 |
 |
GEO help: Mouse over screen elements for information. |
|
| Status |
Public on Jul 24, 2023 |
| Title |
ChIP-seq HDAC2 pMN day4 progenitors rep2 |
| Sample type |
SRA |
| |
|
| Source name |
motor neuron progenitor cells
|
| Organism |
Mus musculus |
| Characteristics |
cell type: ES-derived motor neuron progenitors
cell line: mES A2lox.Cre
treatment: untreated
differentiation stage: day4
chip antibody: rabbit anti HDAC2 (Abcam, ab7029)
|
| Treatment protocol |
Selected samples were treated with doxycycline (DOX) for 14-16hrs prior to collection. See SAMPLES section.
|
| Growth protocol |
Motor neuron differentiation of ESCs was performed as previously described (Rhee et al., 2016; Tan et al., 2016), with slight modifications. Briefly, ESCs were trypsinized (Thermo Fisher Scientific, SC25300054) and seeded at 5 x 105 cells in 12mL of medium (Advanced DMEM/F12:Neurobasal (1:1) Medium; Thermo Fisher Scientific, 12634-010, 21103-049) supplemented with 10% Knockout Serum Replacement (Thermo Fisher Scientific, 10828-028), 2 mM Glutamax (Invitrogen, 35050061), and 0.1 mM b-mercaptoethanol (Sigma-Aldrich, M3148) to initiate formation of embryoid bodies (day0). Embryoid bodies were split 1:4 on day2 of differentiation, and patterning was induced by supplementing fresh media with 1μM all-trans retinoic acid (RA, Sigma, R2625) and 0.5μM Smoothened agonist of hedgehog signaling (SAG, EMD Millipore, 566660). Doxycycline (DOX; Thermo Fisher Scientific, NC0424034) was added to the cultures at 1μg/mL on late day3 of differentiation for 24hrs. DAPT (Selleck, S2215) was added to the culture medium at 5μM on early day4 (~12hrs after DOX addition). DOX was washed on late day4 by replacing medium in both treated and control conditions. Fresh RA, SAG, and DAPT were re-added to this medium. On day5, the medium was again replaced, and only DAPT was re-added to the cultures. Embryoid bodies were collected on day4 (~16hrs post DOX induction) and day6 for analysis. For ChIP experiments, the same conditions were used but scaled to seed 4 x 106 cells on day0, and 1X Penicillin-Streptomycin (Thermo Fisher Scientific, 15140-122) was maintained in the cultures throughout the differentiation. DAPT was not added to cells collected on day4.
|
| Extracted molecule |
genomic DNA |
| Extraction protocol |
40-50 million cells were fixed with 2mM DSG, followed by 1% formaldehyde. Lysates were clarified from sonicated nuclei, and HDAC2-DNA complexes were isolated with antibody.
Purified DNA fragments were processed according to the Illumina/Solexa standard sequencing protocol.
|
| |
|
| Library strategy |
ChIP-Seq |
| Library source |
genomic |
| Library selection |
ChIP |
| Instrument model |
Illumina HiSeq 2000 |
| |
|
| Description |
HDAC2_logCPM.txt
RawCounts_HDAC2.txt
SigHDAC2.txt
|
| Data processing |
RNA-seq: Reads were mapped to the mouse genome (mm10) using the RNA-seq alignment algorithm in the STAR (Spliced Transcripts Alignment to a Reference) software package (version 2.3.1) (Dobin et al., 2013).
RNA-seq: To estimate the relative abundances of transcripts, Cufflinks software was used (version 2.2.1) (Trapnell et al., 2013). FPKM values (Fragments Per Kilobase of exon per Million fragments mapped) were calculated per individual gene.
RNA-seq: Differential RNA expression across cohorts was assessed using EdgeR (version 3.20.9) and limma (version 3.34.9) software (Phipson et al., 2016; Ritchie et al., 2015; Robinson et al., 2010). Specifically, raw counts of all replicate samples in the pMN, iNkx2.2, and iTNmut groups were analyzed together, and all genes which did not have a count of 10 in at least two samples were excluded and considered to be not expressed. Read counts in each replicate were normalized to their new respective library size, and the data were normalized using the trimmed mean of M-values (TMM) method with the calcNormFactors function (Robinson and Oshlack, 2010).
RNA-seq: Significant differences were then determined in EdgeR/limma through linear modeling with precision weights (Law et al., 2014) and were based on adjusted p-values obtained through Benjamini and Hochberg’s method to control false discovery rate set at 5% (Benjamini and Hochberg, 1995).
Supplementary_files_format_and_content: RNA-seq: raw readcount text file used as input into EdgeR/limma
Supplementary_files_format_and_content: RNA-seq: test file of results of EdgeR/limma analysis
Supplementary_files_format_and_content: RNA-seq: text file of abundance measurements (FPKM values)
ChIP-seq: All sequencing data sets were aligned using Bowtie (version 0.12.7) (Langmead et al., 2009) to the mm10 genome assembly.
