12). In order to further investigate the mechanisms underlying the promoter-specific effects of myocardin we have compared the ability of myocardin to activate two single CArG boxcontaining genes, the smooth muscle-specific telokin gene and the widely expressed cfos gene.
Results demonstrate that myocardin and its dimerization deficient LZ (leucine zipper) mutant is capable of strongly trans-activating single CArG box containing, smooth muscle-specific, telokin promoter and to induce telokin expression in 10T1/2 cells although the myocardin LZ mutant is less effective than the wild type myocardin. In contrast, myocardin had no effect on c-fos promoter activity or c-fos gene expression in 10T1/2 cells. Knocking down endogenous myocardin in SMC cells by siRNA decreased telokin promoter activity and endogenous telokin expression. Analysis of a series of chimeric and mutant telokin and c-fos reporter genes demonstrated that in the c-fos promoter the Ets binding site (EBS), which binds ets factors, partially blocks the activation of this promoter by myocardin, however, an additional region between -300 and +39 is required to prevent myocardin activation of the c-fos promoter. Conversely, multiple cis-elements in telokin promoter are required for maximal myocardin activation.
We propose that the gene specificity of myocardin is dependent on combinations of multiple positive and negative cis elements and their trans binding factors. QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) (6). All promoter reporter genes were constructed by cloning fragments of promoters into the pGL 2 B luciferase vector (Promega, Madison, WI). The mouse and rabbit telokin promoterluciferase reporter gene used includes nucleotides -190 to +181 (T370) and -256 to +147 (T400), respectively, of the telokin gene as described previously (13). The SM22aluciferase reporter gene includes nucleotides -475 to +61 of mouse SM22a (14,15). The SM a-actin promoter fragment extended from nucleotide -2,555 to +2,813 (9) and the SM-MHC promoter from -4,200 to +11,600 (16). The Egr1 and c-fos luciferase reporter genes spanned from -637 to +79 and -605 to +39, respectively. The minimal TK promoter used comprised nucleotides -113 to +20 of the thymidine kinase gene. All mutant reporter gene constructs were initially generated in pCR pBlunt vector (Invitrogen, Carlsbad, CA) by QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) and then transferred to pGL 2 b vector. The resultant plasmids were sequenced to verify the integrity of the insert. Transfection was carried out as previously described (17). The level of promoter activity was evaluated by measurement of the firefly luciferase relative to the internal control Renilla luciferase using the Dual Luciferase Assay System essentially as described by the manufacturer (Promega, Madison, WI). A minimum of six independent transfections was performed and all assays were replicated at least twice. Results are reported as the mean ± SEM.
Reverse transcription coupled to PCR. Total RNA was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA). A pair of unique primers for telokin was designed as sense 5 ' -G A C A C C G C C T G A G T C C A A C C T C C G -3 ' a n d a n t i s e n s e 5 ' -

Results
Myocardin trans-activates the telokin promoter. In contrast to many smooth muscle  Figure 3C). To confirm that telokin promoter activity is myocardin dependent in SMCs, plasmid-based myocardin siRNA or a scrambled siRNA control pshuttle plasmid were transiently co-tranfected into A10 SMCs together with telokin promoter reporter genes and luciferase activity determined. As shown in figure 3D, the activity of the rabbit telokin promoter, but not the thymidine kinase promoter was significantly reduced to approximately 40% of control levels in A10 cells transfected with either 300ng or 600ng of myocardin siRNA plasmid.
