TSS in the chromosome and plasmids pSYSA, G, M, and X of Synechocystis 6803. (A) (Upper) Three closely spaced TSS are found in the region upstream of trxA/slr0623, encoding one of four thioredoxins in Synechocystis 6803, illustrated by the 199 primary (+) sequencing reads starting in this region. These reads start at position −49 (TSS1, seven reads), −38/−36 (TSS2, 184 reads), and −32 (TSS3, eight reads). (Lower) All three TSS exhibit a reliable score for a −10 element at a distance of six plus or minus one nucleotides from the first transcribed nucleotide. TSS2 and TSS3 correspond to two previously detected major primer extension products (). The total number of sequencing reads from the (+) and (−) cDNA libraries is shown in light gray along the length of the gene in the upper panel. (B) Details of annotation and classification of 3,527 TSS into 1,165 gTSS giving rise to mRNA, 1,112 aTSS producing asRNA, 821 iTSS for internal sense transcripts, and 429 nTSS for candidate ncRNAs. Case a: A TSS was classified as gTSS if the TSS was located 0–100 nt upstream of an ORF or if at least 1 of the 454 reads reached into the coding sequence or the 10 nt in front of it. Case b: The TSS is located antisense to an annotated gene or within ≤50 bp of its 5′ or 3′ UTR. Case c: The TSS is positioned within an annotated sequence. Case d: Putative ncRNA (nTSS). (C) Overlaps between different categories of TSS. Many TSS associate with multiple categories according to . Thus, 102 of the 1,165 gTSS actually are located antisense, and 62 are located in sense orientation within another annotated gene. One example is the slr1470 mRNA starting from an iTSS within rnpA (slr1469). We also observed 21 aTSS that were located gene-internally because of overlapping transcription. An example is the aTSS for the asRNA to slr0320, which starts internally within the rpoD (sll0306) coding sequence. None of the 429 nTSS overlapped one of the other three categories; therefore these TSS are not shown in the diagram.

Occurrence of 3,213 TSS along a linear plot of the Synechocystis 6803 chromosome. The genome position is drawn along the x axis and is given in nucleotides. Mapped TSS for the forward strand are plotted above the x axis, and mapped TSS for the reverse strand are plotted below the x axis. The number of sequence reads is given as a proxy for gene expression on the y axis (logarithmic scaling). The location of each of the TSS according to served for classification of the respective TSS as gTSS from which an mRNA would originate (blue), nTSS for a putative ncRNA (green), aTSS for antisense transcripts (red), or iTSS (gray).

Gene-internally located TSS (iTSS). (A) Example for the reannotation of start codons based on a newly discovered TSS. (Upper) Reads of (+) and (−) libraries mapping to the rfbB (slr0809) region are shown according to . (Lower) An alignment of rfbB sequences (5′ region) from the cyanobacteria Microcystis aeruginosa NIES 843, Synechococcus elongatus PCC7942, Cyanothece sp. ATCC 51142, and Synechocystis 6803 is shown. The here discovered TSS is displayed by an arrow downstream of the original start codon (boxed in red). The highly conserved reannotated AUG and the possible Shine–Dalgarno element are boxed in green. The −10 box of the new TSS is boxed in purple. (B) Distribution of 732 iTSS relative to their positions within the reading frames in which they are located. The number of iTSS belonging to each percentile ia plotted along the y axis, and the nucleotide positions are plotted along the x axis (in %). (C) The ntcA gene and its four mapped TSS. The light blue box shows the region from which a very abundant sense transcript originates. Microarray probes are indicated by short vertical tabs linked by short colored lines, and their expression values are plotted on the left y axis. (D) Northern hybridization. RNA was extracted from cultures grown in a time-course experiment following the removal of nitrate (in hours). The induction of the ntcA full-length mRNA (arrow 1) at 4 h can be reversed by the readdition of nitrate (4r). The sense transcripts originating at the iTSS at +249 accumulate in the form of two dominant bands (arrows 2). (E) Comparison of dRNA-seq data with previous tiling microarray analysis () validates the accumulation of short sense transcripts from iTSS within genes encoding the glycolate oxidase subunit F (glcF), the ferrichrome-iron receptor (fhuA), or the transketolase (tktA). The gene regions from which short sense transcripts originate are drawn in light blue; confirmed asRNAs are shown in red. The normalized expression values (left y axis) are plotted for microarray probes (black dots). Read numbers for the (+) and (−) libraries are plotted in dark and light gray, respectively (log₂ scale on right y axis).

The hydrogenase (hox) operon is framed by two highly expressed ncRNAs, SyR1 () and ncr0700. (A) Both strands are shown with the location of annotated genes (blue boxes), 5′ UTRs (light gray), internal sense RNAs (light blue), asRNAs (red), and intergenic ncRNA genes (yellow). The TSS of the hox operon () is indicated by the black arrow. The read numbers for the enriched library (+) are plotted in dark gray, and reads for the untreated library (−) are in light gray and are given in log₂ scale (right y axis). The normalized log₂ expression values of four different microarray experiments are plotted in blue (control), black (dark incubation), yellow (incubation at high light), and green (CO₂ depletion). The scale for the microarray data are given at the left y axis. All probes of a single RNA feature are connected by lines. (B) Mean expression (Upper) and mean fold changes (FC) of the two short ncRNAs SyR1 and ncr0700 (Lower Right) under control conditions (co), dark incubation, incubation at high light (HL), and CO₂ depletion (−CO₂). (C) The SyR1 RNA was overexpressed under control of the petJ promoter, causing a severe phenotype with a loss of pigmentation (Left, cuvettes; Right, whole-cell absorption spectra).