ChIP-seq: Peak calling of Sox2, Flag, Grg, Med1, and p300 was carried out using the GEM algorithm (Guo et al., 2012) with multi-condition settings to simultaneously identify all regions enriched over background in pMN, iNkx2.2, and iTNmut progenitors. Peak calling of H3K27ac, H3K4me1, H3K27me3, and HDAC2 was performed with the MACS software (“--- broad”; version2.0.9) (Zhang et al., 2008).
ChIP-seq: Read-count intensity was obtained within +/- 500bp of all GEM identified peaks and +/- 1kb around the midpoints of all MACS-defined regions. Read counts of individual replicates were then normalized to their respective sequencing library size, and only loci above a given threshold in both replicates (when n = 2) were used for subsequent analysis.
ChIP-seq: GEM peaks were then further processed as follows. After filtering out very low-occupied regions, each ChIP-seq dataset (e.g. Grg) contained loci that were bound in at least one of the three cell types (e.g. iNkx2.2, iTNmut, and pMN progenitors). If multiple ChIP-seq peaks resided within 500bp of each other in a given cell type, they were merged into one peak using Bedtools (version 2.17.0) (Quinlan, 2010). The locus with the highest read density (+/- 100bp) was identified using Tidyverse (version 1.2.1) (Wickham H., 2019) and utilized as the center of the occupied region. Read counts for all replicates across the three cell types were obtained within +/- 500bp from the center of all merged peaks as defined above, regardless of their cell-type origin. These datasets were then imported into the EdgeR (version 3.20.9) and limma (version 3.34.9) software packages (Phipson et al., 2016; Ritchie et al., 2015; Robinson et al., 2010) to globally identify all significantly altered occupancy regions among pMN, iNkx2.2, and iTNmut progenitors for ChIP-seq datasets with an n = 2. Read counts in each replicate were normalized to their new respective library size and additionally quantile normalized to account for variations in ChIP-seq quality using the voom function (Hansen et al., 2012; Law et al., 2014). Significant differences in occupancy were then determined through linear modeling with precision weights (Law et al., 2014) and were based on adjusted p-values obtained through Benjamini and Hochberg’s method to control false discovery rate set at 5% (Benjamini and Hochberg, 1995).
ChIP-seq: MACS-defined peaks were similarly processed. After filtering out very low-occupied regions, the remaining peaks in each individual replicate were merged into one peak if they resided within 1kb of each other. Peaks that overlapped with each other in both replicates were then further merged into one peak to create the dataset for each cell type (e.g. pMN H3K27ac) and the midpoint of the newly merged peak was defined as the center. To globally identify all differential occupancy loci as described above for GEM-defined regions, peaks that overlapped among the three cell types (e.g. H3K27ac sites overlapping in pMN, iNkx2.2, iTNmut) were merged into one peak and the midpoint of the newly merged peak was defined as the center. These shared peaks and the remaining cell-type specific peaks were combined into one dataset, and read counts for all replicates across the three cell types were obtained within +/- 1kb from the center of the peaks regardless of their cell-type origin (e.g. pMN H3K27ac ChIP-seq signal was recorded at not only pMN-specific and shared coordinates but also iNkx2.2- and iTNmut-specific loci). These datasets were then further analyzed in EdgeR (version 3.20.9) and limma (version 3.34.9) software packages (Phipson et al., 2016; Ritchie et al., 2015; Robinson et al., 2010), as described above for GEM-identified loci.
ChIP-seq: Final readcounts for all peaks of each ChIP-seq were converted to counts per million (CPM), quantile-normalized, and log2 transformed (logCPM). These logCPM values were then extracted to calculate significant differences in occupancy at specific regions (e.g. subsets of enhancers bound by Nkx2.2). Significance was determined using the Mann-Whitney test or the Kruskal-Wallis test with Dunn's post hoc analysis.
Genome_build: mm10
Supplementary_files_format_and_content: ChIP-seq: peak text files generated by GEM or MACS software
Supplementary_files_format_and_content: ChIP-seq: raw readcount text files for peaks used as input into EdgeR/limma software and to make logCPM files
Supplementary_files_format_and_content: ChIP-seq: text file of results of EdgeR/limma analysis
Supplementary_files_format_and_content: ChIP-seq: logCPM text files for all peaks used in EdgeR/limma analysis
|
| |
|
| Submission date |
Jul 27, 2020 |
| Last update date |
Jul 24, 2023 |
| Contact name |
Elena Abarinov |
| E-mail(s) |
elenavabarinov@gmail.com
|
| Organization name |
University of Colorado at Denver
|
| Street address |
1775 Aurora Ct
|
| City |
Aurora |
| State/province |
Colorado |
| ZIP/Postal code |
80045 |
| Country |
USA |
| |
|
| Platform ID |
GPL13112 |
| Series (1) |
| GSE155229 |
Genome-wide analysis of enhancers repressed by Nkx2.2 in spinal motor neuron progenitors |
|
| Relations |
| BioSample |
SAMN15654869 |
| SRA |
SRX8833474 |
| Supplementary file |
Size |
Download |
File type/resource |
| GSM4697766_iNkx2.2_HDAC2_peaks_rep1.txt.gz |
716.3 Kb |
(ftp)(http) |
TXT |
SRA Run Selector |
| Processed data provided as supplementary file |
| Raw data are available in SRA |
| Processed data are available on Series record |
|
|
|
|
 |
