Cappable-seq short RNA sequencing

15 ml cell culture was rapidly mixed with 30 ml pre-cooled RNAprotect Bacteria Reagent (Qiagen) placed in an ice bath. Cells were harvested by centrifugation (5 min at 4000 × g at 4 °C). Pellets were immediately subjected to RNA isolation using the mirVana miRNA isolation kit (Ambion) with an initial resuspension buffer volume of 200 µl following the protocol for small RNAs (20-200 nt length). Library preparation and deep sequencing was conducted at Vertis Biotechnologie (Germany). In brief, 5′-triphosporylated RNA was capped with 3′-desthiobiotin-TEG-GTP (NEB)) using the Vaccinia virus Capping enzyme (NEB) and biotinylated RNA species were subsequently enriched by affinity purification using streptavidin beads yielding 0.6–1.3% of the sRNA preparation. The eluted RNA was poly-adenylated using E. coli Poly(A) polymerase and 5′-ends were converted to mono-phosphates by incubation with RNA 5′ Pyrophosphohydrolase (NEB). Subsequently, an RNA adaptor (5′-ACACTCTTTCCCTACACGACGCTCTTCCGATCT-3′) was ligated to the newly formed 5′-monophosphate structures. First-strand cDNA synthesis was performed using an oligo(dT)-adaptor primer and the M-MLV reverse transcriptase at 42 °C. The resulting cDNA was finally PCR-amplified (12 cycles) with TruSeq Dual Index sequencing primers (Illumina) and Herculase II Fusion DNA Polymerase (Agilent). The libraries were sequenced on an Illumina NextSeq 500 system with 75 bp read length. In order to remove poly(A)-tails and adaptors, we trimmed the reads using Cutadapt⁹³ in two rounds with the following settings to prevent trimming of naturally occurring A-rich RNAs due to the low GC-content of the S. solfataricus genome: (i) -a “{A15}” -e 0 -m 15 to remove all poly(A) stretches of at least 15 nt length plus downstream regions and (ii) -a “A{15}”X -e 0 -O 5 to terminal shorter poly(A) stretches of minimum 5 nt length. Trimmed and untrimmed reads were split into separate fastq files using awk. Both fastq files were aligned to the S. solfataricus genome using bowtie v1.2.2⁷⁶ (parameters -v 1 -m 1 –-best –-strata -S) with untrimmed 75 nt reads shortened to 71 nt (−3 4). The bam file output was merged, sorted and indexed using SAMtools⁷⁷. Bam files were imported into the R environment using the rsamtools and GenomicRanges packages and filtered to ensure a unique sequence of the initial 20 bp within the S. solfataricus genome required to map the reads in R using the Biostrings package. TSS-RNA were defined as RNAs with a 5′-end within 20 nt of a mapped or predicted TSS. The two biological replicates showed good reproducibility of TSS-RNA occupancy with a Spearman correlation of 0.98 for 438 mappable promoters.

To calculate the fraction of TSS-RNAs with a length shorter than 50 nt for each TU, a minimum read count of 10 TSS-RNAs per TU per replicate was used and values were averaged between the two biological replicates.