Molecular Endocrinology, doi:10.1210/me.2006-0025
Molecular Endocrinology 20 (6): 1261-1275
Copyright © 2006 by The Endocrine Society
The Peroxisome Proliferator-Activated Receptor N-Terminal Domain Controls Isotype-Selective Gene Expression and Adipogenesis
Sarah Hummasti and
Peter Tontonoz
Howard Hughes Medical Institute, Molecular Biology Institute and Department of Pathology and Laboratory Medicine, University of California, Los Angeles, California 90095
Address all correspondence and requests for reprints to: Peter Tontonoz, M.D., Ph.D., Howard Hughes Medical Institute, UCLA School of Medicine, Box 951662, Los Angeles, California 90095-1662. E-mail: ptontonoz{at}mednet.ucla.edu.
 |
ABSTRACT
|
|---|
Peroxisome proliferator-activated receptors (PPAR
, PPAR
, and PPAR
) are important regulators of lipid metabolism. Although they share significant structural similarity, the biological effects associated with each PPAR isotype are distinct. For example, PPAR and PPAR regulate fatty acid catabolism, whereas PPAR controls lipid storage and adipogenesis. The different functions of PPARs in vivo can be explained at least in part by the different tissue distributions of the three receptors. The question of whether the receptors have different intrinsic activities and regulate distinct target genes, however, has not been adequately explored. We have engineered cell lines that express comparable amounts of each receptor. Transcriptional profiling of these cells in the presence of selective agonists reveals partially overlapping but distinct patterns of gene regulation by the three PPARs. Moreover, analysis of chimeric receptors points to the N terminus of each receptor as the key determinant of isotype-selective gene expression. For example, the N terminus of PPAR confers the ability to promote adipocyte differentiation when fused to the PPAR DNA binding domain and ligand binding domain, whereas the N terminus of PPAR leads to the inappropriate expression of fatty acid oxidation genes in differentiated adipocytes when fused to PPAR. Finally, we demonstrate that the N terminus of each receptor functions in part to limit receptor activity because deletion of the N terminus leads to nonselective activation of target genes. A more detailed understanding of the mechanisms by which the individual PPARs differentially regulate gene expression should aid in the design of more effective drugs, including tissue- and target gene-selective PPAR modulators.
|
INTRODUCTION
|
|---|
CONTROL OF LIPID metabolism is a complex process, the dysregulation of which is linked to a range of metabolic disorders including diabetes, atherosclerosis, obesity and hypertension. The peroxisome proliferator-activated receptors (PPAR, PPAR, and PPAR), members of the nuclear receptor superfamily, are key regulators of lipid and carbohydrate metabolism (1, 2, 3). Synthetic ligands for PPAR (fibrates) and PPAR (thiazolidindiones) are in widespread clinical use for the treatment of hyperlipidemia and diabetes, respectively (4, 5, 6). Thus, elucidating the function of these receptors is important for a thorough understanding of both normal physiology and disease.
A decade of research on PPARs has established that the three family members serve as mediators of lipid signaling in multiple tissues and cell types (1, 2, 3, 7). In response to the binding of fatty acids and fatty acid derivatives, these transcription factors regulate the expression of genes involved in lipid and energy homeostasis. However, the specific biological functions associated with each isotype are distinct. Extensive investigation of PPAR has established an essential role for this protein in both adipogenesis and adipocyte function. Ectopic expression of PPAR is sufficient to drive the adipogenic program, and loss of PPAR expression renders cells incapable of becoming adipocytes (8, 9, 10, 11). On the other hand, PPAR is well known for its role in fat catabolism in the liver. PPAR ligands have been shown to induce expression of genes involved in fatty acid uptake and ß-oxidation (1, 12). The role of PPAR is less well characterized, but increasing evidence suggests a role in the control of fatty acid oxidation in many tissues including adipose tissue and skeletal muscle (13, 14). Treatment with a PPAR-specific ligand has been shown to improve lipid profiles in mice and monkeys (15, 16).
Although different functions have been attributed to the individual PPARs, their mode of action is similar. They all bind as heterodimers with the retinoid X receptor (RXR) to PPAR response elements (PPREs) in the promoters of target genes where they recruit coactivators to regulate transcription (17). In addition, studies suggest that all three PPARs bind to similar DNA response elements. It is therefore unclear what is responsible for the different actions associated with the individual PPAR isoforms, including such opposing processes as fat utilization by PPAR and PPAR and fat storage by PPAR. These distinct activities can be explained to some extent by their different patterns of expression: PPAR in the liver, kidney and heart; PPAR predominantly in fat and macrophages; and PPAR ubiquitously (9, 18). However, it is also possible that each PPAR isotype has specific target genes that contribute to its distinct biological effects.
