Cross species comparison of C/EBPa and PPARg profiles in mouse and human adipocytes reveals interdependent retention of binding sites

Background: The transcription factors peroxisome proliferator activated receptor g (PPARg) and CCAAT/enhancer binding protein a (C/EBPa) are key transcriptional regulators of adipocyte differentiation and function. We and others have previously shown that binding sites of these two transcription factors show a high degree of overlap and are associated with the majority of genes upregulated during differentiation of murine 3T3-L1 adipocytes. Results: Here we have mapped all binding sites of C/EBPa and PPARg in human SGBS adipocytes and compared these with the genome-wide profiles from mouse adipocytes to systematically investigate what biological features correlate with retention of sites in orthologous regions between mouse and human. Despite a limited interspecies retention of binding sites, several biological features make sites more likely to be retained. First, co-binding of PPARg and C/EBPa in mouse is the most powerful predictor of retention of the corresponding binding sites in human. Second, vicinity to genes highly upregulated during adipogenesis significantly increases retention. Third, the presence of C/EBPa consensus sites correlate with retention of both factors, indicating that C/EBPa facilitates recruitment of PPARg. Fourth, retention correlates with overall sequence conservation within the binding regions independent of C/EBPa and PPARg sequence patterns, indicating that other transcription factors work cooperatively with these two key transcription factors. Conclusions: This study provides a comprehensive and systematic analysis of what biological features impact on retention of binding sites between human and mouse. Specifically, we show that the binding of C/EBPa and PPARg in adipocytes have evolved in a highly interdependent manner, indicating a significant cooperativity between these two transcription factors. Background The adipose tissue plays a central role in maintaining whole body lipid and glucose homeostasis as well as insulin sensitivity [1,2]. Adipocytes are derived from fibroblastic precursors in the adipose tissue through a tightly regulated differentiation process. The molecular basis for the regulation of adipogenesis has been studied extensively in vitro using a variety of preadipocyte cell culture models. In particular, studies of the 3T3-L1 cell line derived from mouse embryo fibroblasts [3] have been valuable for gaining insight into the ordered cascade of molecular events required for adipogenesis [4-6]. More recently, human cell culture models have become available including the SGBS cell line derived from preadipocytes from a patient with Simpson-Golabi-Behmel syndrome [7]. The transcription factor peroxisome proliferator activated receptor (PPARg) is a key regulator of adipogenesis, required for in vitro as well as in vivo differentiation of adipocytes [8,9]. In addition to PPARg, a number of other important transcriptional regulators of adipocyte differentiation have been identified [6], including members of the CCAAT/enhancer binding protein (C/EBP) family [10-12]. C/EBPa is induced late in adipocyte differentiation and is known to cooperate with PPARg in induction of at least a subset of adipocyte-specific genes. In addition, these two * Correspondence: [email protected]; [email protected] † Contributed equally Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark The Bioinformatics Centre, Department of Biology and Biomedical Research and Innovation Centre, Copenhagen University, Ole Maaløs Vej 5, DK-2200, Copenhagen N, Denmark Full list of author information is available at the end of the article Schmidt et al. BMC Genomics 2011, 12:152 http://www.biomedcentral.com/1471-2164/12/152 © 2011 Schmidt et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. factors induce the expression of each other [13,14]. Two other members of the C/EBP family, C/EBPb and -δ, are directly involved in the transcriptional induction of PPARg and C/EBPa [6]. We recently used chromatin immunoprecipitation (ChIP) combined with deep sequencing (ChIP-seq) to generate genome-wide maps of the binding sites of PPARg and its heterodimerization partner retinoid × receptor (RXR) during differentiation of 3T3-L1 adipocytes [15]. In addition, we profiled RNA polymerase II (RNAPII) occupancy to measure active transcription at different time points during differentiation. This study revealed that PPARg:RXR binding was highly enriched in the vicinity of genes upregulated during adipogenesis. In fact, the majority (75%) of all highly up-regulated genes have PPARg:RXR binding in the immediate vicinity [15]. Similarly, Lazar and colleagues [16] and others [17,18] used ChIP in combination with hybridization