Skip to main content
Fig. 2 | BMC Biology

Fig. 2

From: Low-cell-number, single-tube amplification (STA) of total RNA revealed transcriptome changes from pluripotency to endothelium

Fig. 2

High-throughput profiling of hPSC and 293-cell transcriptomes based on STA. a Representative denaturing PAGE (6%) of the preamplified cDNA from 0 (left) and 100 (right) hESCs. One hundred hESCs were sorted directly into 2 μl of lysis buffer and subject to STA. All 15-cycle preamplified products were purified with PEG/NaCl and electrophoresed. Two different widths (N: narrow; W: wide) of gel slices were cut for library preparation. The 21-cycle preamplified and purified products of lysis buffer alone served as a no-cell control. b Gene body coverage chart of all aligned reads against all transcripts based on GENCODE v22. c The distribution of reads among genomic features. Strandedness was taken into account for counting in the order of exon (exon, GENCODE v22), intron (transcript minus exon, GENCODE v22), tRNA (evidence-based annotation of tRNA, GENCODE v22), and repeat (GRCh38-based RepeatMasker track on UCSC Genome Browser) features, and the rest was considered to be unannotated. The numbers on top of each column indicate reads successfully aligned to the GRCh38 genome assembly. The percentages in parentheses are the numbers of successfully aligned reads divided by the total reads in the respective libraries. d The RNA-type summary of the exon reads in the three libraries based on GENCODE v22. The percentages indicate the numbers of reads divided by the number of exon reads. e Upper: Venn diagram showing the overlap of the miRNAs identified (values >0) in TP100N and two other hESC reference sequencing data sources (Solexa and 454) in the literature. Lower: Scatter plot demonstrating the association of miRNA quantifications between TP100N and the referenced Solexa data. Each value of the reference sample was multiplied by 10,000 because only normalized data were available with the reference samples. Blue, green, and red colors indicate the top 10 most highly expressed miRNAs in TP100N, reference sample, and both, respectively. f Scatter plot demonstrating the association of rlog-normalized (rld) miRNA quantifications of 293 cells between a library prepared with STA (293FT M) and that with conventional ligation-based method (SRX763661). Identical genome alignment (GRCh38) and feature assignment (GENCODE v22) were performed, and all miRNAs (raw counts >0) were used for rlog transformation (blind = TRUE) with DESeq2. g Upper: Venn diagram showing the overlap of protein-coding RNAs identified (values >0 in each sample) in 293FTM and those in reference RNA-Seq data (SRX475877) from 293 cells. Lower: Scatter plot demonstrating the association of rlog-normalized (rld) protein coding-RNA quantifications of 293 cells between a library prepared with STA (293FTM) and reference SRX475877 RNA-Seq data. Identical genome alignment (GRCh38) and feature assignment (GENCODE v22) were performed, and all protein-coding RNAs (raw counts >0) were used for rlog transformation (blind = TRUE) with DESeq2. Representative histone-coding RNAs enriched in the STA-based library are highlighted in red. qPCR (h) and denaturing PAGE (12%) of the semi-quantitative PCR products (i) of miR-302c, miR-367, miR-let7a, and miR-196b. hiPSC (DF19-9-7T) and 293FT cells were lysed, and 100-, 10-, and 1-cell equivalents of lysates were used for STA. Preamplification of 15, 18, and 21 cycles were used for 100-, 10-, and 1-cell lysates, respectively, and 1/500 of the preamplified and purified eluents was used for qPCR. The optimal cycle numbers (threshold cycle plus 4) for semi-quantitative PCR in i were based on the qPCR analysis in h. RNAU6 served as a loading control

Back to article page