Despite a decade of work on PPARs, this issue of isotype-specific target genes has not been rigorously explored. Although prior studies in the literature have addressed the issue indirectly, all of these suffer from one of the following limitations: 1) highly specific PPAR ligands were not used; 2) cells were treated with ligands without taking into account the level of expression of each receptor in the cell (e.g. PPAR is not expressed in many cell types, so the observation that PPAR ligands do not induce target genes in these cells is uninformative); and 3) PPAR-specific ligands were used without any demonstration that their effects were PPAR mediated. We have taken a systematic approach to the characterization of PPAR-dependent gene expression. We have used retroviral vectors to develop fibroblast cell lines that express comparable amounts of each PPAR, and we have characterized gene expression in these cells in the presence of highly specific ligands for each receptor. In the present study, we show that although many of the established PPAR target genes are equally responsive to all three receptors, preferential targets do exist for each isotype. In addition, we outline a previously unrecognized role for the N terminus of the receptor in controlling target gene specificity and limiting adipogenic activity.
|
RESULTS
|
|---|
The ability of individual PPAR isotypes to regulate specific target genes has not been adequately explored. Such analysis has been limited by the lack of a cell line or tissue where all three PPARs are expressed at similar levels. To overcome this limitation, we used retroviral expression vectors to ectopically express each of the three PPAR isotypes in NIH-3T3 cells (Fig. 1A
). Individual stable clones exhibit large variability in gene expression, as exemplified by differences in their ability to undergo adipogenesis. The use of the retroviral system circumvents this problem by allowing large pools of stable clones to be analyzed. Furthermore, the level of receptor expression achieved using this system is within a physiological range. For example, the level of PPAR expression in NIH-PPAR cells is comparable to that observed in adipose tissue in vivo (Fig. 1B) (8). Although PPAR is expressed at a low level in NIH-3T3 cells, the level of retrovirally expressed PPAR is substantially higher, limiting the influence of endogenous PPAR. Thus, this approach facilitates analysis of the activity of each receptor in the same cell type at comparable levels of expression (Fig. 1).
Once cell lines expressing each PPAR isotype were established, ligand-responsive gene expression was analyzed by treating confluent cells with ligand or vehicle [dimethyl sulfoxide (DMSO)] control. The synthetic ligands GW409544 (PPAR), GW7845 (PPAR), and GW501516 (PPAR) were used at a concentration (100 nM) at which they are more than 100-fold selective (data not shown). RNA was isolated after 48 h of ligand treatment and subjected to microarray analysis using Affymetrix (Santa Clara, CA) arrays. This time point allows sufficient time for the PPARs to activate gene expression, while keeping the focus on early effects of each nuclear receptor (e.g. before induction of adipocyte differentiation by PPAR). The results of the microarray analysis, presented in the Venn Diagram in Fig. 1C
, demonstrate that many target genes are common to all three receptors. Such common targets include many established PPAR-responsive genes such as the fatty acid binding protein aP2 and CD36 (Table 1). In addition, unique targets for each receptor were predicted by the array analysis (27 for PPAR, 33 for PPAR, and 93 for PPAR). Some of these are also known targets (for example, adiponectin for PPAR and glycerol kinase for PPAR). However, novel targets were also revealed for each receptor, such as the growth suppressor HRAS like suppressor 3 (HRASLS3, a PPAR target), and the Wnt signaling molecule frizzled 7 (FZD7, a PPAR target; Table 1). Moreover, PPAR and PPAR were found to share a striking number of targets, 54, many of which are involved in fatty acid oxidation pathways. Far fewer genes were found to be common targets between PPAR and either PPAR or PPAR (11 and 15, respectively). The entire microarray data set is presented in Supplemental Table 1, which is published on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org.
Results of the microarray analysis were confirmed by quantitative real-time PCR for a subset of target genes (Fig. 2). All of the gene expression assays presented in this paper were performed on duplicate or triplicate samples and repeated with at least two independently derived stable cell lines. Expression of the relatively nonselective targets, aP2, perilipin, uncoupling protein 2, and adipose differentiation-related protein (ADRP) in NIH-vector, -PPAR, -PPAR and -PPAR cell lines is shown in Fig. 2A. Note that even within this group slight preferences for one receptor can be recognized. For example, aP2 responds best to PPAR, whereas ADRP responds best to PPAR. Nevertheless, the behavior of these genes is clearly distinguished from those shown in Fig. 2, B–E. Genes selectively responsive to PPAR (adiponectin, HRASLS3, Tob1) are shown in Fig. 2B. Genes selectively responsive to PPAR (17-ß-HSD and glycerol kinase) are shown in Fig. 2C. Genes selectively responsive to PPAR (Fzd7 and PDK4) are shown in Fig. 2D. Finally, genes that respond comparably to both PPAR and PPAR (CYP4B1 and HMGCS2) are shown in Fig. 2E. Note, no reproducible difference in target gene regulation was observed between the two PPAR isoforms, PPAR1 and PPAR2 (Fig. 2 and data not shown).
The approach outlined above allowed us to identify isotype-selective targets for PPAR, PPAR, and PPAR in a controlled system. We next addressed the question of how this differential regulation of gene expression is achieved. Chimeric PPAR expression constructs were generated in which either the N-terminal region or both the N terminus and DNA binding domain (DBD) were swapped as depicted in Fig. 3A. These chimeric constructs were cloned into retroviral vectors and used to create stable NIH-3T3 cell lines. Real-time PCR analysis of mRNA expression levels revealed comparable levels of mRNA encoding each receptor (Fig. 3B). Moreover, the chimeric constructs were functional transcription factors, as evidenced by their ability to regulate the target gene aP2, when ligand for the appropriate ligand binding domain (LBD) was added (Fig. 3, C and D).
