The coding and noncoding architecture of the Caulobacter crescentus genome

Abstract

Caulobacter crescentus undergoes an asymmetric cell division controlled by a genetic circuit that cycles in space and time. We provide a universal strategy for defining the coding potential of bacterial genomes by applying ribosome profiling, RNA-seq, global 5'-RACE, and liquid chromatography coupled with tandem mass spectrometry (LC-MS) data to the 4-megabase C. crescentus genome. We mapped transcript units at single base-pair resolution using RNA-seq together with global 5'-RACE. Additionally, using ribosome profiling and LC-MS, we mapped translation start sites and coding regions with near complete coverage. We found most start codons lacked corresponding Shine-Dalgarno sites although ribosomes were observed to pause at internal Shine-Dalgarno sites within the coding DNA sequence (CDS). These data suggest a more prevalent use of the Shine-Dalgarno sequence for ribosome pausing rather than translation initiation in C. crescentus. Overall 19% of the transcribed and translated genomic elements were newly identified or significantly improved by this approach, providing a valuable genomic resource to elucidate the complete C. crescentus genetic circuitry that controls asymmetric cell division.

Figure 1. Genome-wide data set integration to map the genetic elements in the C. crescentus genome.

5′PPP transcription start site (TSS) (Zhou et al. [unpublished data]) (red spheres with black bar), RNA-seq density (black bars), ribosome footprints (green ribosomes), and LC-MS peptide coverage (yellow bars) shown for a single gene (CCNA_00515) between 528700 and 532200 bp. 5′PPP data generated from Tobacco Acid Pyrophosphate (TAP) enriched 5′ global RACE. 5′ PPP fragments plotted with Y-axis scale in #reads. Base hydrolyzed (-OH) RNA-seq data plotted with Y-axis scale in log(#reads+1). Micrococcal nuclease (MNase) protected ribosome footprints plotted with Y-axis scale in log(#reads+1). LC-MS-identified tryptic peptides are mapped onto their respective positions of the CDS with potential ribosomal initiated N-terminal residue in purple. Respective genomic features are highlighted including transcriptional start site (TSS), 5′ untranslated region (UTR), coding region (blue bar), and 3′ UTR of the expressed element. Y-axis scales are similar in all subsequent figures.

Figure 2. Mapping the C. crescentus coding DNA sequence architecture.

A. Mapping of the correct start codon for ftsA. Shown below are phase contrast images of cells containing a high copy plasmid with either the newly identified start codon (pBXftsA) or the old start codon (pBX-ftsAΔ_N1-18) grown in M2G before and after induction with xylose for 6 hours. Scale bar is 6.05 µm. B. Putative small leader CDS identified on the trpS mRNA. C. Ribosome profiling identification of an intergenic small CDS. D. Alternative translation initiation site identified in the CCNA_02448 mRNA allows translation of two in-frame protein isoforms. Internal start codon was verified by LC-MS.

Figure 3. Role of the Shine-Dalgarno sequence in translation initiation and pausing.

A. Global lack of SD sites in front of start codons. mRNA affinity to the aSD site on the ribosome was calculated using the Free2bind package . −4.4 kcal/mol is the cutoff for SD identification based on the predicted annealing between the aSD and translation initiation region as in . B. Translation initiation site motif derived from all start codons in the genome generated in MEME . C. Ribosome occupancy profiles reveal pausing identified by peaks of ribosomes above the average read density (black line). Stronger pauses are shown by a larger peak height. Y-axis value is #reads on linear scale. D. Plot of the normalized cross-correlation function between pauses in the ribosome occupancy profiles and the presence of SD sequences. The plot is centered at the A-site of ribosome pauses and the peak of correlation occurs in proximity to the aSD site on the ribosome. E. Plot of SD binding affinity for the aSD compared to the occupancy of ribosomes translating them. r is the correlation coefficient.

Figure 4. Transcription start site and RNA-seq-derived transcript architecture reveals mRNA complexity.

A. Metagene plot of the normalized RNA-seq reads centered on the 5′ PPP sites identified by 5′ global RACE (Zhou et al. [unpublished data]). RNA-seq reads are mapped to the 5′ nucleotide with an enriched peak resulting from partial shearing of the RNA . B. Global distribution of 5′ UTR lengths for all C. crescentus mRNAs with identified 5′ ends shown in blue with Y-axis scale on the left. Cumulative distribution of 5′ UTRs less than 200 nt shown in red with Y-axis scale the right. C. Leaderless dnaQ mRNA, where transcription is initiated at the 5′ nucleotide of the initiating ATG. D. Long 5′ UTR (150 nt) of the dnaA mRNA. E. TSS selection yields alternative translation products: A full length CDS is translated in the +1 reading frame measuring 804 nt. An internally initiated transcript encodes a 360 nt CDS that is translated in the +2 reading frame highlighted in red. LC-MS peptides corresponding to both CDSs are shown in yellow. F. Alternative transcripts drive two different start codons for ftsW with the position of each translation start site (initiating ATG codon) marked below. ATG 2 is in the same reading frame as ATG 1. Shown on the left is a low agar swarmer plate assay for motility and cell division defects, with the cell division inhibitor sidA as a positive control and the ileS leader CDS (CCNA_03934) as a negative control. Shown in the middle is a cell length distribution from cells containing a xylose inducible high-copy plasmid (pBXSPA) with ftsWs or no insert. Shown on the right is the localization of ftsWs-YFP expressed from a low copy plasmid pftsWs-YFP where the promoter for ftsW has been replaced by two transcription terminators.

Figure 5. C. crescentus noncoding RNA architecture.

A. Previously unannotated intergenic small RNA. B. Small non-coding RNA with a TSS encoded within the 3′ region of CCNA_01035. C. CCNA_03120 mRNA with an extended 3′ UTR overlapping the CCNA_03121-3 operon on the antisense strand.

Figure 6. Complex regulation of C. crescentus operons.

A. Classical ribosomal protein operon containing rpmI & rplT. B. Polarity with decreased RNA read density at the 3′ CDSs in the CCNA_00161-4 operon. C. Transcription attenuation through the ivlL leader CDS to regulate expression of the ilvIH operon. D. Alternative TSS in the rbfA truB rpsO operon can drive differential CDS expression. E. Potential operon cleavage site between CCNA_03729 and CCNA_03730 by the presence of a 5′ monophosphate on the RNA. The higher RNA stability the CCNA_03729 RNA can allow higher 3′ gene expression levels.