 |  |
| ||||
Copyright © The Author 2005. Published by Oxford University Press. All rights reserved Characterization of cis- and trans-acting elements in the imprinted human SNURF-SNRPN locus *To whom correspondence should be addressed. Tel: +1 352 392 6472; Fax: +1 352 392 2953; Email: tpyang/at/ufl.edu Received April 21, 2005; Revised July 27, 2005; Accepted August 5, 2005. The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact journals.permissions/at/oupjournals.org  This article has been cited by other articles in PMC. | |||||||||
Abstract The imprinted SNRPN locus is a complex transcriptional unit that encodes the SNURF and SmN polypeptides as well as multiple non-coding RNAs. SNRPN is located within the Prader-Willi and Angelman syndrome (PWS/AS) region that contains multiple imprinted genes, which are coordinately regulated by a bipartite imprinting center (IC). The SNRPN 5′ region co-localizes with the PWS-IC and contains two DNase I hypersensitive sites, DHS1 at the SNRPN promoter, and DHS2 within intron 1, exclusively on the paternally inherited chromosome. We have examined DHS1 and DHS2 to identify cis- and trans-acting regulatory elements within the endogenous SNRPN 5′ region. Analysis of DHS1 by in vivo footprinting and chromatin immunoprecipitation identified allele-specific interaction with multiple regulatory proteins, including NRF-1, which regulates genes involved in mitochondrial and metabolic functions. DHS2 acted as an enhancer of the SNRPN promoter and contained a highly conserved region that showed allele-specific interaction with unphosphorylated RNA polymerase II, YY1, Sp1 and NRF-1, further suggesting a key role for NRF-1 in regulation of the SNRPN locus. We propose that one or more of the regulatory elements identified in this study may also contribute to PWS-IC function. | |||||||||
INTRODUCTION SNURF-SNRPN (hereafter termed SNRPN) is a bicistronic imprinted gene on human chromosome 15 that encodes two polypeptides, the SmN splicing factor involved in RNA processing (1), and the SNURF (SNRPN upstream reading frame) polypeptide of unknown function (2). SNRPN also encodes a long (~460 kb) alternatively spliced RNA transcript that contains several families of snoRNAs (3) and extends downstream to partially overlap the UBE3A gene in the anti-sense orientation. The SNRPN promoter is associated with a CpG island that is hypermethylated on the maternally inherited allele and hypomethylated on the paternally inherited allele (4). Two alternative upstream promoters and alternatively spliced non-coding exons expressed at low levels add to the complexity of the locus (5). The gene is transcribed exclusively from the paternally inherited chromosome and shows highest levels of expression in the brain and heart (2). Furthermore, SNRPN is located within an imprinted gene cluster in chromosome 15 q11–q13 that is associated with the Prader-Willi syndrome (PWS) and Angelman syndrome (AS). PWS and AS are two clinically distinct neurogenetic disorders linked to a single imprinted domain on chromosome 15 containing at least eight genes distributed across ~2 Mb [reviewed in (6)]; the similarly imprinted syntenic region in the mouse is located on chromosome 7C (6). PWS arises from loss of function of genes in this region that are expressed exclusively from the paternal chromosome, while AS arises from loss of expression or mutation of the maternally expressed UBE3A gene. Multiple genetic mechanisms lead to the allele-specific loss of gene expression in AS and PWS, including deletions of the entire imprinted region, uniparental disomy (UPD), and microdeletions that encompass the 5′ region of the SNRPN gene and/or a region upstream of SNRPN. High-resolution mapping of these microdeletions has led to the delineation of a bipartite imprinting control region or imprinting center (IC). All of the microdeletions associated with PWS share a 4.3 kb deleted region termed the PWS smallest region of deletion overlap (PWS-SRO), which includes the SNRPN promoter region, first exon, and part of the first intron (7). Similarly, the microdeletions associated with AS share a 0.8 kb AS-SRO located ~35 kb upstream from exon 1 of SNRPN (8). Thus, the IC is composed of two distinct functional components, the PWS-IC (including but not necessarily limited to the PWS-SRO) that appears to be required for maintenance of the paternal epigenotype in somatic cells (9,10), and the AS-IC (including the AS-SRO), which appears to be required for establishment of the maternal epigenotype during oogenesis (11). Currently, the mechanisms of PWS-IC and AS-IC function in establishing and/or maintaining imprinted gene expression across the domain are not well understood. The co-localization of the PWS-SRO with the SNRPN promoter suggests that transcription factor binding to the promoter and subsequent transcriptional activation of the SNRPN locus may be integral to PWS-IC function (12). Several studies have identified cis-acting elements in the SNRPN promoter region that affect promoter function in transient expression assays and regulate expression or imprinting of mouse transgenes. However, the corresponding trans-acting factors have not been identified and no specific transcription factors have yet been shown to bind to the endogenous SNRPN locus. SNRPN promoter function was first reported to be contained within an interval between positions −207 and +53 by using transient expression assays of reporter constructs that included an exogenous SV40 enhancer (13). Employing the same strategy (and the SV40 enhancer), the minimal SNRPN promoter was subsequently shown to be located between nucleotides −71 and +51, a region that contains a 7 bp element (SBE) between nucleotides −57 and −51 and an element around position +17 that act as positive regulators of transcription (14). In addition, a repressor sequence was mapped downstream of the minimal promoter in the 3′ region of exon 1 (14). The SBE coincides with one of the six sequences that are phylogenetically conserved between the 5′ flanking region of the human and mouse SNRPN genes (7). The SBE has also been identified as a functional promoter element in the mouse Snrpn gene (15). Subsequent analysis of the mouse Snrpn promoter in a transgene construct containing the human AS-SRO and the minimal mouse Snrpn promoter expressing a reporter gene identified several additional DNA sequences that appear to be involved in imprinted expression of the transgene (16). These included two de novo methylation signals (DNS), an allele discrimination signal (ADS) and two separate signals responsible for the maintenance of the paternal imprint (MPI1 and MPI2). The mechanisms and the associated trans-acting factors that mediate the function of these cis-acting elements remain unknown. Despite the fact that transgenic mice containing constructs with the human AS-SRO and the human (17) or the mouse (18) SNRPN promoters assume the correct imprinted state of the transgene, deletion of the endogenous mouse Snrpn promoter has no apparent effect on PWS-IC function (19), suggesting that sequences outside of the promoter may be sufficient to maintain the function of the PWS-IC in the mouse (16). In the current study, we have used multiple strategies to identify cis-acting elements within the endogenous human SNRPN 5′ region and to determine the corresponding trans-acting regulatory factors that interact in an allele-specific manner with the promoter. We also have identified and analyzed cis-acting and allele-specific trans-acting elements associated with an enhancer located within the first intron of SNRPN just downstream of the PWS-SRO. The potential role of these regulatory elements in PWS-IC function is discussed. | |||||||||
MATERIALS AND METHODS Cell culture EBV-transformed lymphoblasts derived from PWS and AS patients were grown in RPMI medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Human SK-N-SH neuroblastoma cells were obtained from the American Type Culture Collection and grown in E-MEM supplemented with 10% FBS and 1% penicillin–streptomycin. Cells were grown at 37°C in 5% CO2.DNase I treatment of permeabilized cells for mapping of hypersensitive sites and Southern blotting Cell permeabilization and DNase I treatment was performed as described previously (20). Genomic DNA from DNase I treated cells was digested with BamHI and size-fractionated in a 0.8% agarose gel, transferred and hybridized as described previously (20). The 2.26 kb hybridization probe was isolated from plasmid pPH-B8 (21) by digestion with BamHI and EcoRI and labeled by random priming.Vector design Constructs for transient expression assays were generated from the pGL3-Basic vector (Promega). Details on vector construction are provided in Supplemental Material.Transient transfection and luciferase reporter assays SK-N-SH cells were transfected with firefly luciferase expression constructs using SuperFect (Qiagen) according to the manufacturer's specifications (see Supplementary Material). Cells were co-transfected with pRL-TK (Promega) which contains the Renilla luciferase gene. Cells were lysed 24 h post-transfection and firefly and Renilla luciferase activities were measured (Dual-Luciferase® Reporter Assay System, Promega) in a Sirius Luminometer V2.2. (Berthold Detection Systems). Firefly luciferase activity was normalized to Renilla luciferase activity. For each construct, the average and standard error of the mean [standard deviation/square root (n), where n is the number of independent experiments] were calculated.Site-directed mutagenesis Sequence-specific mutations in reporter constructs were performed using the QuikChange® XL Site-Directed Mutagenesis Kit (Stratagene) following the manufacturer's instructions. Mutated sequences were designed to include a new restriction site, and mutant clones were identified by restriction digestion and verified by sequencing. Supplementary Table S1 shows the wild-type and mutant sequences. Mutant sequences from constructs (c–g) in Figure 2G were excised with XbaI and cloned into the XbaI site of construct (c) in Figure 4B to generate the constructs (d–h) in Figure 4B.