Maximal myocardin activity on the telokin promoter requires multiple ciselements. As both telokin and c-fos promoters contain single CArG boxes we by guest on March 23, 2020 http://www.jbc.org/ Downloaded from determined whether the specific sequences of the CArG box within the telokin promoter contributes to ability of myocardin to activate the promoter. Reporter genes were generated in which the telokin promoter CArG box was mutated to the c-fos gene CArG box sequence or the SM22a gene CArG-near sequence or to a sequence no longer able to bind SRF. These mutant reporter genes were co-transfected together with myocardin and luciferase activity determined ( Figure 4A). Mutant telokin promoter reporter genes containing either a c-fos or SM22a CArG box were activated by myocardin similar to the wild-type telokin promoter. As expected a mutant telokin promoter that was unable to bind SRF showed no activation by myocardin, showing that the intact CArG is critical for the myocardin activation ( Figure 4A). These data demonstrated that SRF binding to the CArG box is necessary for myocardin activation of the telokin promoter but the sequence of the CArG box does not explain the ability of myocardin to activate the telokin promoter as opposed to the c-fos promoter.
To define the minimal regions of the telokin promoter required for myocardin activation, the ability of myocardin to activate a series of deletion constructs was determined ( Figure 4B). Results from this analysis suggest that the regions between -80 and -66 (an AT-rich region) and between +36 to +82 are important for myocardin activation. In contrast, deletion of residues -190 to -80 or +82 to +171 did not alter the ability of myocardin to activate the promoter, suggesting that these regions are not important for this effect. Deletion of the region from +36 to +82 or from -80 to -66 decreased the ability of myocardin to activate the promoter over 10-and 20-fold, respectively. These data demonstrated that the CArG box together with regions from +36 to +82 and -80 to -66 are necessary for myocardin activation of the telokin promoter. To determine if these regions are sufficient to confer myocardin activation, the telokin CArG box, -66 to -80 region (AT-rich region) and +36 to +82 region were fused to a minimal TK promoter, alone or in combination. Each of these regions alone was not sufficient to confer a large amount of myocardin activation ( Figure 4C). Although the CArG element alone increased activation to 11-fold, when all three elements were present the ability of myocardin to activate the minimal TK promoter was increased to 50-fold. These data suggest that multiple cis-elements of telokin promoter are necessary and largely sufficient to confer maximal activation by myocardin.

The ets binding site (EBS) in c-fos promoter partially blocks myocardin
activation. It has been reported that the SRF binding affinity of the c-fos CArG box is higher than the SM22a CArG boxes and the variations among CArG boxes of c-fos and SM22a influence cell type specificity of expression (19). To determine if the specific sequence of the c-fos CArG box is important for the lack of response of this promoter to myocardin, the CArG box was mutated to the telokin CArG box sequence or to a sequence unable to bind SRF. Analysis of these mutant reporter genes demonstrated that c-fos promoters containing either the native or telokin CArG box sequence were poorly activated by myocardin ( Figure 5B) and, as expected, a mutant c-fos promoter that was unable to bind SRF showed no activation by myocardin. These data together with those obtained from the mutant telokin promoters described in figure 4A, suggest that the precise sequence of the CArG boxes in the c-fos and telokin promoters does not account for the promoter-specific effects of myocardin.
It has recently been reported that growth signals can repress smooth muscle-specific genes by triggering the displacement of myocardin from SRF by ELK1, an ets family member that competes for the myocardin docking site on SRF through a structurally related SRF-binding motif (20,21). binding site adjacent to the CArG box, sequence alignment between the telokin and cfos promoters revealed a significant degree of sequence similarity in this AT-rich region ( Figure 5A). This sequence similarity allowed us to determine if the sequences immediately 5' of the CArG boxes are important for the promoter-specific effects of myocardin. When the EBS region in the c-fos promoter was mutated to the corresponding sequence in the telokin promoter the mutant c-fos promoter remained largely refractory to myocardin activation (activation was increased from 4-fold to 10fold, figure 5B). In addition, when the corresponding region of the telokin promoter was mutated to match the EBS and surrounding nucleotides of the c-fos promoter this did not prevent myocardin from activating the promoter ( Figure 5C). Similar results were obtained when the CArG box sequences were also switched in conjunction with the ATrich/EBS sequences ( Figure 5C). These data suggest that there are additional regulatory regions within the c-fos promoter that prevent myocardin activation of the promoter. To begin to further identify these regions a truncated promoter was generated (-324 to +39) in which the region 5' of the ets binding site was deleted. This truncated construct had similar myocardin activation to the wild type c-fos promoter (Figure 6 B).
