Diribonuclease activity eliminates toxic diribonucleotide accumulation

Abstract

RNA degradation is a central process required for transcriptional regulation. Eventually, this process degrades diribonucleotides into mononucleotides by specific diribonucleases. In Escherichia coli, oligoribonuclease (Orn) serves this function and is unique as the only essential exoribonuclease. Yet, related organisms, such as Pseudomonas aeruginosa, display a growth defect but are viable without Orn, contesting its essentiality. Here, we take advantage of P. aeruginosa orn mutants to screen for suppressors that restore colony morphology and identified yciV. Purified YciV (RNase AM) exhibits diribonuclease activity. While RNase AM is present in all γ-proteobacteria, phylogenetic analysis reveals differences that map to the active site. RNase AM_Pa expression in E. coli eliminates the necessity of orn. Together, these results show that diribonuclease activity prevents toxic diribonucleotide accumulation in γ-proteobacteria, suggesting that diribonucleotides may be utilized to monitor RNA degradation efficacy. Because higher eukaryotes encode Orn, these observations indicate a conserved mechanism for monitoring RNA degradation.

Keywords: CP: Molecular biology; RNA degradation; RNase AM; YciV; diribonuclease; diribonucleotides; essentiality; linear; oligoribonuclease.

Figure 1.. Transposon suppressor screen and complementation of suppressor for restoration of Δorn colony size identified yciV

(A) Map of the transposon insertions. Tn #1 is in cpxR (PA14_22760), Tn #2 is in yciI (PA14_22780), and Tn #3 and #4 are in yciB (PA14_22800). (B) Photograph of plates showing in colony size of transposon suppressor mutant compared to PA14 and PA14 Δorn. (C) Photograph of tubes containing overnight culture of transposon mutant, PA14, and PA14 Δorn for visualization of bacterial aggregates. (D) The genomic region surrounding the transposon insertion sites. (E) Photograph of colony size expressing indicated gene fragments in Δorn. Bacterial cultures were diluted and dripped on LB agar plate containing carbenicillin for 30 h. The quantification of colony sizes is shown on the plot. (F) Degradation of AAAAAGG or pGpG by Δorn complemented with suppressor fragments. ³²P-pGpG (top, 1 μM total) and ³²P-AAAAAGG (bottom, 1 μM total) by whole-cell lysates of Δorn complemented with indicated gene fragments. Samples were stopped at the indicated time (min) and analyzed by 20% denaturing PAGE.

Figure 2.. Catalytic site residues of yciV are required for diribonuclease activity

(A) Photograph of PA14 Δorn expressing yciV or alanine substitution alleles were grown on LB agar plate containing 1 mM IPTG for 30 h. Graph shows colony size as measured by Fiji software. (B) Degradation of ³²P-pGpG (total 1 μM) by whole-cell lysates PA14 Δorn complemented with yciV or variants at indicated alanine substitutions was assessed. Samples are stopped at the indicated times (min) and analyzed by 20% denaturing PAGE. All data shown represent the average of triplicate independent experiments. (C) The graph shows the quantification of triplicate data by the amount of remaining pGpG shown in (B). (D) Cleavage of pAp(2AP) by purified RNase AM_Pa or RNase AM_Pa with indicated alanine substation. (E) Cleavage of (2AP)ApsGG and AApsG(2AP) by purified RNase AM_Pa shows that RNase AM_Pa is a 5′ to 3′ exonuclease. For control samples, enzyme was not added.

Figure 3.. RNase AM homologs from different γ-proteobacteria have reduced diribonuclease activity

(A) Photograph of PA14 Δorn expressing yciV from indicated γ-proteobacteria were grown on LB agar plate containing 1 mM IPTG for 30 h. Graph shows colony size as measured by Fiji software. (B) Cleavage of pAp(2AP) by purified RNase AM orthologs from γ-proteobacteria. (C) Degradation of ³²P-pGpG (top) or ³²P-AAAAAGG by whole-cell lysates PA14 Δorn expressing yciV from indicated γ-proteobacteria was assessed. Samples are stopped at the indicated times (min) and analyzed by 20% denaturing PAGE. All data shown represent the average of triplicate independent experiments.