Dimethyl sulfate (DMS) treatment of cells and naked DNA for in vivo footprinting DMS treatment of ~2 × 107 lymphoblast cells and DMS treatment of purified DNA were performed as described by Hornstra and Yang (22).Ligation-mediated PCR (LMPCR) LMPCR was performed essentially as described previously (20). The sequence of LMPCR primers and cycling conditions for the PCR are provided in Supplementary Table S2.LMPCR DNA sequencing ladders Ladders were generated from genomic DNA as described previously (22) and were used to determine the exact position of each guanine residue in autoradiograms of LMPCR-amplified DMS-treated samples.Electrophoresis, transfer and hybridization of sequencing gels LMPCR-amplified products were analyzed by size-fractionation in a DNA sequencing gel, followed by electrotransfer, hybridization and autoradiography essentially as described previously (22). Radiolabeled probes were generated as described in Supplementary Material.Chromatin immunoprecipitation (ChIP) ChIP was performed essentially as described by Leach et al. (23) on human lymphoblasts. A total of 107 cells were used in each immunoprecipitation (IP) reaction with antibodies against YY1 (Santa Cruz Biotechnology sc-1703; 5 μg), NRF-1 (kindly provided by Richard C. Scarpulla; 4 μl), Sp1 (Santa Cruz Biotechnology sc-59; 10–12.5 μg), CTCF (Upstate 06-917; 5–20 μg), unphosphorylated RNA polymerase II (Covance MMS-126R, 40–60 μg), H3 dimethyl-K4 (Abcam ab7766-50; 2 μg), acetylated H4 (Upstate 06-866; 1 μl) and H3 acetyl-K9 (Upstate 06-942; 5 μg). To account for non-specific binding, each experiment included a mock IP reaction to which no antibody was added; these reactions routinely yielded no PCR products. Immunoprecipitated DNA was analyzed by PCR using PCR conditions and primer sequences shown in Supplementary Table S3. The linearity of the amplification was verified by diluting the input DNA (i.e. the fraction of the unbound chromatin in the mock IP) by 1:2, 1:4 and 1:8. Genomic regions outside of the PWS-AS domain that are known to be associated with specific factors were analyzed as a positive control for the IP reaction with antibodies against those factors (Supplementary Figure S3A). PCR products were size-fractionated in 5% TBE (89 mM Tris, pH 8.3, 89 mM boric acid and 2 mM EDTA) polyacrylamide gels and gel bands were visualized and quantified by SyBr-green staining in a fluorescence scanner (Storm 860, Molecular Dynamics). The fraction bound was calculated as percentage of input DNA; average and standard error of the mean across experiments are shown. | |||||||||
RESULTS The location of DNase I hypersensitive (DH) sites in chromatin within and surrounding the SNRPN promoter were mapped by Southern blot analysis (Figure 1) and confirmed previous reports by Schweizer et al. (24) and Ohta et al. (7). Our studies identified two strong hypersensitive sites specific to the paternal chromosome: DHS1—which is located in the promoter region of the SNPRN gene and DHS2—which is located in the first intron roughly 1.5 kb downstream of the transcription initiation site. DHS2 is located just outside and downstream of the PWS-SRO (7). Because DH sites are commonly associated with cis-acting regulatory elements in eukaryotic chromatin, and because these DH sites were mapped within or immediately adjacent to the PWS-SRO, we used various strategies to identify and characterize both cis- and trans-acting regulatory elements within DHS1 and DHS2 of the endogenous SNRPN locus.
In vivo footprint analysis of DHS1 To identify cis-acting regulatory elements associated with DHS1 within intact cells, we performed DMS in vivo footprint analysis of the SNRPN promoter region using LMPCR (22). As shown in Figure 2A, the region analyzed included a series of potential transcription factor binding sites identified by the TRANSFAC database, as well as six phylogenetically conserved sequences within the SNRPN promoter region (7). The paternal and maternal SNRPN alleles were assayed independently using cultured lymphoblasts from AS and PWS patients, respectively, carrying either UPD or deletion (Del) of the AS/PWS region. Figure 2A shows the relative position and region analyzed with each LMPCR primer set on the upper and lower strands of the endogenous SNRPN promoter region from position −490 to position +150.Each region and strand of interest was analyzed in multiple DNA sequencing gels from at least two separate DMS treatments of cells, and footprints were further confirmed in two different cell lines (UPD and Del) for each parental allele. Although some of the footprints were subtle, all were highly reproducible (confirmed in 80–100% of the autoradiograms containing each site). Furthermore, transient expression assays (see below) confirmed four of the six footprinted sites as bona fide cis-acting regulatory elements. Weak footprints were likely to be due to the fact that these in vivo footprinting assays were performed in lymphoblast cells, which express relatively low levels of SNRPN mRNA (21) and, therefore, unlikely to have a highly active SNRPN promoter maximally occupied by transcription factors. In vivo footprints were detected as bands of decreased or increased intensity relative to neighboring bands in each in vivo-treated sample compared with the pattern of relative band intensities in control naked DNA samples purified from the same cells. Figure 2B–F shows representative DNA sequencing gels for the in vivo footprint analysis; results are summarized in Figure 3.