In addition, changing the EBS and CArG box from the c-fos promoter to corresponding telokin promoter sequences, within the context of this truncated promoter, had no further effect on myocardin activation ( Figure 5B). In a reciprocal experiment changing the AT-rich region and CArG box of the telokin promoter to the corresponding sequences in the c-fos promoter, within the context of a -80 to +82 minimal telokin promoter did not prevent myocardin from activating the telokin promoter ( Figure 5C).
Together these data suggest that sequences in the c-fos promoter between -300 and +39 and between -55 and +82 of the telokin promoter are responsible for the promoter specific effects of myocardin on these two genes.
To determine if the telokin +36 to +82 region is sufficient to confer myocardin responsiveness to the c-fos gene, this fragment was added to c-fos promoter and the ability of myocardin to activate the promoter determined. This chimeric promoter showed only a small increase in myocardin activation to 8-fold as compared to the 4- fold activation of the wild-type c-fos promoter ( Figure 5D). When the +36 to +82 region was added in combination with the telokin AT-CArG sequence, no further activation of the promoter was observed. These data imply that the positive elements within the telokin promoter are not able to override the negative elements located between -300 to +39 of the c-fos promoter.

Discussion
Our data demonstrate that myocardin increases telokin expression through a CArGdependent mechanism that requires the cooperative activity of multiple cis-acting regulatory elements. Conversely the inability of myocardin to activate the growth factor responsive c-fos gene appears to result from both the lack of these key cooperative positive regulatory elements together with the presence of multiple negative elements that help prevent myocardins' activation of the promoter.
It has been proposed that the ability of myocardin to specifically activate cardiac and smooth muscle-specific genes is dependent on cooperative interaction of pairs of CArG boxes. This would explain why growth-regulated genes such as c-fos, that contain a single CArG box, are not activated by myocardin (6). However, another early growth response gene, Egr-1, that has 5 CArG boxes located in the 5' flanking promoter sequence, was not activated by myocardin ( Figure 1A) or MRTF-A (7,8). In addition, the proximal SM a-actin or SM myosin heavy chain promoters, that each contain two CArG boxes are not sufficient to drive SMC-specific transgene expression (9,10) whereas the telokin promoter which contains only one CArG box is sufficient to drive SMC-specific expression in transgenic mice (22). In the current study, we have further shown that the telokin promoter is strongly activated by myocardin (Figure 1, 4A), that myocardin can activate endogenous telokin expression in 10T1/2 cells ( Figure 2) and that knocking down endogenous myocardin in SMC cells decreases telokin expression ( Figure 3).
Furthermore, although a myocardin LZ mutant, which is not able to dimerize, activated the telokin promoter ( Figure 2D) and induced endogenous telokin expression in 10T1/2 fibroblast cells (Figure 2A,C) it did so much less effectively compared to the wild type myocardin. These data suggest that a myocardin monomer is sufficient to induced telokin and other smooth muscle-specific gene expression when expressed at high levels. However, at more physiological levels of expression it is likely that the ability of myocardin to dimerize is important for its ability to activate smooth muscle genes, including those such as the telokin gene, that contain only a single CArG box in their promoter regions. Taken together, these data would suggest that the ability of paired CArG box elements to promote myocardin dimerization is not sufficient to account for the smooth and cardiac muscle-specific effects of myocardin.