Next, we compared gene expression profiles between cells expressing PPAR, PPAR, and the chimeric constructs. Consistent with the results shown above, PPAR-selective targets, represented here by HRASLS3, were regulated by ligand in PPAR- but not PPAR-expressing cell lines (Fig. 3C). Likewise, PPAR-specific targets, such as Cyp4B1, were regulated by PPAR, but not PPAR. Swapping the N-terminal domain resulted in loss of isotype-selective target gene expression by both PPAR and PPAR. Specifically, replacing only the amino terminus of PPAR with that of PPAR resulted in the ability of PPAR to regulate the PPAR-selective target Cyp4B1 (asterisk, Fig. 3C lower panel). However, this same chimeric construct did not lose its ability to regulate the PPAR target gene HRASLS3, despite gaining the ability to regulate PPAR-selective targets. Similar results were seen when replacing the PPAR N-terminal region with that of PPAR. This chimeric receptor gained the ability to regulate HRASLS3 (asterisk, Fig. 3C, middle panel). Replacing the DBD along with the N terminus had no additional affect on target gene regulation when compared with swapping the N terminus alone (Fig. 3D). This result suggests that it is the N-terminal domain and not the DBD that is the primary determinant of isotype-selective regulation. Similar results were obtained for the other isotype-selective targets examined, including the nonselective targets perilipin and PPAR angiopoietin-related protein, the PPAR-specific targets adiponectin and Tob1, and the PPAR/-selective target HMGCS2 (data not shown).
We also used FLAG-tagged receptors to facilitate determination of receptor protein expression in each cell line. As with the untagged receptors shown in Fig. 3, real-time PCR analysis confirmed comparable mRNA expression for each receptor (Fig. 4A). Moreover, Western analysis confirmed that FLAG-tagged chimeric vectors gave rise to comparable levels of wild-type and chimeric PPAR protein expression in the stable cell lines (Fig. 4B). Analysis of target gene expression in cell lines expressing the FLAG-tagged cell receptors yielded similar patterns of target gene expression as observed with the untagged receptors (Fig. 4C). As in Fig. 3, swapping the N terminus of PPAR and PPAR altered the isotype-selective pattern of target gene expression.
The results of Fig. 3 strongly suggest that the N terminus and not the DBD of PPARs is the key determinant of isotype selectivity. To further address this issue, we examined the ability of each PPAR isotype to bind to the PPREs of various target genes. EMSAs were performed using equivalent amounts of in vitro-translated PPAR, PPAR, or PPAR in the presence or absence of RXR. The sequence of each of the PPREs analyzed is shown in Fig. 5A. All three PPARs bound to PPREs from the aP2, adiponectin, CYP4B1, and HMGCS2 genes (Fig. 5B), although modest differences in affinity for the various elements were noted. For example, PPAR generally bound less well to most of the response elements. However, these differences did not correlate with the response of the endogenous target genes to the different PPARs. PPAR bound relatively better to the PPRE from the nonselective target aP2 than to the PPRE from the PPAR-selective targets CYP4B1 and HMGCS2. Together with the results of the chimeric receptor studies (Figs. 3 and 4), these data indicate that differences in DNA binding and PPRE recognition cannot account for the isotype selectivity of PPAR target genes.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 5. PPAR Isotype-Selective Gene Expression Does Not Correlate with Differential Binding to PPREs
A, Sequence comparison of PPREs from the aP2, adiponectin, CYP4B1, and mHMGCS2 genes. B, EMSAs of PPAR/RXR heterodimer binding to various PPREs. In vitro-translated PPAR and/or RXR were incubated with radiolabeled PPREs as indicated. Comparable amounts of PPAR isotype were used as determined by incorporation of labeled methionine during in vitro translation. DR, Direct repeat.
|
|
The above results pointed to an important role for the amino terminus of PPARs in differential regulation of gene expression. However, it remained unclear whether it was the loss of the cognate N terminus or the addition of the foreign N terminus that resulted in the gained ability to regulate previously isoform-specific targets. To address this question, we created additional constructs in which the amino terminus was deleted. Real-time PCR confirmed comparable levels of receptor mRNA in each cell lines (Fig. 6A). Surprisingly, deletion of the N terminus of PPAR resulted in regulation of the PPAR targets HMGCS2 and CYP4B1, whereas loss of the N terminus of PPAR led to up-regulation of the PPAR target HRASLS3 (Fig. 6B). These data strongly suggest that the N-terminal region of the receptor has a negative regulatory function that prevents certain isotypes from regulating certain targets. With the N terminus removed, each PPAR becomes a permissive receptor that regulates all target genes, at least in the context of the NIH-3T3 system employed here. However, the N terminus does not suppress target gene expression in the chimeric constructs (Fig. 6), suggesting that this repressive effect occurs only in the context of the cognate receptor.
Next, we endeavored to correlate our data on PPAR isotype-selective gene expression with the biological process of adipocyte differentiation. As previously reported, PPAR but not PPAR promotes lipid accumulation (as assessed by Oil Red O staining) in NIH-3T3 cells (Fig. 7). Previous work with PPAR in the NIH-3T3 system used low potency nonselective agonists such as carbaprostacyclin (19). Our data confirm that PPAR has no adipogenic potential in this system, even in the presence of the highly potent and selective agonist GW1516. It follows that target genes specifically downstream of PPAR must be important for the adipocyte conversion of these cells. Indeed, swapping the amino terminus between PPAR and PPAR resulted in two chimeric receptors (N-/ and N-/) that each promoted differentiation when treated with ligand for 5 d (Fig. 7). This activity is consistent with the ability of these constructs to regulate PPAR-selective target genes (Figs. 3 and 4).