Figure 2B shows in vivo footprint P1 at positions +56 and +58, which was identified with primer set TY528 on the lower strand of the paternal allele. Examination of relative band intensities revealed enhanced DMS reactivity at position +58, and protection from DMS modification at position+56 in the in vivo-treated samples containing the paternal allele (lanes 6–8) compared with the control naked genomic DNA samples (lanes 5). This pattern was not observed on the maternal allele (compare the naked DNA sample in lane 1 and the in vivo-treated samples in lanes 2–4). In vivo footprint P1 is located at the SNRPN translation initiation site within a sequence that shows weak similarity to an AP1 binding site. Furthermore, the P1 in vivo footprint lies within a previously reported negative cis-acting regulatory element mapped by transient expression assays within exon 1 of SNRPN (14). Footprints P2–P5 were detected with LMPCR primer set TY641 on the upper strand exclusively on the paternal allele (Figures 2C–E). Figure 2C shows enhanced DMS reactivity at positions −4 (footprint P2) and −13 (footprint P3) in the in vivo-treated samples from the paternal allele (lanes 6–8) compared with the naked DNA control (lane 5) and that was not observed on the maternal allele (compare lane 1 with lanes 2–4). Footprint P2 is associated with a potential E2F binding site, and P3 lies within one of the three potential CTCF binding sites located in the SNRPN 5′ region (see Figure 3). In vivo footprint P4, shown in Figure 2D, is detected at position −34 as a band of enhanced DMS reactivity exclusively on the paternal allele (compare lanes 5–6 with lane 4, and compare lanes 2 and 3 with lane 1). Searches of the transcription factor database did not identify a known transcription factor binding site associated with P4. However, footprint P4 is flanked on both sides by potential Sp1 binding sites (Figure 3). Figure 2E shows footprint P5 where two bands at positions −56 and −58 were protected from DMS modification exclusively on the paternal allele (compare lanes 5 and 6 with lanes 1 and 4, as well as lanes 2 and 3); this was confirmed in similar analysis using primer set TY540 (data not shown). Footprint P5 is associated with a potential binding site for NRF-1 (nuclear respiratory factor-1). P5 is also the only footprint that coincides with one of the six phylogenetically conserved sequences in the promoter region (Figure 3) and is included within the SBE (14). Figure 2F shows footprints M6 and P6 that were identified with primer set TY528 on the lower strand at positions −84 and −85 on the maternal and the paternal alleles, respectively. The band at position −85 showed enhanced DMS reactivity relative to bands above and below it in samples containing only the in vivo-treated paternal allele (lanes 6–8) but neither in control naked DNA samples (lanes 1 and 5) nor in in vivo-treated samples carrying only the maternal allele (lanes 2-4). This indicates that footprint P6 is specific to the paternal allele. In contrast, the band at position −84, termed footprint M6, showed enhanced DMS reactivity only in the in vivo-treated samples carrying the maternal allele (lanes 2–4). Because the finding of an in vivo footprint on the transcriptionally repressed maternal allele was somewhat unexpected, footprint M6 was confirmed by DNase I in vivo footprinting (data not shown). The sequence shared by footprinted sites M6 and P6 lies within a potential NRF-1 binding site that partially overlaps with a potential CTCF binding site. Because only altered DMS reactivity at guanine residues was used as the criterion for bona fide in vivo footprints, altered DMS reactivity at other nucleotides, such the adenine at position −83 in Figure 2F, was not deemed to reflect a footprint. In addition, nucleotides of apparent altered DMS reactivity that were not reproducible in at least 80% of the corresponding autoradiograms were also not considered to be footprinted (e.g. the apparent footprint at position −79 in Figure 2F, which was not reproducible in other autoradiograms). Analysis of the SNRPN 5′ flanking region with the remaining four LMPCR primer sets (Figure 2A) identified no additional in vivo footprints. Functional analysis of in vivo footprinted elements in DHS1 To examine the potential function of the in vivo footprinted sites, transient expression assays were performed in which each footprinted sequence was mutated in a reporter construct containing a 756 bp fragment of the SNRPN promoter region (from position −676 to +80) driving the firefly luciferase gene. Figure 2G shows the relative luciferase activity of the mutant constructs compared with a wild-type control construct in human neuroblastoma SK-N-SH cells (cells which showed readily detectable SNRPN mRNA levels in northern blots; data not shown).Mutations of the paternal-specific sites P2 and P5 [(d) and (f) in Figure 2G] significantly reduced the expression of the reporter gene, suggesting the association of P2 and P5 with cis-acting elements involved in SNRPN promoter function. Construct (g) in Figure 2G showed a reduction in reporter gene activity to ~1/3 of wild-type levels upon mutation of the M6/P6 site. This reduction was most likely due to effects on interactions associated with the P6 footprint on the transcriptionally active paternal allele and not with the M6 footprint specific to the silent maternal allele. Mutation of the sequences associated with footprinted sites P1 and P4 [(c) and (e) in Figure 2G] did not alter significantly the expression levels of the reporter. However, a potential role for footprint P1 is described below. Footprint P3 was not examined for effects on promoter function. Identification of an activator function associated with DHS2 To investigate the possibility that sequences within DHS2 might have an effect on SNRPN promoter function, a 2.2 kb (EcoRI–SmaI) genomic DNA fragment (termed 2.2-DHS2) that included DHS2 and flanking sequences was inserted into luciferase reporter constructs that included 756 bp of the SNRPN promoter (from positions −676 to +80) and analyzed by transient expression assays in SK-N-SH cells (Figure 4A). Constructs in which 2.2-DSH2 was cloned downstream from the luciferase gene in the forward [Figure 4A, (b)] or reverse orientations (c), and upstream of the promoter in the forward (d) or reverse (e) orientations showed an 8-, 2-, 5- and 8-fold increase in reporter activity, respectively, compared with a control construct lacking 2.2-DHS2 [(Figure 4A, (a)]. This strongly suggested that 2.2-DSH2 can act as an activator of the SNRPN promoter independent of orientation and position.The activation function of 2.2-DHS2 was then examined in the context of the SNRPN mutations of in vivo footprinted sites shown in Figure 2G. We used a luciferase reporter construct that contained 756 bp of the SNRPN promoter as before, plus 2.2-DHS2 [Figure 4A, construct (e)]. As shown by construct (d) of Figure 4B, mutation of site P1 (which had no effect on reporter activity in the absence of the enhancer; see Figure 2G) resulted in a ~50% increase in reporter activity in the presence of 2.2-DHS2 [compared with control construct (c)]. This suggested that regulatory elements associated with in vivo footprint P1 may be involved in negatively regulating the activation of the SNRPN promoter by 2.2-DHS2. Mutation of sites P2 and M6/P6 [Figure 4B, (e) and (h)] in the presence of 2.2-DHS2 had effects similar to that seen in the absence of 2.2-DHS2 (compare Figure 2G with Figure 4B). However, mutation of the footprint P5 resulted in a smaller reduction of promoter activity in the presence of 2.2-DHS2 [see Figure 4B, (c) versus (g)] compared with the absence of 2.2-DHS2 [see Figure 2G, (b) versus (f)]. Mutation of site P4 [construct (f)] had no effect on reporter activity, either in the presence [Figure 4B, (f)] or in the absence [Figure 2G, (e)] of 2.2-DHS2, which suggests that the factor(s) binding this site may play a role in functions other than direct promoter activity. Effect of the intronic activator on other promoters in the AS/PWS region The human SNRPN locus includes several upstream promoters that are functional only from the paternal allele at low levels (5). Therefore, as shown in Figure 4C, we investigated the effect of 2.2-DHS2 on two upstream SNRPN promoters, U1A and U1B. Each of these promoter regions was cloned into a luciferase reporter construct that included 2.2-DHS2 positioned downstream of the reporter in the forward orientation [as in Figure 4A, construct (b)] and analyzed using transient expression assays in SK-N-SH cells as before. Both the U1A and U1B promoters were activated significantly by 2.2-DHS2 relative to a construct lacking 2.2-DHS2, though to a slightly lesser extent than the major SNRPN promoter itself (5.1- and 5.7-fold versus 7.7-fold, respectively). We also examined the effect of the 2.2-DHS2 sequence on promoter activity of other imprinted genes within the AS/PWS region, including the paternal MKRN3 gene located ~1.4 Mb upstream of SNRPN, and the maternal UBE3A gene located ~0.5 Mb downstream of SNRPN. As shown in Figure 4C, both the UBE3A and MKRN3 promoters by themselves showed higher levels of basal promoter activity than the SNRPN promoter and were both activated by the 2.2-DHS2 fragment (2.0- and 1.9-fold, respectively), though to a lesser extent than any of the SNRPN-related promoters (5.1- to 7.7-fold). Analysis of the SV40 promoter (that was activated up to 14-fold by 2.2-DHS2; data not shown) further corroborated data that 2.2-DHS2 can function as an activator of heterologous promoters. Furthermore, 2.2-DHS2 did not repress the maternally expressed UBE3A promoter in these transient expression assays, which suggests that 2.2-DHS2 is unlikely to play a direct role in repressing maternal genes on the paternally inherited chromosome.Although 2.2-DHS2 does not activate the SNRPN promoter with equal efficiency in all orientations and positions (see Figure 4A), and distances (data not shown), or activate all heterologous promoters to the same degree (Figure 4C, and data not shown for the SV40 promoter), its activity in transient expression assays nonetheless meets the classical criteria for enhancer function. Identification of regulatory elements in 2.2-DHS2 To identify candidate cis-acting DNA sequences within 2.2-DHS2 that could mediate its activator function, we searched for evolutionarily conserved sequences in the human 2.2-DHS2 and the entire first intron of the mouse and rat Snrpn genes. As shown in Figure 5A, comparison of these sequences identified a single ~80 bp region that showed significant conservation among these species. Sequence analysis of the conserved sequence with the TRANSFAC database revealed highly conserved potential binding sites for the transcription factors Sp1, YY1 and NRF-1. Sp1 is a well-characterized ubiquitous transcriptional activator (25). NRF-1 regulates a variety of genes involved in mitochondrial function and energy metabolism (26), and a binding site for NRF-1 was also in vivo footprinted in the SNRPN promoter region (see Figure 3). YY1 is a ubiquitous transcription factor associated with transcriptional initiation, activation and repression [reviewed in (27)]. In addition, there are other highly conserved nucleotide sequences (from nucleotides +1302 to +1310, and from nucleotides +1333 to +1343 of the human sequence) that are not associated with known transcription factor binding sites.