to genomic tiling microarrays (ChIP-chip) or cloning followed by sequencing (ChIP-PET) to map PPARg binding sites in 3T3-L1 adipocytes (reviewed in [19]). Intriguingly, these studies have revealed that the C/EBPa consensus site is highly over-represented under the binding regions of PPARg. Lazar and colleagues profiled C/EBPa binding sites in mature 3T3-L1 adipocytes and found a remarkable overlap between C/EBPa and PPARg binding (> 60% of all PPARg sites) on a genome-wide scale [16]. Importantly, 60% of the genes induced during adipogenesis have both C/EBPa and PPARg binding sites within 50 kb of a transcription start site (TSS), and knockdown studies indicated that both C/EBPa and PPARg are required for robust gene expression of a few selected adipocyte specific target genes. Cumulatively, these results indicate that both PPARg and C/EBPa are directly involved in the activation of the majority of adipocyte-specific genes and that they cooperate through binding to adjacent sites on DNA. Genome-wide profiling has also made it possible to study the evolution of gene regulation by mapping the sites for the same transcription factors in different species. This is typically done by aligning genomes of the two species and tabulating the number of detected sites in one species that are in the corresponding region in the other species. This means that a genomic region might be highly conserved in terms of nucleotides but may or may not bind the transcription factor in question in both species. To distinguish between genome sequence conservation and transcription factor binding site conservation, we will use the word “retention” to describe binding sites that are present in both species in the corresponding genomic region. Notably, whereas the functional gene targets of a particular transcription factor are generally well conserved between species, it has been shown that the majority of binding sites for all transcription factors investigated to date are species-specific [20-25], reviewed in [26]. This is surprising given that sequence conservation has been successfully used to enhance regulatory site prediction in proximal promoters (phylogenetic foot printing) [27,28]; however, this might reflect that these older studies were focused on limited sets of sites often located in tissuespecific promoters, while genome-wide methods, such as ChIP-seq, are independent of previous annotations. Consistent with the species specificity of transcription factor binding sites, Rosen and colleagues recently compared PPARg binding in 3T3-L1 adipocytes and in vitro differentiated primary human adipose stromal cells (hASC), and found that only 21.3% of the murine binding sites were retained in human adipocytes. By contrast, the association of PPARg with adipocyte gene regulation appeared to be better retained than the specific binding sites, and the overall gene expression profiles were well conserved [21]. Rosen and colleagues also showed that genes associated with a conserved PPARg binding site are more likely to be upregulated during adipogenesis than genes associated with a species specific site [21], indicating that retention of PPARg binding is increased near upregulated genes. While previous reports agree that retention of transcription factor binding sites is limited, systematic analysis of the biological features determining whether a site is retained or not has not been performed. In particular, the interdependence between retention of the transcription factor binding sites of two transcription factors has not been investigated. Here, we used ChIP-seq to generate genome-wide binding profiles of C/EBPa and PPARg in human SGBS adipocytes, compared these to previously published profiles in mouse 3T3-L1 adipocytes [15,16], and systematically studied what features affect whether a binding site is retained between species or not. We find that PPARg binding sites have higher retention near genes upregulated during adipogenesis, and that regions bound by both factors are even more likely to be retained. Interestingly, PPARg binding site retention in these co-bound regions is increased by the presence of a C/EBPa consensus site, suggesting that C/EBPa may facilitate PPARg binding to DNA. At the same time, and independent of C/EBPa and PPARg sequence patterns, sequence conservation in the larger region surrounding the actual binding sites has a positive impact on retention of both C/EBPa and PPARg binding sites, indicating that other DNA sequence patterns also affect binding of these two factors to DNA. Results Genome-wide mapping of C/EBPa and PPARg binding in human SGBS adipocytes To compare C/EBPa and PPARg binding in mouse and human adipocytes, we used ChIP-seq to generate Schmidt et al. BMC Genomics 2011, 12:152 http://www.biomedcentral.com/1471-2164/12/152 Page 2 of 16