Surprisingly, deletion of the N terminus of PPAR, which also results in regulation of PPAR targets, also led to lipid accumulation (Fig. 8). Thus, contrary to what one might have predicted, the PPAR receptor DBD and LBD do possess the inherent capacity to activate adipogenic gene expression, but the N-terminal domain of the receptor prevents this from occurring in the context of the holoreceptor. On the other hand, PPAR maintained both its ability to induce differentiation (Fig. 8) and to regulate PPAR-specific target genes (Figs. 6), despite removal of the N-terminal region or replacement of the amino terminus with that of PPAR. Chimeric constructs in which both the N terminus and the DBD were swapped showed comparable differentiation potential to those in which only the N terminus was swapped (data not shown). Similar results were also obtained when cell lines expressing the FLAG-tagged chimeric receptors (Fig. 4) were assayed for their potential to undergo adipocyte differentiation (data not shown).
Unexpectedly, microscopic examination of the differentiated 3T3 cultures revealed morphological differences in the size and extent of lipid accumulation in the different cell lines (Fig. 9A). In fact, both the chimeric constructs and the amino terminal deletions exhibited smaller lipid droplets, indicating that the lipid content per cell was decreased. Because PPAR is known for its role in lipid accumulation, whereas PPAR is thought to be involved in fatty acid oxidation, we hypothesized that a cycle of lipid accumulation and break down could be occurring in cells expressing the chimeric receptors as a result of simultaneous activation of both the PPAR and PPAR gene programs. To test this hypothesis, we collected RNA from the various cell lines at d 5 (once they had accumulated lipid). Indeed, the altered lipid accumulation in the chimeric receptor-expressing cells lines correlated closely with altered expression of lipid metabolic genes (Fig. 9B). As expected, PPAR, PPAR, and both chimeric receptors promoted expression of the nonspecific target aP2. Thus, although aP2 is commonly used as a marker of differentiation, our results emphasize that aP2 induction alone is not necessarily indicative of differentiation. The PPAR-expressing cells in our study were completely undifferentiated yet expressed aP2 (see Figs. 8 and 9A). On the other hand, expression of the adipocyte marker adipsin and endogenous PPAR gene, which are not directly regulated by PPARs, accurately reflected the differentiated state of the various cell lines (Fig. 9 and data not shown). Consistent with their morphological differentiation, cells overexpressing PPAR or the N-/ or N-/ chimeric receptors, exhibited increased expression of adipocyte-associated genes (adipsin, perilipin, lipoprotein lipase) compared with PPAR or control cells (Fig. 9B and data not shown). Moreover, swapping the N-terminal domains of PPAR and PPAR led to the inappropriate coexpression of both PPAR- and PPAR-specific target genes in differentiated adipocytes (Fig. 9B). Both the N-/ and N-/ chimeric lines showed increased expression of genes involved in fatty acid catabolism (AOX, HMGCS2, PDK4, LCAD). Note, the cells expressing the chimeric receptor with the N terminus of PPAR fused to PPAR (N-/) not only had the smallest lipid droplets (Fig. 9A) but also expressed these catabolic genes to the highest degree (Fig. 9B). Once again, swapping the DBD along with the N terminus had no additional affect on regulation of target genes (not shown).
|
DISCUSSION
|
|---|
Numerous studies over the past 10 yr have advanced our understanding of the role of PPARs in the control of metabolic gene expression (1, 2, 3, 7). Much of this work has focused on common mechanisms of PPAR action, such as their binding as RXR heterodimers to similar sequences in target gene promoters. Yet only limited data exists to address the issue of how (or even if) the PPARs differ in the mechanism by which they regulate gene expression. The distinct biological actions of these receptors have led many in the field to assume that such differences must exist. However, whether this distinct biology results from the differential tissue expression of the PPARs or to intrinsic differences between the receptors, such as selective cofactor binding or subtle differences in promoter preference, remains unclear.
To examine intrinsic differences in activity, we compared the ability of each PPAR to induce target gene expression when expressed at comparable levels in the same cell type. Our results show that although the PPARs share many targets in common, preferential targets also exist for each receptor. Overall, our results on isotype preference in NIH-3T3 cells correlate remarkably well with established patterns of PPAR target gene expression in vivo. For example, many of the genes most highly regulated by PPAR are prominently expressed in liver and involved in pathways of fatty acid utilization. On the other hand, the observation that PPAR poorly activates genes involved in lipid catabolism is consistent with the key role of PPAR in promoting lipid accumulation and adipogenesis. It should be noted that the approach employed here does not distinguish between absolute and relative selectivity in PPAR-dependent gene expression. We have classified genes as selective or nonselective based on an arbitrarily defined cut-off (>2-fold induction by ligand) in our array studies. In most cases, it is likely that we are observing relative rather than absolute selectivity. We cannot exclude the possibility that higher levels of receptor expression or a different agonist might confer greater induction of a particular gene. At the same time, the difference in the ability of PPAR and PPAR to promote differentiation is absolute, and therefore relative gene expression differences must be biologically meaningful.
Many of the targets we identified by microarray analysis, both shared and selective, are well-characterized PPAR targets (1, 2, 3). For example, aP2, CD36, ADRP, and PPAR angiopoietin-related protein are induced by all three PPARs. On the other hand, our results suggest that induction of adiponectin is selective for PPAR, 17ß-hydroxysteroid dehydrogenase is selective for PPAR, and LCAD is selective for PPAR. Consistent with the known functions of these receptors, PPAR- and PPAR-selective targets include numerous genes in the peroxisomal and mitochondrial fatty acid oxidation pathways (12, 13, 20). Furthermore, PPAR and PPAR share a striking number of common targets, likely reflective of their similar biological roles.
