GEO Accession viewer

NCBI > GEO > Accession Display

Not logged in | Login

GEO help: Mouse over screen elements for information.

Sample GSM2948019

Query DataSets for GSM2948019

Status

Public on Dec 25, 2018

Title

HDL_SRBIKO_13

Sample type

SRA

Source name

HDL_SRBIKO

Organism

Mus musculus

Characteristics

strain background: C57BL/6
genotype/variation: SRBIKO; Srbi-/-
animal id#: 13
tissue: HDL
barcode: 40

Growth protocol

Conventional Housing

Extracted molecule

total RNA

Extraction protocol

Plasma, basal bile, urine, and livers were collected from wild-type (WT) and SR-BI-deficient (B6;129S2-Scarb1tm1Kri/J, SR-BI KO) mice. Mice were anesthetized with urethane (1g/kg, i.p.). The common bile duct was ligated and the gall bladder cannulated to divert bile into collection tubes. Basal bile was collected for a period of 30 min. Mice were then exsanguinated, blood was collected from the abdominal aorta in EDTA coated tubes and placed on wet ice, and tissues were dissected and snap frozen in liquid nitrogen. Plasma and tissues were stored at -80oC prior to analysis. All animal procedures were completed under active and approved IACUC protocols. To separate HDL and apolipoprotein B (APOB)-containing lipoproteins from mouse plasma, 200 µL of 0.22-µm filtered-plasma samples were diluted to 500 µL in size-exclusion chromatography (SEC) running buffer (10 mM Tris-HCl, 0.15 M NaCl, 0.2% NaN3) and injected an ÄKTA SEC system (GE Healthcare) with three in-series Superdex-200 Increase gel filtration columns (10/300 GL; GE Healthcare). Samples were applied to the column with a flow rate of 0.3 mL/min at room temperature and eluate collected as 72 x 1.5 mL fractions using a F9-C 96-well plate fraction collector (GE Healthcare). Each fraction was analyzed for total protein (BCA; Pierce), total cholesterol (Raichem), and triglycerides (Raichem) to identify fractions corresponding with HDL and APOB particles. Due to the SEC set-up, we were not able to separate VLDL from LDL particles, and thus, we collected fractions covering both lipoprotein classes, referred to here as APOB. Fractions corresponding with each lipoprotein group were pooled, concentrated with Amicon Ultra-4 10 kDa centrifugal filters (Millipore) to <200 µL volume, and protein concentrations were quantified by BCA assays (Pierce). Based on the distribution of total cholesterol, triglycerides, and protein, fractions corresponding to HDL and APOB were collected, pooled, and concentrated. To differentiate lipoprotein sRNA signatures from liver and biofluids, and determine the impact of SR-BI-deficiency, samples were collected from Scarb1-/- (SR-BI KO) and wild-type (WT) mice. Total RNA was extracted from HDL (WT N=7, SR-BI KO N=7) and APOB (WT N=7, SR-BI KO N=7) particles, as well as livers (WT N=7, SR-BI KO N=7), bile (WT N=7, SR-BI KO N=6), and urine (WT N=5, SR-BI KO N=6). RNA was isolated from equal inputs of either bile (volume), liver (mg), HDL (protein) or APOB (protein) using miRNAEasy Mini kits (Qiagen). Specifically, 30 µL of primary bile, 120 μg of APOB, 180 μg of HDL or 20 mg of liver were added to 1 mL of Qiazol. Livers were homogenized in Qiazol with High-Impact Zirconium beads using a Bead Bug Homogenizer (Benchmark Scientific). After removal of beads, subsequent steps for liver RNA extraction were followed according to manufacturer’s protocol. Bile, APOB and HDL RNA isolations were processed according to manufacturer’s protocol, except that after addition of ethanol, samples were incubated at -80oC overnight before application to isolation columns, and were eluted with a volume of 50 μL. Liver RNA samples were quantified by Take3 plates (BioTek).
NEXTflex Small RNA Library Preparation Kits v3 for Illumina® Platforms (BioO Scientific) were used to generate cDNA libraries for sRNA-seq. Briefly, 1 µg of liver total RNA was used as input for adapter ligation, as per manufacturer’s protocol. For bile, APOB and HDL RNA, 10.5 μL (21%) of the RNA isolation eluate was used as input for adapter ligation. Library generation was performed according to manufacturer’s protocol (BioO Scientific) with a modification to the amplification step, as liver libraries received 18 cycles and bile, APOB and HDL libraries received 25 cycles. After amplification, samples were size-selected using a Pippin-Prep (Sage Science) -- set for a range of 135-200 nts in length -- and subsequently purified and concentrated using DNA Clean and Concentrator 5 kit (Zymo). Individual libraries were then screened for quality by High-Sensitivity DNA chips using a 2100 Bioanalyzer (Agilent) and quantified using High-Sensitivity DNA assays with Qubit (Life Technologies). Equal concentrations of all individual libraries were pooled for multiplex sequencing runs, and concentrated using DNA Clean and Concentrator 5 kit (Zymo). For rigor in down-stream comparisons, all 66 sequencing libraries were randomized and run independently on three individual sequencing lanes. Single-end sequencing (75 cycles) of multiplexed libraries were performed on an Illumina NextSEQ 500 at the Vanderbilt Technologies for Advanced Genomics (VANTAGE) core laboratory. Each library was sequenced at an average depth of 16.28 million reads/sample.