Sequence analysis of the region associated with DHS2 also indicates that it contains a CpG island [(according to the criteria established by Takai and Jones (28)], which includes the 80 bp phylogenetically conserved region. This CpG island in DHS2 is distinct and separate from the CpG island at the SNRPN promoter. The CpG island in the promoter region has been well characterized in normal individuals and shown to be hypermethylated on the maternal allele and hypomethylated on the paternal allele (6). Analysis of the CpG island associated with DHS2 (and the 80 bp conserved sequence) also showed a similar pattern of allele-specific methylation. Digestion of genomic DNA from PWS and AS patients with methyl-sensitive restriction enzymes (HhaI to analyze CpGs at +1325 and +1375, and EaeI to analyze CpGs at +1354 and +1360) followed by PCR across the restriction site indicated that the paternal allele is hypomethylated and the maternal allele is hypermethylated (see Supplementary Figure S1). To determine whether cis-acting elements within the 80 bp conserved sequence contributed to the activation function of the 2.2-DHS2 fragment, subfragments of 2.2-DHS2 containing the conserved sequence were analyzed for the activation of the SNRPN promoter by transient expression assays as before (Figure 5B). A 94 bp BstNI–HaeIII subfragment (Bs-Ha in Figure 5B) containing the entire 80 bp conserved sequence activated the SNRPN promoter 4-fold over the promoter alone. Thus, the 80 bp conserved sequence was termed the conserved activator sequence (CAS). However, the BstNI–HaeIII subfragment exhibited only about 2/3 and 1/2 of the activation activity of a fragment that included sequences flanking the CAS (the N-S subfragment in Figure 5B) and the entire 2.2-DHS2 fragment, respectively. This suggested that the CAS requires additional sequences in 2.2-DHS2 to generate the full promoter activation activity of 2.2-DHS2. However, it is unclear what these sequences might be, since we have not identified additional regions of significant sequence conservation in 2.2-DHS2 outside of the CAS and no other DNase hypersensitive sites are detected in 2.2-DHS2 on the paternal allele (24). We also demonstrated a role for the putative YY1 binding site in the activation function associated with 2.2-DHS2. Transient expression assays yielded a 4.5-fold reduction in the activation of the SNRPN promoter by 2.2-DHS2, which was mutated at the putative YY1 binding site compared with wild-type 2.2-DHS2 (see Supplementary Material and Supplementary Figure S2A). Furthermore, electrophoretic mobility shift assay analyses and supershift assays with YY1 antibodies indicated that the potential YY1 binding site in the CAS can interact with YY1 in vitro (see Supplementary Figure S2B and C). ChIP analysis of transcription factor binding in vivo ChIP analysis was used to examine the binding of Sp1, YY1, NRF-1 and CTCF to DHS1 (i.e. the SNRPN promoter region) and to the CAS within DHS2 in vivo using lymphoblasts derived from AS and PWS patients. As shown in Figure 6A, five regions in the SNRPN locus were assayed by ChIP analysis: (1) the U1A upstream promoter; (2) 1.3 kb upstream of the SNRPN promoter; (3) the SNRPN promoter; (4) the intronic CAS and (5) 1.3 kb downstream of the CAS. The analysis showed strong association of YY1 with the intronic CAS in vivo on the transcriptionally active paternal allele (i.e. in AS cells) and not on the maternal allele (i.e. in PWS cells, Figure 6B). In contrast, similar ChIP analysis of the non-imprinted glucocorticoid receptor gene, known to be bound by YY1 at an upstream regulatory region (29), showed association of YY1 in this region on both alleles (i.e. in both the AS- and PWS-derived cell lines; see Supplementary Figure S3A). Analysis of NRF-1 by ChIP assays demonstrated the association of NRF-1 with both the SNRPN promoter region as well as the intronic CAS (Figure 6B) specifically on the paternal allele. The association of NRF-1 with the SNRPN promoter is most likely to be at the potential NRF-1 binding sites associated with paternal footprints P5 and/or P6 (Figure 3). As a positive control for NRF-1 association, ChIP analysis of the 3′ non-coding region of the non-imprinted myotonic dystrophy protein kinase gene (DMPK) that contains potential NRF-1 binding sites (as indicated by analysis of the TRANSFAC database; S. Rodriguez-Jato, unpublished data) detected NRF-1 binding in this region in both the AS- and PWS-derived cell lines (see Supplementary Figure S3A). The U1A upstream promoter of the SNRPN locus also contains a potential NRF-1 binding site (based on analysis of the TRANSFAC database); however, no association of NRF-1 with U1A was detected on either allele (Figure 6B). This is not unexpected since the upstream SNRPN transcripts are not expressed in blood (5) and the U1 promoters have been reported not to be nuclease hypersensitive in lymphoblast cells (24). Analysis of Sp1 association using ChIP assays showed interaction of Sp1 with both the promoter region as well as the CAS only on the paternal allele (Figure 6B).