Our results indicate that PPARs possess significant inherent capacity to control the expression of highly tissue-specific genes even when expressed in a heterologous cell type. The ability of PPAR to induce expression of adipocyte genes in NIH-3T3 cells is well documented (8, 19). However, we have further shown that expression of PPAR in fibroblasts induces the expression of a number of catabolic genes whose expression is normally limited to hepatocytes. Thus, although tissue-specific cofactors are also likely to exert a level of control on PPAR-dependent gene regulation, some degree of selectivity can be achieved even with the coactivator set expressed in a generic fibroblast line such as NIH-3T3. Together, our results strongly suggest that intrinsic differences in the transcriptional activities of the three PPARs make a significant contribution to their different biological activities.
A minority of the genes that we classify here as isotype-selective have been characterized differently in the literature by other groups. For example, previous work from Lazar and colleagues (21, 22) has reported that glycerol kinase is responsive to PPAR ligands and that the PPRE in the promoter binds PPAR. Our results suggest that glycerol kinase is preferentially responsive to PPAR in the NIH-3T3 system. This apparent discrepancy may simply be the result of quantitative differences. We do not contend that PPAR has no ability to regulate glycerol kinase expression, only that this ability is weaker than that of PPAR in our system. Note, we did observe an increase in glycerol kinase in response to PPAR ligand in Table 1; however, this increase fell below our arbitrary cut-off (1.5-fold for PPAR vs. 5.3-fold for PPAR). Our results do not exclude the possibility that under different conditions (e.g. higher PPAR expression or different agonist) induction of glycerol kinase by PPAR would be observed. Alternatively, it is possible that certain coactivators necessary for regulation of this gene in adipocytes are not present in NIH-3T3 cells. In general, the isotype preference we observe is consistent with the expression pattern of PPAR target genes under physiological conditions where only endogenous activators are present. For example, glycerol kinase and fatty acid oxidation genes such as acyl-coenzyme A oxidase are most highly expressed in liver (where PPAR is high) and poorly expressed in fat (where PPAR is high). However, these genes are significantly induced by pharmacolgist agonists in adipose tissue in vivo (21, 22, 23), albeit typically to a lower level of expression than observed for the same genes in liver. It seems likely that the use of highly efficacious synthetic agonists may overcome isotype preferences in certain contexts, and that this may contribute to their pharmacological effects.
In addition to clarifying the isotype-selective response of many known PPAR targets, we also present a list of previously unidentified target genes for each receptor. These genes may be unknown players in established pathways regulated by PPARs or may provide clues to novel functions of the individual PPARs. For example, although all three PPARs play roles in lipid metabolism, only PPAR efficiently promotes adipocyte differentiation. It follows that selective PPAR targets must be critical for adipogenesis. Surprisingly, a number of the known adipocyte-associated target genes appear to be regulated by all three isoforms despite the unique role of PPAR in adipogenesis (e.g. aP2, CD36, ADRP). The PPAR-selective targets identified here should now be evaluated for potential roles in the differentiation process. On the other hand, PPAR appears to regulate several members of the Wnt signaling pathway (including FZD7 and LRP5), suggesting a previously unrecognized role for PPAR in regulation of this pathway. It is also possible that genes controlled by PPAR and PPAR may be antagonistic to adipocyte differentiation. At the very least, the ability of these receptors to induce expression of fatty acid catabolic genes limits lipid accumulation in adipocytes (Fig. 9). A more detailed understanding of the function of PPAR isotype-selective targets is expected to provide further insight into the unique role of these three proteins in lipid and energy homeostasis.
The data presented here also point to an important function for the N-terminal domain of PPARs in controlling target gene activation. One might reasonably have expected the DBD and/or LBD to determine the specificity of PPAR target gene activation. Indeed, prior studies have reported differences in the binding affinities of the PPARs to various PPREs (19). However, our data our inconsistent with the suggestion that differences in DNA binding contribute to differential gene regulation. It is also reasonable to suggest that the different PPAR LBDs may interact with different coregulator proteins (21). This may be true, but our data also show that the LBD alone is not sufficient to define isotype-specific gene expression. Using chimeric constructs in which various domains were swapped between PPAR and PPAR, we have shown the N-terminal activation function (AF)-1 domain to be responsible for both differences in regulation of gene expression and the ability to direct adipogenesis. Thus, the distinct biological activity of the PPARs not only results from distinct expression patterns, but also stems from intrinsic differences localized to the N-terminal region of each receptor.
Several prior studies have investigated the importance of the N-terminal domain of PPAR for adipogenesis. Deletion of the N terminus was previously reported to lead to increased adipogenic activity of the receptor (8), and phosphorylation of PPAR in the N terminus by MAPK was shown to inhibit differentiation (24, 25). Spiegelman and colleagues (26) also reported that fusion of the N terminus of PPAR to PPAR results in a receptor capable of promoting differentiation and linked this gain of function to the observation that the N terminus of PPAR interacts with the coactivator PPAR coactivator-2. However, our study is the first to correlate PPAR adipogenic activity with control of gene expression. Taken together, our data do not support the hypothesis that the N terminus of PPAR has a unique function that confers adipogenic activity. First, we showed that a receptor containing the N terminus of PPAR and the PPAR DBD/LBD still promotes differentiation. Second, we established that the N terminus actually serves to restrict the transcriptional activity of PPAR to its particular target gene set. Surprisingly, deletion of the N terminus does not result in loss of function; rather, receptors lacking the N-terminal domain gain the ability to activate new target genes. For example, loss of the PPAR N-terminal domain (either by deletion or replacement with that of PPAR) allows the receptor to activate PPAR target genes. Upon differentiation, this leads to the inappropriate simultaneous activation of both lipid storage and catabolic genes in white adipocytes. However, adding the PPAR N-terminal domain does not suppress PPAR target genes when fused to the PPAR DBD/LBD, suggesting that the repressive effect of this domain only occurs in the context of the cognate receptor.