Figure 4.. RNase AM structure and key sequences

(A) Phylogenetic tree and key sequence conservation of RNase AM proteins. Shown is an unrooted phylogenetic tree of RNase AM sequences, with sequences of interest highlighted by triangles on the branch tips. The branch tips are colored by taxonomic order for the organism for each RNase AM sequence, with the color key shown to the right. Only the top 10 most populated taxonomic orders are shown for clarity; all others are in gray. The major clades are labeled by the most representative taxonomic order. Only the top 10 most populated taxonomic orders and top 18 taxonomic families are shown for clarity; all others are in gray. (B) Sequence logos for major clades in the RNase AM phylogeny. Each logo shows the information entropy of important sequence motifs. The residue numbering for Pseudomonas aeruginosa is shown labeled on top of the sequences. Stars highlight conservation in the Pseudomonadales clade that is not conserved in Enterobacterales or Burkholderiales. (C) Crystal structure of RNase AM_Vc. The cartoon shows the structure of V. cholerae RNase AM bound to divalent metal ions and two sulfate molecules (shown here as spheres) at the active site. Close-up views highlight key residues that interact with the ligands. The conserved sequence features shown in (B) are labeled and colored in blue or purple. (D) Predicted diribonucleotide-binding site and pose. Computational docking of pGpG to the crystal structure of RNase AM_Vc resulted in a top-ranked solution in which the diribonucleotide was placed at the active site of the enzyme. Ligands from the crystallization are shown as transparent spheres. The inset shows a close-up view of the docking solution and crystallographic ligands, in addition to AMP from the crystal structure of RNase AM_Cv (PDB: 2YB1). The side chain of residue F¹⁵⁴ in RNase AM_Vc is shown as sticks. (E) Structural model of RNase AM_Pa. The AF2 model of P. aeruginosa RNase AM was aligned with the crystal structure of RNase AM_Vc. The model is shown as a cartoon putty, with the putty thickness and color gradient indicating the RMSD between corresponding residues. An insertion and a predicted C-terminal helix in RNase AM_Pa is shown and labeled in light green. The positions of Pseudomonadales-specific sequence changes are highlighted as black spheres representing their corresponding Cα positions.

Figure 5.. Proteomic analysis of P. aeruginosa Δorn and complemented strains

(A) Normalized protein abundances of each sample (x axis) and each protein (y axis) are clustered by Euclidean distance with a complete linkage method. The clustering resulted in the grouping of samples relative to strain under two discriminate sub-trees. (B) A principal-component analysis of the normalized protein abundance data. (C) A bar graph representing the abundance of Orn_Pa in arbitrary units in each strain. (D) A bar graph representing the abundance of RNase AM_Pa in arbitrary units in each strain.

Figure 6.. Diribonuclease activity is required in γ-proteobacteria

(A) PCR product was transformed into E. coli and orn_Ec was replaced with chloramphenicol acetyltransferase gene via lambda red recombination. The number of Cm^R colonies is indicated in parentheses. The number of colonies in which the orn gene is substituted with the chloramphenicol acetyltransferase gene was determined by PCR. (B) P. aeruginosa harboring a co-integrant for the Δorn deletion was resolved by growth in the absence of selection followed by counterselection on sucrose. Colonies were assessed by PCR to determine the number of colonies that reverted to wild type or deleted the orn gene.

Figure 7.. Model for flux of ribonucleotides

Estimate of RNA distribution in growing P. aeruginosa cells based on previous studies on E. coli and P. aeruginosa. The box represents a P. aeruginosa cell. The gray area containing circles represents the amount of nucleotides (nt) in each category of RNA that would be present in both genetic backgrounds (wild type [WT] or Δorn). For ribosomes, the amount of nt is based on an estimate of 10,000–70,000 ribosomes in an E. coli cell. Each ribosome has ~4,400 nt for a total of ~4.4 × 10⁷–3.1 × 10⁸ nt. This model uses a cell with 40,000 ribosomes for WT P. aeruginosa and Δorn. Each of the mononucleotides (adenine, guanine, cytosine, and uracil) includes mono-, di-, and tri-phosphate versions and has been measured in E. coli for a combined total of 5.2 mM.^, In this estimate, there would be ~5.2 × 10⁶ nt in the cell as mononucleotides for WT P. aeruginosa and Δorn. When degraded, these larger RNA molecules will generate dinucleotides. For WT P. aeruginosa (indicated in the yellow portion), Orn is present and will cleave dinucleotides into mononucleotides. A prior estimate of pGpG shows that there is ~10 μM in the cell. Assuming all of the linear diribonucleotides are present at similar concentrations, there would ~160 μM or ~1.6 × 10⁵ nt per cell. In Δorn mutants (indicated in the green portion), the concentration of pGpG increases to ~70 μM, which extrapolates ~1.12 × 10⁶ nt for all 16 linear dinucleotides per cell.