Despite the fact that two in vivo footprints in the SNRPN promoter region were associated with potential CTCF binding sites (Figure 3), we have been unable to detect in vivo interactions involving CTCF on either the paternal or the maternal allele of the SNRPN promoter region by ChIP assays (Supplementary Figure S3B). As a positive control in these assays, we analyzed the previously reported association of CTCF with the 3′ non-coding region of the DMPK gene (30) and confirmed the interaction of CTCF at this locus in both AS- and PWS-derived cell lines (Supplementary Figure S3A). Thus, it is likely that CTCF does not interact with either the maternal or paternal alleles of the SNRPN promoter in lymphoblast cells. Association of RNA polymerase II with the intronic CAS YY1 has been shown to interact with and recruit RNA polymerase II to core promoters and activate transcription (27). Therefore, we examined the possibility that the intronic CAS, with the possible involvement of YY1, could act as a site for recruitment of pol II to the SNRPN locus. Antibodies specific to the unphosphorylated form of RNA pol II were used in ChIP assays of the same SNRPN regions as before (Figure 6A); unphosphorylated RNA pol II is the non-processive form typically associated with transcription initiation at promoters. As shown in Figure 6B, ChIP analysis revealed strong association of unphosphorylated pol II with both the CAS and the promoter specifically on the paternal allele. Furthermore, the level of unphosphorylated pol II on the paternal allele was higher at the CAS compared with the SNRPN promoter region. No significant amplification of the regions 1.3 kb upstream and downstream of the promoter and the CAS, respectively, was observed. Analysis of the U1A upstream promoter, which is inactive in blood, also did not detect association of pol II. These findings are consistent with a model of transcriptional activation in which the CAS may act to recruit unphosphorylated RNA pol II to the SNRPN locus (see Discussion).Analysis of histone modifications in DHS1 and DHS2 Allele-specific patterns of histone modification associated with the SNRPN promoter region have been reported previously (31,32) and consist of increased levels of H3 and H4 acetylation (H3 and H4 Ac) and H3 lysine 4 methylation (H3-K4 Me) on the paternal allele, and H3 lysine 9 methylation (H3-K9 Me) on the maternal allele. However, the elevated levels of H3-K4 diMe and H3-K9 diMe in the promoter region decrease drastically ~200 and ~600 bp, respectively, downstream from the transcription initiation site (32). To determine whether the downstream CAS has a distinct pattern of allele-specific histone modifications, we used ChIP analysis on AS- and PWS-derived lymphoblasts as before. We analyzed levels of H4 Ac, H3-K9 Ac and H3-K4 diMe, all of which are generally associated with transcriptional activation and are commonly elevated at the 5′ and regulatory regions of active genes (33,34). As shown in Figure 6C, the levels of all these modifications in histones associated with the transcriptionally active paternal allele are elevated at the CAS (and SNRPN promoter), and lower at the inactive U1A promoter and regions flanking the promoter and CAS. Significant levels of these histone modifications were not observed at any region assayed on the maternal allele. Of particular interest are the high levels of H3-K4 diMe at the CAS, given that the levels of H3-K4 diMe are known to drastically decrease 200 bp downstream from the SNRPN transcription initiation site (32). This suggests that the SNRPN promoter region and CAS are associated with two separate and distinct regions of chromatin, both characterized by histone modification patterns typical of regulatory regions in active chromatin (rather than co-localization of these two regions within one continuous stretch of chromatin having the same histone modification patterns). These results are consistent with the notion that the CAS is a regulatory region within the endogenous SNRPN locus. | |||||||||
DISCUSSION We have analyzed the 5′ region of the SNRPN gene for cis- and trans-acting regulatory elements that may mediate SNRPN transcription. These studies focused on the two DNase I hypersensitive sites located within the SNRPN promoter (DHS1) and first intron (DHS2). Both hypersensitive sites co-localize with CpG islands that are differentially methylated on the maternal and paternal alleles, a characteristic of regulatory regions associated with imprinted genes (35). In vivo footprint analysis of the SNRPN promoter region DHS1 is located within the PWS-SRO and co-localizes with the SNRPN promoter. In vivo footprint analysis of the endogenous promoter was used to identify cis-acting elements associated with six footprints on the paternal allele (P1–P6) and one on the maternal allele (M6). In vivo footprint P1 is located adjacent to the translation initiation site in exon 1 and is included within a previously described repressor sequence (14). However, our analysis showed that the repressor function appears to be limited to negatively regulating the activation of the SNRPN promoter by the enhancer associated with DHS2 (Figure 4B). Footprint P2, a potential E2F binding site, appears to be a cis-acting element essential for SNRPN promoter function in the presence or in the absence of the enhancer (Figures 2G and 4B). E2F consists of a family of transcription factors that participate in transcription activation and repression (36). The sequence associated with footprint P3 bears a resemblance to a potential CTCF binding site (37). CTCF has been implicated in the regulation of the imprinted H19/IGF2 locus where it is associated with an intergenic insulator (38). However, we have been unable to demonstrate by ChIP assays that CTCF is bound in the SNRPN promoter region on either the paternal or maternal alleles. Therefore, in vivo footprint P3 is likely to be due to binding of a factor other than CTCF in lymphoblasts. It is also possible that CTCF is bound to this site in cell types other than lymphoblasts, possibly in the germ line. By transient expression assays, footprint P4 does not appear to play a direct role in SNRPN promoter function (Figures 2G and 4C) and does not seem to correspond to any known transcription factor binding site. However, it is located between two potential binding sites for Sp1, which our ChIP analysis showed to be associated with the SNRPN promoter region (Figure 6B). This suggests that footprint P4 may be the result of Sp1 interaction with one or both of the flanking Sp1 binding sites. Footprint P5 is associated with a potential NRF-1 binding site, which is phylogenetically conserved between the human and mouse SNRPN promoters. The interaction of NRF-1 with the paternal promoter region in vivo was demonstrated by ChIP (Figure 6B). In addition, footprint P5 overlaps with the previously reported SBE cis-acting element in the SNRPN promoter (14). Our analysis also found that P5 is essential for promoter function (Figure 2G). Footprints P6 on the paternal allele and M6 on the maternal allele show two distinctly different in vivo footprint patterns that localize to adjacent guanine nucleotides (Figure 2F). These data suggest that P6 and M6 represent interaction of different trans-acting factors at this sequence that perform different functions on the maternal and paternal alleles; on the maternal allele, the factor responsible for M6 may participate in the silencing of the SNRPN gene, while on the paternal allele, the factor representing P6 may contribute to SNRPN promoter function and gene activation as suggested by mutation analysis of the M6/P6 sequence (Figure 2G). The sequence associated with footprints P6 and M6 corresponds to overlapping potential NRF-1 and CTCF binding sites (Figure 3). ChIP assays in lymphoblasts confirmed the interaction of NRF-1 with the paternal allele of the SNRPN 5′ region (Figure 6B), but failed to detect association of