Also unexpected was our observation that deletion of the N terminus of PPAR is sufficient to confer the ability to promote adipocyte differentiation. This result reveals that the DBD and LBD of PPAR possess the intrinsic ability to promote differentiation and activate the key adipogenic target genes but are prevented from doing so in the context of the holoreceptor by the action of the N-terminal domain. Precisely how the N-terminal region acts to limited target gene expression and modulate adipogenic activity remains to be determined. One possibility is that differential recruitment of positive or negative cofactors to the N-terminal domain could be involved, as has previously been reported for PPAR coactivator-2 (26). However, it is also tempting to speculate that the N-terminal domain interacts with the LBD and AF-2 domains of the receptor to modulate cofactor recruitment by the AF-2. The N-terminal domain of PPARs also contains multiple potential sites for phosphorylation and SUMOylation that may be involved in modulating target gene expression (24, 27, 28, 29). Pinpointing the mechanism by which the individual PPARs diverge to regulate distinct targets should aid in the design of more effective drugs, including tissue and target gene-selective PPAR modulators.
|
MATERIALS AND METHODS
|
|---|
Reagents and Stable Cell Lines
The synthetic ligands GW9544 (PPAR), GW7845 (PPAR), and GW1516 (PPAR) were provided by Jon Collins and Timothy Willson (GlaxoSmithKline, Research Triangle Park, NC). Ligands were dissolved in DMSO before use in cell culture. All experiments were performed using these ligands at 100 nM concentration. Expression constructs containing full-length cDNAs for PPAR1, PPAR2, PPAR, and PPAR (8, 19) were packaged into retrovirus by transient transfection of Phoenix A cells as previously described (8). NIH-3T3 cells were infected at 50% confluence with approximately equal titers of virus. Stable cell lines were selected with 2 µg/ml puromycin for at least 1 wk.
Cell Culture
Stably expressing NIH-3T3 cell lines were maintained in DMEM containing 10% bovine calf serum. For 48 h, gene expression studies cells were switched to 10% fetal bovine serum with dexamethasone (1 µM) and insulin (5 µg/ml) at confluence. 100 nM ligand or DMSO was also added at confluence. For assays of differentiation, cells were treated at confluence with dexamethasone (1 µM), and insulin (5 µg/ml), in DMEM containing 10% fetal bovine serum, for 2 d. Cells were subsequently cultured in DMEM containing 10% fetal bovine serum and insulin. Ligand was first added at confluence and media with fresh ligand was added every 1–2 d. After 5 d, cells were either fixed and stained with Oil Red O as previously described (8) or processed for RNA analysis.
RNA and Protein Analysis
RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA). Sybrgreen and Taqman real-time quantitative PCR assays were performed using an Applied Biosystems (Foster City, CA) 7700 or 7900HT sequence detector as described (30). Results show averages of duplicate experiments normalized to 36B4. PPAR mRNA expression levels in stable cell lines were analyzed by real-time PCR and the relative amounts of each receptor in each cell line was compared using the 2(-
Ct) calculation method (31). Primer and probe sequences are available on request. Western blot analysis was preformed as described on 2x FLAG-tagged PPAR and chimeric constructs (32). EMSAs were performed as described (33) using proteins that were in vitro transcribed/translated using TnT quick-coupled transcription/translation system (Promega, Madison, WI). The sequences of oligonucleotides used for EMSAs were (only one strand shown): aP2 5'-catgcttactGGATCAgAGTTCAcagatc-3', HMGCS2 5'-gat-cttgttctgAGACCTtTGGCCCagtttttc-3', Cyp4B1 5'-gatcaaggatgaGGACCTaTGACCTctttgggg-3', adiponectin 5'-gatcagg-catcagggatagAGGTCAgGTGTCAa-3'.
DNA Microarray Analysis
NIH-3T3 cells stably expressing PPAR, PPAR, or PPAR were cultured in DMEM containing 10% fetal bovine serum, dexamethasone (1 µM), insulin (5 µg/ml), and either vehicle or appropriate ligand (100 nM) for 48 h. Total RNA was isolated using Trizol reagent and further purified with a QIAGEN (Valencia, CA) RNeasy total RNA isolation kit. Duplicate samples of RNA (each derived from a pool of two or more identically treated cultures) were processed by the UCLA microarray facility for hybridization and visualization on murine 430A Affymetrics chips. The results of the microarrays were analyzed with Genespring and GeneChip Analysis Suite software (Affymetrix). Each sample was analyzed in duplicate and only genes shown to have statistically significant differences in gene expression are included. The resulting list of target genes for each receptor, or combination of receptors, thus contains less than 5% false positives based on statistical analysis done using the ANOVA test and multiple testing correction using Benjamini and Hochberg false discovery rate.