Although siRNA mediated knockdown of myocardin resulted in decreased telokin, SM22a and calponin expression in A10 cells no changes in the level of expression of SM a-actin or the 130kDa MLCK were observed (Figure 3). These latter findings are puzzling in light of our data ( Figure 2) and that of others that have shown that myocardin induces expression of SM a-actin and the 130kDa MLCK in 10T1/2 cells. At least one explanation for this apparent discrepancy could be that in A10 cells much of the expression of SM a-actin and the 130kDa MLCK may be occurring through myocardin independent mechanisms. This is particularly likely for these two proteins as expression of neither protein is restricted to smooth muscle cells. For example, SM a-actin is expressed in skeletal muscle myoblasts that do not express myocardin and the 130kDa MLCK is expressed in most adult cell types (23,24). Together these data suggest that the expression of SM a-actin and the 130kDa MLCK may occur by myocardin dependent pathways in some cell types and by myocardin independent pathways in other cells that do not express myocardin.
Consistent with previous reports (2-4), our results demonstrate that an intact CArG element is required but not sufficient for telokin promoter activation by myocardin ( Figure 4). Although essential for myocardin activation the precise sequence of the CArG box has little effect on the ability of myocardin to activate the promoter. Within the telokin promoter at least two additional regions (-80 to -66 and +36 to +82) are required to act in concert with the CArG box to facilitate high levels of promoter activation by myocardin ( Figure 4). Although this combination of elements is sufficient to confer a significant amount of myocardin activation to a minimal thymidine kinase promoter, these elements are not sufficient to confer increased myocardin responsiveness to the c-fos promoter ( Figure 5D). These data would suggest that, in addition to lacking key positive acting cis-regulatory elements, the c-fos promoter also contains a negative regulatory region located between nucleotides -300 and +39 that blocks the activity of these positive elements.
Based on a recent report demonstrating that growth signals can repress smooth muscle-specific genes by triggering the displacement of myocardin from SRF by ELK1, an ets family protein, it is logical to propose that the inability of myocardin to activate the c-fos promoter is likely to be due to binding of ets to the serum response element of the c-fos promoter (20,21). Similarly, there are multiple ets binding sites surrounding the CArG boxes in the Egr-1 promoter that may be responsible for the inability of myocardin to activate this promoter despite the presence of multiple CArG boxes (25). However, although our data demonstrated that the ets binding site in the c-fos promoter is partially involved in inhibiting the activation of this promoter by myocardin, when the ets-binding site and CArG box in the c-fos promoter were replaced with the corresponding sequences from the telokin promoter the mutant c-fos promoter remained refractile to myocardin stimulation ( Figure 5). Conversely when the AT-rich region and CArG box from the telokin promoter were replaced with the ets binding site and CArG box from the c-fos promoter the mutant telokin promoter remained strongly activated by myocardin.
Together these data would suggest that the regions 3' of the CArG boxes of the c-fos (-300 to +39) and telokin (-55 to +82) promoters, rather than the EBS/AT-rich sequences or CArG boxes, are critical for determining the promoters' responsiveness to myocardin.
Presumably each region binds to specific trans-acting factors that interact with myocardin and/or SRF to modify their function. The -55 to +82 region in the telokin gene is highly conserved between species being 90% identical between mouse, rabbit and human genes also suggesting that this region contains important regulatory elements (26). The identity of the key regulatory factors is currently unknown, however, it is tempting to speculate that these factors may comprise part of the transcription initiation complex that forms on each gene. This proposal arises from the observation that the putative regulatory regions span the transcription initiation sites and that the telokin and c-fos promoters utilize different cis-elements to initiate transcription. The telokin promoter initiates transcription from multiple start sites spanning approximately 70-80bp in a TATA-independent manner whereas the c-fos gene is a classical TATAdependent gene with a single transcription initiation site. Although the core components of the transcription initiation complex will be identical on both genes, additional accessory factors may be promoter specific. This raises the possibility that specific components of the transcription initiation complexes may be required for myocardin to strongly activate transcription.
In summary, our results suggest that the promoter-specific effects of myocardin, and likely also MRTF-A, result from a complex interaction of positive regulatory elements that include at least one CArG box in responsive genes together with negative regulatory elements in unresponsive genes that block the activity of myocardin family members.