CTCF with either the paternal or the maternal allele (Supplementary Figure S3B). This would suggest that footprint P6 on the paternal allele may be generated by interaction with NRF-1, and that it is unlikely that CTCF interacts with neither P6 nor M6 in these cells Therefore, it is currently unclear what factor(s) is generating the M6 footprint on the maternal allele in these lymphoblast cells.Our in vivo footprint analysis of the SNRPN promoter included all six sequences that are phylogenetically conserved between the human and the mouse homologous regions (7). However, only one of these sequences was in vivo footprinted in lymphoblasts (footprint P5), which suggests that the other five phylogenetically conserved sequences might not be functional cis-acting elements in lymphoblasts and may be restricted to function in other somatic cell types, the germ line, and/or in early embryo development. In addition, Kantor et al. (16) have identified a series of sequences within the mouse Snrpn promoter (not including the SBE) that are involved in the correct imprinting of a transgene in mice that includes the mouse Snrpn promoter and the human AS-SRO. However, these mouse promoter sequences do not overlap with any of the six phylogenetically conserved sequences and are not present in the human SNRPN promoter. Therefore, it is not possible at this time to draw a comparison between the sequences identified by Kantor et al. (16) and the footprints identified in our study. The region associated with DHS1 is located within the PWS-SRO and has been proposed to be involved in PWS-IC function (7,24) as well as in SNRPN promoter function. Thus, one or more of the cis-acting regulatory sequences associated with DHS1 identified here by in vivo footprinting may be essential not only for SNRPN expression as shown by our transient expression assays, but may also be functional elements of the PWS-IC. The enhancer associated with DHS2 Transient expression assays demonstrated that intronic sequences associated with DHS2 could function as an enhancer of the SNRPN promoter (Figure 4A) and included a highly conserved region we have termed the CAS. The CAS contains evolutionarily conserved binding sites for three known transcription factors, NRF-1, YY1 and Sp1 (Figure 5A), as well as two other highly conserved sequences that may serve as binding sites for transcription factors that have not yet been identified. One of these latter two sequences in the CAS (from nucleotides +1302 to +1310) is highly homologous to the MPI2 sequence in the mouse Snrpn promoter, which has been shown to be involved in maintenance of paternal imprinting of a construct in transgenic mice that contains the human AS-SRO and the mouse Snrpn promoter, though the corresponding DNA-binding factor for MPI2 element in the mouse promoter has not been identified (16).In addition to activating the SNRPN promoter, our data have shown that the intronic enhancer can also function to activate transcription of heterologous promoters in transient expression assays, including the MKRN3 promoter (Figure 4C). Thus, it is conceivable that the enhancer could also contribute to PWS-IC function by activating the promoters of other genes in the AS/PWS domain that are expressed only from the paternal chromosome (e.g. MKNR3, NDN and MAGEL2). This would be consistent with the model of PWS-IC function proposed by Brannan and Bartolomei (39) in which the PWS-IC acts as a positive regulator of genes expressed only from the paternal chromosome. ChIP analysis showed that NRF-1, YY1 and Sp1 are all preferentially associated with the CAS region on the paternal allele in vivo (Figure 6B). Gel-shift analysis of the potential YY1 binding site demonstrated the ability of YY1 to bind to this site in vitro (Supplementary Figure S2C), and transient expression assays showed that this YY1 binding site is essential for the activation function of this intronic region (Supplementary Figure S2A). YY1 is a ubiquitous factor that has been shown to act as a transcriptional initiator, activator and repressor (27), suggesting the possibility that YY1 could be involved in mediating both activation and silencing of the SNRPN locus, depending upon the parent-of-origin of the allele. A role for YY1 in the regulation of imprinted genes has been postulated for the imprinted PEG3 locus where YY1 appears to act as an insulator-binding protein (40). YY1 is also reported to associate with the nuclear matrix (41), suggesting a possible role for the CAS in tethering this region of the paternally inherited SNRPN locus to the matrix. Similar to previous reports for the promoter region of SNRPN (31,32), the intronic CAS on the paternal chromosome (but not on the maternal chromosome) is associated with histone modification patterns characteristic of transcriptionally active chromatin (33) (Figure 6C). This is consistent with a role for the CAS in regulating SNRPN transcription in vivo. In addition, both YY1 and NRF-1 are known to interact with histone modifying enzymes (27,42), suggesting that these factors may have a role in establishing and/or maintaining differential patterns of histone modification on the paternal and/or maternal chromosomes. The role of NRF-1 in the regulation of the SNRPN locus DNA sequence analysis and ChIP assays indicate that NRF-1 interacts with both the promoter region and the CAS on the paternal allele (Figure 6B). The upstream SNRPN promoters U1A and U1B also contain potential NRF-1 binding sites; however, no interaction of NRF-1 was detected with either allele of the U1A promoter (Figure 6B), which is not active in blood cells (5). In addition, we have identified by sequence analysis a conserved potential NRF-1 binding site in the NDN promoter region, which coincides with a sequence that is in vivo footprinted on the paternal NDN allele only (43). The fact that NRF-1 may be regulating at least some of the genes in the PWS/AS region is interesting because of the involvement of NRF-1 in the regulation of genes related to mitochondrial biogenesis and function, metabolism (including growth factor receptors and factors involved in glucose homeostasis), DNA replication and transcriptional regulation (26). This suggests that genes in the AS/PWS region and genes that function in metabolism and in cellular energetics may be co-regulated through the common transcriptional regulator NRF-1. This would further suggest a potential link between energy metabolism and aspects of the PWS phenotype (e.g. obesity and growth factor deficiency). However, the resting metabolic rate of PWS patients does not seem to differ from that of normal obese individuals (44).Potential roles of the CAS Using antibodies against unphosphorylated pol II in ChIP assays, we found that the non-elongating form of pol II was significantly enriched at the CAS (Figure 6B), an intronic region that would normally be expected to be associated only with elongating phosphorylated pol II. This suggested that the CAS may be involved in recruiting and accumulating the initiating form of pol II at the SNRPN locus. This recruitment of pol II to the CAS could be mediated by YY1, which is known to interact with and recruit pol II to promoters (27), and enhanced by Sp1 and NRF-1. Furthermore, recruitment of pol II to the CAS may serve as a mechanism for facilitating activation and transcription of the SNRPN gene by transfer of pol II from the CAS to the promoter and/or by increasing the local concentration of pol II in the vicinity of the SNRPN promoter. These mechanisms of promoter activation by distal regulatory elements have been proposed for a variety of other loci. For example, the locus control region (LCR) of the major histocompatibility complex (MHC) class II locus has been proposed to recruit pol II and transfer it to the promoter via either looping or tracking mechanisms (45). Alternatively, the LCR of the β-globin locus (46) has been postulated to recruit RNA pol II (and other trans-acting