PPAR Chimeric and Deletion Constructs
A PCR-based method using bridge primers, SalI overhangs, and the original PPAR-pBabe plasmids as templates was used to create chimeric constructs. The AF-1 transcriptional activation domain of PPAR (coding for amino acids 1–138) was fused to the DNA binding, ligand binding, dimerization, and AF2 transcriptional activation domain of PPAR (amino acids 73–440). Similarly, the N-terminal AF-1 domain of PPAR (amino acids 1–72) was fused to PPAR from the DBD onward to the C terminus (amino acids 139–505). In addition, constructs swapping both the AF-1 domain and adjacent DBD (amino acids 1–137 for PPAR, 1–203 for PPAR) were created. N-terminal deletions (amino acids 1–127 for PPAR2 and 1–43 for PPAR) were also generated. PCR products were then TA-cloned (Invitrogen), sequenced, and cloned into the SalI site of pBabe. The pBabe chimeric constructs were used to create NIH-3T3 cell lines stably expressing the chimeric receptors as described above for the wild-type receptors.
|
FOOTNOTES
|
|---|
Disclosure Statement: S.H. has nothing to declare. P.T. consults for consults for Asahi Kasei, Sumitomo and Bristol-Myers Squibb and received lecture fees from Bristol-Myers Squibb.
First Published Online March 23, 2006
Abbreviations: ADRP, Adipose differentiation-related protein; AF, activation function; CYP4B1, cytochrome p450, superfamily IV B, polypeptide 1; DBD, DNA binding domain; DMSO, dimethyl sulfoxide; FZD7, frizzled 7; HMGCS2, 3-hydroxy-3-methylglutaryl-CoA synthase 2; HRASLS3, HRAS like supressor 3; LBD, ligand binding domain; PPAR, peroxisome proliferator-activated receptor; PPRE, PPAR response elements; RXR, retinoid X receptor.
Received for publication January 13, 2006.
Accepted for publication March 15, 2006.
|
REFERENCES
|
|---|
- Desvergne B, Wahli W 1999 Peroxisome proliferator-activated receptors: nuclear control of metabolism. Endocr Rev 20:649–688[Abstract/Free Full Text]
- Rosen ED, Spiegelman BM 2000 Molecular regulation of adipogenesis. Annu Rev Cell Dev Biol 16:145–171[CrossRef][Medline]
- Lehrke M, Lazar MA 2005 The many faces of PPAR. Cell 123:993–999[CrossRef][Medline]
- Isseman I, Green S 1990 Activation of a member of the steriod hormone receptor superfamily by peroxisome proliferators. Nature 347:645–649[CrossRef][Medline]
- Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Wilson TM, Kliewer SA 1995 An antidiabetic thiazolidinedione is a high affinity ligand for perosixome proliferator-activated receptor (PPAR). J Biol Chem 270:12953–12956[Abstract/Free Full Text]
- Forman BM, Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans RM 1995 15-Deoxy- 12,14-prostaglandin J2 is a ligand for the adipocyte determination factor PPAR . Cell 83:803–812[CrossRef][Medline]
- Castrillo A, Tontonoz P 2004 Nuclear receptors in macrophage biology: at the crossroads of lipid metabolism and inflammation. Annu Rev Cell Dev Biol 20:455–480[CrossRef][Medline]
- Tontonoz P, Hu E, Spiegelman BM 1994 Stimulation of adipogenesis in fibroblasts by PPARg2, a lipid-activated transcription factor. Cell 79:1147–1156[CrossRef][Medline]
- Tontonoz P, Hu E, Graves RA, Budavari AI, Spiegelman BM 1994 mPPAR 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev 8:1224–1234[Abstract/Free Full Text]
- Rosen ED, Sarraf P, Troy AE, Bradwin G, Moore K, Milstone DS, Spiegelman BM, Mortensen RM 1999 PPAR is required for the differentiation of adipose tissue in vivo and in vitro. Mol Cell 4:611–617[CrossRef][Medline]
- Barak Y, Nelson MC, Ong ES, Jones YZ, Ruiz-Lozano P, Chien KR, Koder A, Evans RM 1999 PPAR is required for placental, cardiac, and adipose tissue development. Mol Cell 4:585–595[CrossRef][Medline]
- Dreyer C, Krey G, Keller H, Givel F, Helftenbein G, Wahli W 1992 Control of the peroxisomal b-oxidation pathway by a novel family of nuclear hormone receptors. Cell 68:879–887[CrossRef][Medline]
- Wang YX, Lee CH, Tiep S, Yu RT, Ham J, Kang H, Evans RM 2003 Peroxisome-proliferator-activated receptor activates fat metabolism to prevent obesity. Cell 113:159–170[CrossRef][Medline]
- Lee CH, Olson P, Evans RM 2003 Minireview: lipid metabolism, metabolic diseases, and peroxisome proliferator-activated receptors. Endocrinology 144:2201–2207[Abstract/Free Full Text]
- Oliver Jr WR, Shenk JL, Snaith MR, Russell CS, Plunket KD, Bodkin NL, Lewis MC, Winegar DA, Sznaidman ML, Lambert MH, Xu HE, Sternbach DD, Kliewer SA, Hansen BC, Willson TM 2001 A selective peroxisome proliferator-activated receptor agonist promotes reverse cholesterol transport. Proc Natl Acad Sci USA 98:5306–5311[Abstract/Free Full Text]