Library strategy

ncRNA-Seq

Library source

transcriptomic

Library selection

size fractionation

Instrument model

Illumina NextSeq 500

Description

Small RNA Seq-fastq, demultiplexed, NextFlex V3
processed data file:
ENV_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
Fungus_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
HMB_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
ex_rRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
ex_tRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
isomiR_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
miRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
othersRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
rRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
snoRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
snRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
tRNA_HDL_SRBIKO_vs_WT_detectedInBothGroup_min5_DESeq2.csv
bacteria_group1_pm_KCV_3018_77_78_79.category.count.txt
bacteria_group1_pm_KCV_3018_77_78_79.read.count.txt
bacteria_group2_pm_KCV_3018_77_78_79.category.count.txt
bacteria_group2_pm_KCV_3018_77_78_79.read.count.txt
fungus_group4_pm_KCV_3018_77_78_79.category.count.txt
fungus_group4_pm_KCV_3018_77_78_79.read.count.txt
KCV_3018_77_78_79.Category.Table.csv
rRNA_pm_KCV_3018_77_78_79.count.txt
tRNA_pm_KCV_3018_77_78_79.read.count.txt
tRNA_pm_KCV_3018_77_78_79.category.count.txt
tRNA_pm_KCV_3018_77_78_79.count.txt
miRBase_pm_KCV_3018_77_78_79.count.txt
miRBase_pm_KCV_3018_77_78_79.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.count.txt
smallRNA_1mm_KCV_3018_77_78_79.miRNA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.miRNA.isomiR.count.txt
smallRNA_1mm_KCV_3018_77_78_79.miRNA.NTA.base.count.txt
smallRNA_1mm_KCV_3018_77_78_79.miRNA.NTA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.miRNA.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.other.count.txt
smallRNA_1mm_KCV_3018_77_78_79.other.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.rRNA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.rRNA.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.snoRNA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.snoRNA.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.snRNA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.snRNA.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.tRNA.aminoacid.count.txt
smallRNA_1mm_KCV_3018_77_78_79.tRNA.aminoacid.read.count.txt
smallRNA_1mm_KCV_3018_77_78_79.tRNA.count.txt
smallRNA_1mm_KCV_3018_77_78_79.tRNA.NTA.count.txt
rRNA_pm_KCV_3018_77_78_79.read.count.txt

