Fragment analyzer

DNA was extracted from plasma and urine using the QIAamp Circulating Nucleic Acid Kit according to the manufacturer’s instructions (Qiagen, Cat #55114). Sequencing libraries were prepared using a single-stranded library preparation as described in Burnham et al.^{13 (link)}. DNA was extracted from oral swab samples using the QIAamp UCP Pathogen Kit according to the manufacturer’s instructions (Qiagen, Cat #50214) and libraries were prepared using the Nextera XT DNA Library Prep Kit (Illumina, Cat #FC-131-1024). All libraries were characterized using the AATI Fragment Analyzer before pooling and sequencing on the Illumina NextSeq 500 platform (paired-end, 2 × 75 bp).

A detailed step-by-step protocol has been deposited in the protocols.io repository (Gressel et al. 2019a ). TT-seq and RNA-seq were performed in 2 biological replicates including RNA spike-ins. Briefly, experiments were performed using $5\times {10}^7$ K562 cells per biological replicate. Cells were kept at optimal growth conditions and supplemented with 5 µg mL⁻¹ of α-amanitin or solvent (water) for 8 h. After 7 h 55 min, a 4-thiouridine (4sU) labeling pulse (Sigma-Aldrich, T4509) was applied for 5 min using 500 µM (see Supplementary material S3). Total RNA was isolated with the QIAzol reagent (# 79306) according to manufacturer’s instructions except for the addition of 150 ng of RNA spike-in pool with QIAzol reagent as previously described (Schwalb et al. 2016 (link); Gressel et al. 2019a ). The Ovation Universal RNA-Seq System (NuGEN) was used for strand-specific library preparation as described (Gressel et al. 2019b (link)). Purified cDNA libraries were analyzed by Fragment Analyzer prior to Illumina sequencing. Sequencing was performed on a HiSeq 2500 (Illumina) in paired-end mode with 50-bp read length.

ChIP assay was performed using 1 Day ChIP kit and Shearing ChIP kit (Diagenode, Denville, USA) according to the manufacturer’s protocol. After lentiviral transduction and cross-linking, mouse kidney fibroblasts were fixed with formaldehyde and the DNA was sheared into small fragments. After incubation with a Myc-tag antibody (Cell signaling, Beverly, USA), the pulled-down complexes were de-crosslinked and treated with proteinase K. The purified DNA samples were proceeded further for library preparation by TruSeq RNA Library Prep Kit v2 (Illumina, USA) and quality control and library validation were performed by Fragment Analyzer and Kapa PCR (Illumina, USA), respectively. The fragments were sequenced by an Illumina HiSeq4000 instrument. For data analysis, a previous established protocol^{48 (link)} was followed. Briefly, the raw reads from two independent biological replicates were first concatenated and then peak calling was performed with MACSII in order to obtain the count numbers. We used settings for narrow peaks (200 bp window size, 200 bp gap size, and false discovery rate of 0.01) in all cases.

Reduced representation bisulphite sequencing libraries were constructed with 200 ng input DNA from each sample (n = 24; Table 1) using the Premium RRBS Kit (Diagenode, Cat. No. C02030032) according to the manufacturer’s instructions. Unmethylated and methylated spike‐in control sequences are included within the commercial kit to allow for assessment of bisulphite conversion efficiency, which in this experiment exceeded 99% for all samples (mean ± SD = 99.57 ± 0.25). Size selection and clean‐up steps were performed with AMPure XP Beads (Beckman Coulter), and bisulphite‐converted libraries were subjected to 13–15 amplification cycles in the final PCR (polymerase chain reaction). Resulting libraries were stored at −80°C prior to submission to the Georgia Genomics and Bioinformatics Core (GGBC; University of Georgia, Athens, GA) for sequencing. Quality of the libraries was assessed using an Agilent Fragment Analyzer, and pooled libraries were sequenced on a single Illumina NextSeq 500 High Output flowcell (75 bp, single‐end) with 20% Illumina PhiX control. Sequencing generated a total of 300 million reads, with 94% meeting or exceeding the Phred score quality threshold of >30. Raw reads were assessed for quality using fastqc (version 0.11.8; https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and multiqc (version 1.8; Ewels et al., 2016 (link)).

Isolated DNA from both FFPE blocks and peripheral blood was prepared for whole exome sequencing (WES) sequencing with commercial kits for 100 bp paired-end read lengths. The RNA isolated from FFPE blocks was used for RNA library preparation for paired-end sequencing with read lengths of 100 bp. The quality of prepared libraries was checked with Fragment Analyzer and sequenced on an Illumina sequencer with expected depth for DNA of 150x and 300x for PBMC and FFPE samples, respectively. RNA sequencing was set to obtain at least 100 million reads for each sample.