Within the Zfp423 gene itself showed reproducible occupancy by Zfp423 in ChIP-PCR and ChIP-seq assays. The stronger of these sites, in Zfp423 intron 5, also showed enhancer activity in heterologous classical promoter-reporter assays in P19 cells. Surprisingly, Zfp423 appears to act as a negative 4 IBP site regulator atZfp423 Binds Autoregulatory Sitesthe stronger of the two sites, suggesting a negative feedback cycle that may be conditional on signaling and cell state.Results Conserved Zfp423-complex Binding Motifs are Enriched at Zfp423 and Ebf GenesTo identify candidate target genes for Zfp423, we first looked for consensus binding sites in regions conserved among vertebrate genomes (Figure 1). Using the web-based SynoR software tool, which uses a matrix representation to account for degeneracy in binding sites [16], we separately examined paired or clustered sites for Zfp423 [11] and its best-characterized binding partners, Ebf (including Olf1, represented by a distinct sequence matrix [17]), SMAD, and Retinoic acid receptor in sequences conserved across vertebrate species pairs (human-chick, mouse-chick, mouse-frog), across a range of parameters for number (1?) of sites and maximum distance (100?00 bp) between sites for at least two component factors. Multiple binding matrices were used for Ebf (EBF_Q6 and OLF1_01) and SMAD (SMAD_Q6, SMAD_Q6_01, and SMAD4_Q6) Title Loaded From File family members. This analysis resulted in a surprisingly small number of sites genome wide; distributions of such sites for 100 bp windows with 2 sites or 150 bp windows for three sites are tabulated (Figure 1A). We identified 60 conserved non-coding sites containing a Zfp423 consensus site within 100 bp of either a consensus motif for one of its known binding partners or a second Zfp423 site, with syntenic site predictions in human, mouse and chicken. Surprisingly, four of these 60 robustly predicted clusters occur in the Zfp423 and Ebf3 genes (Figure 1B,C). Broadening the criteria to allow up to 200 bp between sites and to allow Ebf-only or SMAD-only clusters finds three additional sites in or adjacent to Ebf1 (Figure 1D). This enrichment of clustered sites for known interacting factors in genes encoding those factors represents a dramatic enrichment above genome-wide expectation and led us to test whether these sites might 23148522 be functional.uniqueness of the amplified sequence and gel electrophoresis to confirm predicted size at endpoint. Similar quantitative results were obtained with two distinct primer sets for Ebf3. Western blot analysis confirmed expression of ZNF423/Zfp423 protein in IMR32 and P19 cells, relative to b-actin and GAPDH loading controls (Figure 2E). Zfp423 frequently appeared as a doublet under gel conditions that optimized its detection (see Methods); that both bands represented legitimate Zfp423 was supported both by their recognition with independent antibodies and by loss of both bands in extracts from either Zfp423-mutant tissues or cells treated with Zfp423-directed shRNA (see below). Based on expression of ZNF423/Zfp423 and at least one EBF/Ebf member, IMR32 and P19 cells were selected for further experiments.Zfp423 Occupies Sites Zfp423 Introns 3 and 5 in Mouse and Human CellsTo test whether predicted sites are occupied in cells with relatively high levels of the indicated factors, we performed a series of chromatin immunoprecipitation (ChIP) experiments (Figure 3). Semi-quantitative PCR after ChIP detected ZNF423 binding above background in IMR32 cells at the.Within the Zfp423 gene itself showed reproducible occupancy by Zfp423 in ChIP-PCR and ChIP-seq assays. The stronger of these sites, in Zfp423 intron 5, also showed enhancer activity in heterologous classical promoter-reporter assays in P19 cells. Surprisingly, Zfp423 appears to act as a negative regulator atZfp423 Binds Autoregulatory Sitesthe stronger of the two sites, suggesting a negative feedback cycle that may be conditional on signaling and cell state.Results Conserved Zfp423-complex Binding Motifs are Enriched at Zfp423 and Ebf GenesTo identify candidate target genes for Zfp423, we first looked for consensus binding sites in regions conserved among vertebrate genomes (Figure 1). Using the web-based SynoR software tool, which uses a matrix representation to account for degeneracy in binding sites [16], we separately examined paired or clustered sites for Zfp423 [11] and its best-characterized binding partners, Ebf (including Olf1, represented by a distinct sequence matrix [17]), SMAD, and Retinoic acid receptor in sequences conserved across vertebrate species pairs (human-chick, mouse-chick, mouse-frog), across a range of parameters for number (1?) of sites and maximum distance (100?00 bp) between sites for at least two component factors. Multiple binding matrices were used for Ebf (EBF_Q6 and OLF1_01) and SMAD (SMAD_Q6, SMAD_Q6_01, and SMAD4_Q6) family members. This analysis resulted in a surprisingly small number of sites genome wide; distributions of such sites for 100 bp windows with 2 sites or 150 bp windows for three sites are tabulated (Figure 1A). We identified 60 conserved non-coding sites containing a Zfp423 consensus site within 100 bp of either a consensus motif for one of its known binding partners or a second Zfp423 site, with syntenic site predictions in human, mouse and chicken. Surprisingly, four of these 60 robustly predicted clusters occur in the Zfp423 and Ebf3 genes (Figure 1B,C). Broadening the criteria to allow up to 200 bp between sites and to allow Ebf-only or SMAD-only clusters finds three additional sites in or adjacent to Ebf1 (Figure 1D). This enrichment of clustered sites for known interacting factors in genes encoding those factors represents a dramatic enrichment above genome-wide expectation and led us to test whether these sites might 23148522 be functional.uniqueness of the amplified sequence and gel electrophoresis to confirm predicted size at endpoint. Similar quantitative results were obtained with two distinct primer sets for Ebf3. Western blot analysis confirmed expression of ZNF423/Zfp423 protein in IMR32 and P19 cells, relative to b-actin and GAPDH loading controls (Figure 2E). Zfp423 frequently appeared as a doublet under gel conditions that optimized its detection (see Methods); that both bands represented legitimate Zfp423 was supported both by their recognition with independent antibodies and by loss of both bands in extracts from either Zfp423-mutant tissues or cells treated with Zfp423-directed shRNA (see below). Based on expression of ZNF423/Zfp423 and at least one EBF/Ebf member, IMR32 and P19 cells were selected for further experiments.Zfp423 Occupies Sites Zfp423 Introns 3 and 5 in Mouse and Human CellsTo test whether predicted sites are occupied in cells with relatively high levels of the indicated factors, we performed a series of chromatin immunoprecipitation (ChIP) experiments (Figure 3). Semi-quantitative PCR after ChIP detected ZNF423 binding above background in IMR32 cells at the.