Data processing

Analysis Pipeline: TIGER
https://www.biorxiv.org/content/early/2018/01/23/246900
Step 1: Pre-processing
To assess raw data quality, FastQC was performed at the raw read level to check for base quality, total read counts, and adapter identification. Cutadapt was then used to trim 3’ adapters from processed reads (-a TGGAATTCTCGGGTGCCAAGG). Cutadapt was then used to remove the first and last 4 bases from the trimmed reads and all trimmed reads <16 nts in length were removed (-m 16 -u 4 -u -4). After trimming, read length distributions were plotted and FastQC was performed on trimmed reads to validate the efficiency of adapter trimming. To generate identical read files, trimmed reads in each sample were collapsed into non-redundant “identical” reads in FASTQ format and copy numbers were recorded for downstream analysis.
Step 2: Host-Genome Alignment (Mouse)
In the Host Genome & Database alignment module, bowtie (v1.1.2) was used to map reads to a costumed database with option (-a -m 100 --best - strata -v 1) which allows 1 mismatch (MM) and 100 multi-mapped loci, and only the best matches were recorded. The costumed database was constructed by the host genome and known sequences of host mature transcripts curated in specific library databases – tRNAs (http://gtrnadb.ucsc.edu/GtRNAdb2/) and rRNA (http://archive.broadinstitute.org/cancer/cga/rnaseqc_download). Counting and differential expression analysis of miRNAs, tDRs, rDRs, snDRs, snoDRs, and other miscellaneous sRNAs (miscRNA), including yDRs and lincDRs, were performed. All prepossessed quality reads were assigned to different classes of annotated sRNAs using distinct rules -- miRNA: 1 MM, ≥16nt, offset -2, -1, 0, 1, 2 and tDR, snDRs, snoDRs, yDRs, and miscRNAs: 1 MM, ≥16nt, overlap ≥0.9 overlap. Based on the extensive genomic coverage of lncRNAs and repetitive elements and conservation of rRNAs, the TIGER pipeline applies more stringent assignment rules for lncDRs and rDRs – perfect match, ≥20 nt, and ≥90% overlap with parent lncRNAs or rRNAs. Furthermore, reads assigned to lncDRs must only be aligned to lncRNA coordinates and not to any other loci in the genome. All reads ≥20 nts in length and not aligned to the costumed database were extracted and tested for alignment as non-host reads. Differential expression of tabulated read counts were performed by DEseq2.
Step 3: Non-Host Alignment Module
Non-host reads were then analyzed using the Non-Host Genome and Non-Host Library modules in parallel. In the Non-Host Genome module, reads were aligned in parallel to two collections of bacterial genomes: a human microbiome (HMB) collection and a hand-curated list of environmental bacteria observed during sequencing of human and mouse lipoproteins. Due to high conservation between bacterial genomes and multi-mapping issues, a different bowtie option (-a -m 1000 --best -strata -v 0) was used which allowed perfect match only and 1000 multi-mapped loci. Reads were aligned to the HMB, ENV, and fungal groups in parallel and, thus, the same reads could have been counted in multiple groups. The fraction of reads that align to both databases (HMB, ENV) and the reads that are unique to specific databases were plotted. Differential expression and high-end analyses, as described above, were performed at the genome level (total normalized read count for each genome) and at the individual read level. In parallel, non-host reads were also analyzed by the Non-Host Library (Gold) module where they were aligned to non-coding RNA databases with same bowtie option as non-host genome analysis. To identify possible non-host miRNAs (xenomiRs) in sRNA-seq datasets, all non-host reads were aligned perfectly to annotated miRNAs in miRBase (miRBase.org) and tabulated. Similarly, non-host reads were aligned to all tRNAs in the GtRNAdb database (GtRNAdb2). Extensive categorical analysis of parent non-host tRNAs were performed at the kingdom, genome (species), amino acid, anti-codon, and fragment (read) levels. All assigned non-host tDRs underwent differential expression analysis, high-end analysis, and data visualization, as described above. Non-host reads were also aligned to prokaryotic and eukaryotic rRNA transcripts in SILVA database (https://www.arb-silva.de). TIGER limits the analysis of non-host rDR to the kingdom level for counting, differential expression analysis and high- end analysis.
Genome_build: mm9
Supplementary_files_format_and_content:
Identical.fastq files are fastq files that contain all reads that survive processing (Step 1),
Count files are comma separated value spreadsheets that list all annotated features (.count.txt) or fragment reads (read.count.txt) or total categorical counts (category.count.txt) for associated with a given category
CSV files are outputs from differential expression analysis using DeSeq2 comparing Wild-type vs SRBIKO

Submission date

Jan 25, 2018

Last update date

Dec 25, 2018

Contact name

Ryan M Allen

E-mail(s)

[email protected]

Phone

3146109328

Organization name

Vanderbilt University Medical Center