factors) by formation of a holocomplex or active chromatin hub (ACH) (47). The formation of an ACH by the PWS-IC to coordinately activate transcription of genes expressed from the paternal chromosome would be consistent with the model for PWS-IC function proposed by Brannan and Bartolomei (39). Thus, we speculate that the PWS-IC may function via formation of an ACH. Regulatory elements within the SNRPN 5′ region, including those associated with both DHS1 and DHS2 (and the CAS) may participate in the formation of a holocomplex (i.e. a chromatin hub) that interacts with and recruits regulatory regions from each of the paternally expressed genes (e.g. MAGEL2, NDN and MKRN3) to form an ACH specifically on the paternally inherited chromosome. This ACH would coordinately facilitate and activate transcription of the genes that are expressed exclusively on the paternal chromosome, in part by creating a highly localized region within the nucleus that is enriched in trans-acting positive regulators of transcription, such as histone acetyltransferases, general transcription factors and RNA polymerase II. Participation of the CAS in formation of the ACH would be consistent with our finding of unphosphorylated pol II at the CAS. As a variation of this model, it is also possible that the PWS-IC acts by facilitating recruitment of genes expressed from the paternal chromosome into transcription factories described by Osborne et al. (48). Both models would be consistent with the long-range interactions reported between differentially methylated regulatory regions that regulate imprinting in the Igf2/H19 domain (49).Participation of the CAS in formation of an ACH would also suggest a role for the CAS in PWS-IC function. The location of the CAS relative to targeted deletions of the mouse Snrpn gene that serve as mouse models for murine AS/PWS imprinting defects provide evidence consistent with such a role. A targeted 35 kb deletion extending 19 kb upstream and 16 kb downstream of the mouse Snrpn promoter region that included both the mouse CAS and Snrpn promoter region resulted in 100% neonatal lethality in mice when paternally inherited (50). This was accompanied by a loss PWS-IC function, including loss of paternal-specific gene expression and DNA methylation patterns of other imprinted genes in the region. These results suggested that the murine PWS-IC is contained within the 35 kb deletion. A second targeted deletion in this region, in which 0.9 kb of the Snrpn promoter region and exon 1 were deleted, showed no disruption of imprinted gene expression or DNA methylation patterns, suggesting that the Snrpn promoter region does not, by itself, function as the PWS-IC in the mouse (19). However, a 4.8 kb deletion that included the previously deleted 0.9 kb promoter region as well as sequences ~1.9 kb upstream and ~2 kb downstream yielded mice with a partial imprinting defect, and 40–50% perinatal lethality when paternally inherited (19). This would suggest that sequences 1.9 kb upstream and/or 2 kb downstream of the previous 0.9 kb promoter deletion contain cis-acting elements that contribute to PWS-IC function in the mouse. The mouse CAS sequence is located ~1.8 kb downstream of the Snrpn transcription initiation site and is, therefore, contained within the 2 kb region downstream of the promoter, suggesting that loss of the CAS could have contributed to the partial imprinting defect observed for the 4.8 kb Snrpn deletion. This would further suggest that the CAS may be a functional component of the PWS-IC in the mouse, and because of the high degree of sequence conservation of the CAS between the humans and rodents, the CAS may also act as a functional component of the human PWS-IC. Because mice carrying the 4.8 kb deletion on the paternal chromosome do not exhibit the full mutant phenotype shown by mice carrying the paternally inherited 35 kb Snrpn deletion, it is conceivable that other regions within the 35 kb deletion (but outside of the 4.8 kb deletion) may also contribute to PWS-IC function in the mouse (16). In addition, Kantor et al. have proposed that components of the PWS-IC may be redundant, such that deletion of one component, e.g. the 0.9 kb Snrpn promoter deletion, may be compensated for by other components of the mouse PWS-IC (16). This functional redundancy in a multicomponent PWS-IC may explain why no human microdeletions that remove the CAS and not the SNRPN promoter have been identified to date. | |||||||||
SUPPLEMENTARY MATERIAL Supplementary Material is available at NAR Online. | |||||||||
[Supplementary Material]
| |||||||||
Acknowledgments This work was supported by the US Public Health Service Grant RO1-HD36436 and RO1-HD36417 from the National Institute of Child Health and Human Development. The authors thank Christopher Glenn for assistance with DNase I hypersensitivity assays, Teresa Kunkel for assistance with the gel mobility shift assays, Richard Scarpulla for anti-NRF-1 antibodies, Edward Seto for purified YY1 protein and oligonucleotides with a YY1 binding site and Jorg Bungert for valuable discussions. Funding to pay the Open Access publication charges for this article was provided by National Institutes of Health Grant RO1-GM44286. Conflict of interest statement. None declared. | |||||||||
REFERENCES 1. Schmauss, C.; Brines, M.L.; Lerner, M.R. The gene encoding the small nuclear ribonucleoprotein-associated protein N is expressed at high levels in neurons. J. Biol. Chem. 1992;267:8521–8529. [PubMed] 2. Gray, T.A.; Saitoh, S.; Nicholls, R.D. An imprinted, mammalian bicistronic transcript encodes two independent proteins. Proc. Natl Acad. Sci. USA. 1999;96:5616–5621. [PubMed] 3. Runte, M.; Huttenhofer, A.; Gross, S.; Kiefmann, M.; Horsthemke, B.; Buiting, K. The IC-SNURF-SNRPN transcript serves as a host for multiple small nucleolar RNA species and as an antisense RNA for UBE3A. Hum. Mol. Genet. 2001;10:2687–2700. [PubMed] 4. Sutcliffe, J.S.; Nakao, M.; Christian, S.; Orstavik, K.H.; Tommerup, N.; Ledbetter, D.H.; Beaudet, A.L. Deletions of a differentially methylated CpG island at the SNRPN gene define a putative imprinting control region. Nature Genet. 1994;8:52–58. [PubMed] 5. Dittrich, B.; Buiting, K.; Korn, B.; Rickard, S.; Buxton, J.; Saitoh, S.; Nicholls, R.D.; Poustka, A.; Winterpacht, A.; Zabel, B., et al. Imprint switching on human chromosome 15 may involve alternative transcripts of the SNRPN gene. Nature Genet. 1996;14:163–170. [PubMed] 6. Nicholls, R.D.; Knepper, J.L. Genome organization, function, and imprinting in Prader-Willi and Angelman syndromes. Annu. Rev. Genomics. Hum. Genet. 2001;2:153–175. [PubMed] 7. Ohta, T.; Gray, T.A.; Rogan, P.K.; Buiting, K.; Gabriel, J.M.; Saitoh, S.; Muralidhar, B.; Bilienska, B.; Krajewska-Walasek, M.; Driscoll, D.J., et al. Imprinting-mutation mechanisms in Prader-Willi syndrome. Am. J. Hum. Genet. 1999;64:397–413. [PubMed] 8. Buiting, K.; Lich, C.; Cottrell, S.; Barnicoat, A.; Horsthemke, B. A 5-kb imprinting center deletion in a family with Angelman syndrome reduces the shortest region of deletion overlap to 880 bp. Hum. Genet. 1999;105:665–666. [PubMed] 9. Bielinska, B.; Blaydes, S.M.; Buiting, K.; Yang, T.; Krajewska-Walasek, M.; Horsthemke, B.; Brannan, C.I. De novo deletions of SNRPN exon 1 in early human and mouse embryos result in a paternal to maternal imprint switch. Nature Genet. 2000;25:74–78. [PubMed] 10. El-Maarri, O.; Buiting, K.; Peery, E.G.; Kroisel, P.M.; Balaban, B.; Wagner, K.; Urman, B.; Heyd, J.; Lich, C.; Brannan, C.I., et al. Maternal methylation imprints on human chromosome 15 are established during or after fertilization. Nature Genet. 2001;27:341–344. [PubMed] 11. Perk, J.; Makedonski, K.; Lande, L.; Cedar, H.; Razin, A.; Shemer, R. The imprinting mechanism of the Prader-Willi/Angelman regional control center. EMBO J. 2002;21:5807–5814. [PubMed] 12. Nicholls, R.D.; Saitoh, S.; Horsthemke, B. Imprinting in Prader-Willi and Angelman syndromes. Trends Genet. 