- Leibowitz MD, Fievet C, Hennuyer N, Peinado-Onsurbe J, Duez H, Bergera J, Cullinan CA, Sparrow CP, Baffic J, Berger GD, Santini C, Marquis RW, Tolman RL, Smith RG, Moller DE, Auwerx J 2000 Activation of PPAR alters lipid metabolism in db/db mice. FEBS Lett 473:333–336[CrossRef][Medline]
- Kliewer SA, Umesono K, Noonan D, Heyman R, Evans RM 1992 Convergence of 9-cis retinoic acid and peroxisome proliferator signalling pathways through heterodimer formation of their receptors. Nature 358:771–774[CrossRef][Medline]
- Braissant O, Foufelle F, Scotto C, Dauca M, Wahli W 1996 Differential expression of peroxisome proliferator-activated receptors (PPARs): tissue distribution of PPAR-, -ß, and - in the adult rat. Endocrinology 137:354–366[Abstract]
- Brun RP, Tontonoz P, Forman BM, Ellis R, Chen J, Evans RM, Spiegelman BM 1996 Differential activation of adipogenesis by multiple PPAR isoforms. Genes Dev 10:974–984[Abstract/Free Full Text]
- Desvergne B, A IJ, Devchand PR, Wahli W 1998 The peroxisome proliferator-activated receptors at the cross-road of diet and hormonal signalling. J Steroid Biochem Mol Biol 65:65–74[Medline]
- Guan HP, Ishizuka T, Chui PC, Lehrke M, Lazar MA 2005 Corepressors selectively control the transcriptional activity of PPAR in adipocytes. Genes Dev 19:453–461[Abstract/Free Full Text]
- Guan HP, Li Y, Jensen MV, Newgard CB, Steppan CM, Lazar MA 2002 A futile metabolic cycle activated in adipocytes by antidiabetic agents. Nat Med 8:1122–1128[CrossRef][Medline]
- Way JM, Harrington WW, Brown KK, Gottschalk WK, Sundseth SS, Mansfield TA, Ramachandran RK, Willson TM, Kliewer SA 2001 Comprehensive messenger ribonucleic acid profiling reveals that peroxisome proliferator-activated receptor activation has coordinate effects on gene expression in multiple insulin-sensitive tissues. Endocrinology 142:1269–1277[Abstract/Free Full Text]
- Hu E, Kim JB, Sarraf P, Spiegelman BM 1996 Inhibition of adipogenesis through MAP kinase-mediated phosphorylation of PPAR. Science 274:2100–2103[Abstract/Free Full Text]
- Zhang B, Berger J, Zhou G, Elbrecht A, Biswas S, White-Carrington S, Szalkowski D, Moller DE 1996 Insulin- and mitogen-activated protein kinase-mediated phosphorylation and activation of peroxisome proliferator-activated receptor gamma. J Biol Chem 271:31771–1774[Abstract/Free Full Text]
- Castillo G, Brun RP, Rosenfield JK, Hauser S, Park CW, Troy AE, Wright ME, Spiegelman BM 1999 An adipogenic cofactor bound by the differentiation domain of PPAR. EMBO J 18:3676–3687[CrossRef][Medline]
- Ohshima T, Koga H, Shimotohno K 2004 Transcriptional activity of peroxisome proliferator-activated receptor is modulated by SUMO-1 modification. J Biol Chem 279:29551–29557[Abstract/Free Full Text]
- Yamashita D, Yamaguchi T, Shimizu M, Nakata N, Hirose F, Osumi T 2004 The transactivating function of peroxisome proliferator-activated receptor is negatively regulated by SUMO conjugation in the amino-terminal domain. Genes Cells 9:1017–1029[Abstract/Free Full Text]
- Zhang B, Berger J, Hu E, Szalkowski D, White-Carrington S, Spiegelman BM, Moller DE 1996 Negative regulation of peroxisome proliferator-activated receptor- gene expression contributes to the antiadipogenic effects of tumor necrosis factor-. Mol Endocrinol 10:1457–1466[Abstract]
- Laffitte BA, Joseph SB, Walczak R, Pei L, Wilpitz DC, Collins JL, Tontonoz P 2001 Autoregulation of the human liver X receptor promoter. Mol Cell Biol 21:7558–7568[Abstract/Free Full Text]
- Bodin L, Beaune PH, Loriot MA 2005 Determination of cytochrome P450 2D6 (CYP2D6) gene copy number by real-time quantitative PCR. J Biomed Biotechnol 2005:248–253[Medline]
- Castrillo A, Diaz-Guerra MJ, Hortelano S, Martin-Sanz P, Bosca L 2000 Inhibition of I
B kinase and IB phosphorylation by 15-deoxy-(12,14)-prostaglandin J(2) in activated murine macrophages. Mol Cell Biol 20:1692–1698[Abstract/Free Full Text] - Castrillo A, Joseph SB, Marathe C, Mangelsdorf DJ, Tontonoz P 2003 Liver X receptor-dependent repression of matrix metalloproteinase-9 expression in macrophages. J Biol Chem 278:10443–10449[Abstract/Free Full Text]
NURSA Molecule Pages Link:
- Nuclear Receptors:
PPARα
|
PPARδ
|
PPARγ
|
RXRα
- Ligands:
GW409544
This article has been cited by other articles:
|
|
|
|
 
S. Hummasti, C. Hong, S. J. Bensinger, and P. Tontonoz
HRASLS3 is a PPAR{gamma}-selective target gene that promotes adipocyte differentiation
J. Lipid Res.,
December 1, 2008;
49(12):
2535 - 2544.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