1998;14:194–200. [PubMed] 13. Huq, A.H.; Sutcliffe, J.S.; Nakao, M.; Shen, Y.; Gibbs, R.A.; Beaudet, A.L. Sequencing and functional analysis of the SNRPN promoter: in vitro methylation abolishes promoter activity. Genome Res. 1997;7:642–648. [PubMed] 14. Green Finberg, Y.; Kantor, B.; Hershko, A.Y.; Razin, A. Characterization of the human Snrpn minimal promoter and cis elements within it. Gene. 2003;304:201–206. [PubMed] 15. Hershko, A.Y.; Finberg, Y.; Kantor, B.; Shemer, R.; Razin, A. The mouse Snrpn minimal promoter and its human orthologue: activity and imprinting. Genes Cells. 2001;6:967–975. [PubMed] 16. Kantor, B.; Makedonski, K.; Green-Finberg, Y.; Shemer, R.; Razin, A. Control elements within the PWS/AS imprinting box and their function in the imprinting process. Hum. Mol. Genet. 2004;13:751–762. [PubMed] 17. Kantor, B.; Kaufman, Y.; Makedonski, K.; Razin, A.; Shemer, R. Establishing the epigenetic status of the Prader-Willi/Angelman imprinting center in the gametes and embryo. Hum. Mol. Genet. 2004;13:2767–2779. [PubMed] 18. Shemer, R.; Hershko, A.Y.; Perk, J.; Mostoslavsky, R.; Tsuberi, B.; Cedar, H.; Buiting, K.; Razin, A. The imprinting box of the Prader-Willi/Angelman syndrome domain. Nature Genet. 2000;26:440–443. [PubMed] 19. Bressler, J.; Tsai, T.F.; Wu, M.Y.; Tsai, S.F.; Ramirez, M.A.; Armstrong, D.; Beaudet, A.L. The SNRPN promoter is not required for genomic imprinting of the Prader-Willi/Angelman domain in mice. Nature Genet. 2001;28:232–240. [PubMed] 20. Chen, C.; Yang, T.P. Nucleosomes are translationally positioned on the active allele and rotationally positioned on the inactive allele of the HPRT promoter. Mol. Cell. Biol. 2001;21:7682–7695. [PubMed] 21. Glenn, C.C.; Saitoh, S.; Jong, M.T.; Filbrandt, M.M.; Surti, U.; Driscoll, D.J.; Nicholls, R.D. Gene structure, DNA methylation, and imprinted expression of the human SNRPN gene. Am. J. Hum. Genet. 1996;58:335–346. [PubMed] 22. Hornstra, I.K.; Yang, T.P. In vivo footprinting and genomic sequencing by ligation-mediated PCR. Anal. Biochem. 1993;213:179–193. [PubMed] 23. Leach, K.M.; Nightingale, K.; Igarashi, K.; Levings, P.P.; Engel, J.D.; Becker, P.B.; Bungert, J. Reconstitution of human beta-globin locus control region hypersensitive sites in the absence of chromatin assembly. Mol. Cell. Biol. 2001;21:2629–2640. [PubMed] 24. Schweizer, J.; Zynger, D.; Francke, U. In vivo nuclease hypersensitivity studies reveal multiple sites of parental origin-dependent differential chromatin conformation in the 150 kb SNRPN transcription unit. Hum. Mol. Genet. 1999;8:555–566. [PubMed] 25. Bouwman, P.; Philipsen, S. Regulation of the activity of Sp1-related transcription factors. Mol. Cell. Endocrinol. 2002;195:27–38. [PubMed] 26. Cam, H.; Balciunaite, E.; Blais, A.; Spektor, A.; Scarpulla, R.C.; Young, R.; Kluger, Y.; Dynlacht, B.D. A common set of gene regulatory networks links metabolism and growth inhibition. Mol. Cell. 2004;16:399–411. [PubMed] 27. Thomas, M.J.; Seto, E. Unlocking the mechanisms of transcription factor YY1: are chromatin modifying enzymes the key? Gene. 1999;236:197–208. [PubMed] 28. Takai, D.; Jones, P.A. Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc. Natl Acad. Sci. USA. 2002;99:3740–3745. [PubMed] 29. Breslin, M.B.; Vedeckis, W.V. The human glucocorticoid receptor promoter upstream sequences contain binding sites for the ubiquitous transcription factor, Yin Yang 1. J. Steroid. Biochem. Mol. Biol. 1998;67:369–381. [PubMed] 30. Filippova, G.N.; Thienes, C.P.; Penn, B.H.; Cho, D.H.; Hu, Y.J.; Moore, J.M.; Klesert, T.R.; Lobanenkov, V.V.; Tapscott, S.J. CTCF-binding sites flank CTG/CAG repeats and form a methylation-sensitive insulator at the DM1 locus. Nature Genet. 2001;28:335–343. [PubMed] 31. Saitoh, S.; Wada, T. Parent-of-origin specific histone acetylation and reactivation of a key imprinted gene locus in Prader-Willi syndrome. Am. J. Hum. Genet. 2000;66:1958–1962. [PubMed] 32. Xin, Z.; Allis, C.D.; Wagstaff, J. Parent-specific complementary patterns of histone H3 lysine 9 and H3 lysine 4 methylation at the Prader-Willi syndrome imprinting center. Am. J. Hum. Genet. 2001;69:1389–1394. [PubMed] 33. Liang, G.; Lin, J.C.; Wei, V.; Yoo, C.; Cheng, J.C.; Nguyen, C.T.; Weisenberger, D.J.; Egger, G.; Takai, D.; Gonzales, F.A., et al. Distinct localization of histone H3 acetylation and H3-K4 methylation to the transcription start sites in the human genome. Proc. Natl Acad. Sci. USA. 2004;101:7357–7362. [PubMed] 34. Valverde-Garduno, V.; Guyot, B.; Anguita, E.; Hamlett, I.; Porcher, C.; Vyas, P. Differences in the chromatin structure and cis-element organization of the human and mouse GATA1 loci: implications for cis-element identification. Blood. 2004;104:3106–3116. [PubMed] 35. Mann, J.R.; Szabo, P.E.; Reed, M.R.; Singer-Sam, J. Methylated DNA sequences in genomic imprinting. Crit. Rev. Eukaryot. Gene Expr. 2000;10:241–257. [PubMed] 36. DeGregori, J. The genetics of the E2F family of transcription factors: shared functions and unique roles. Biochim. Biophys. Acta. 2002;1602:131–150. [PubMed] 37. Chao, W.; Huynh, K.D.; Spencer, R.J.; Davidow, L.S.; Lee, J.T. CTCF, a candidate trans-acting factor for X-inactivation choice. Science. 2002;295:345–347. [PubMed] 38. Lewis, A.; Murrell, A. Genomic imprinting: CTCF protects the boundaries. Curr. Biol. 2004;14:R284–R286. [PubMed] 39. Brannan, C.I.; Bartolomei, M.S. Mechanisms of genomic imprinting. Curr. Opin. Genet. Dev. 1999;9:164–170. [PubMed] 40. Kim, J.; Kollhoff, A.; Bergmann, A.; Stubbs, L. Methylation-sensitive binding of transcription factor YY1 to an insulator sequence within the paternally expressed imprinted gene, Peg3. Hum. Mol. Genet. 2003;12:233–245. [PubMed] 41. Bushmeyer, S.M.; Atchison, M.L. Identification of YY1 sequences necessary for association with the nuclear matrix and for transcriptional repression functions. J. Cell. Biochem. 1998;68:484–499. [PubMed] 42. Izumi, H.; Ohta, R.; Nagatani, G.; Ise, T.; Nakayama, Y.; Nomoto, M.; Kohno, K. p300/CBP-associated factor (P/CAF) interacts with nuclear respiratory factor-1 to regulate the UDP-N-acetyl-alpha-d-galactosamine: polypeptide N-acetylgalactosaminyltransferase-3 gene. Biochem. J. 2003;373:713–722. [PubMed] 43. Hanel, M.L.; Lau, J.C.; Paradis, I.; Drouin, R.; Wevrick, R. Chromatin modification of the human imprinted NDN (necdin) gene detected by in vivo footprinting. J. Cell. Biochem. 2005;94:1046–1057. [PubMed] 44. Goldstone, A.P.; Brynes, A.E.; Thomas, E.L.; Bell, J.D.; Frost, G.; Holland, A.; Ghatei, M.A.; Bloom, S.R. Resting metabolic rate, plasma leptin concentrations, leptin receptor expression, and adipose tissue measured by whole-body magnetic resonance imaging in women with Prader-Willi syndrome. Am. J. Clin. Nutr. 2002;75:468–475. [PubMed] 45. Masternak, K.; Peyraud, N.; Krawczyk, M.; Barras, E.; Reith, W. Chromatin remodeling and extragenic transcription at the MHC class II locus control region. Nature Immunol. 2003;4:132–137. [PubMed] 46. Johnson, K.D.; Christensen, H.M.; Zhao, B.; Bresnick, E.H. Distinct mechanisms control RNA polymerase II recruitment to a tissue-specific locus control region and a downstream promoter. Mol. Cell. 2001;8:465–471. [PubMed] 47. de Laat, W.; Grosveld, F. Spatial organization of gene expression: the active chromatin hub. Chromosome Res. 2003;11:447–459. [PubMed] 48. Osborne, C.S.; Chakalova, L.; Brown, K.E.; Carter, D.; Horton, A.; Debrand, E.; Goyenechea, B.; Mitchell, J.A.; Lopes, S.; Reik, W., et al. Active genes dynamically colocalize to shared sites of ongoing transcription. Nature Genet. 2004;36:1065–1071. [PubMed] 49. Murrell, A.; Heeson, S.; Reik, W. Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 into parent-specific chromatin loops. Nature Genet. 2004;36:889–893. [PubMed] 50. Yang, T.; Adamson, T.E.; Resnick, J.L.; Leff, S.; Wevrick, R.; Francke, U.; Jenkins, N.A.; Copeland, N.G.; Brannan, C.I. A mouse model for Prader-Willi syndrome imprinting-centre mutations. Nature Genet. 1998;19:25–31. [PubMed] | |||||||||
