Molecular recognition of RhlB and RNase D in the Caulobacter crescentus RNA degradosome

Abstract

The endoribonuclease RNase E is a key enzyme in RNA metabolism for many bacterial species. In Escherichia coli, RNase E contributes to the majority of RNA turnover and processing events, and the enzyme has been extensively characterized as the central component of the RNA degradosome assembly. A similar RNA degradosome assembly has been described in the α-proteobacterium Caulobacter crescentus, with the interacting partners of RNase E identified as the Kreb's cycle enzyme aconitase, a DEAD-box RNA helicase RhlB and the exoribonuclease polynucleotide phosphorylase. Here we report that an additional degradosome component is the essential exoribonuclease RNase D, and its recognition site within RNase E is identified. We show that, unlike its E. coli counterpart, C. crescentus RhlB interacts directly with a segment of the N-terminal catalytic domain of RNase E. The crystal structure of a portion of C. crescentus RNase E encompassing the helicase-binding region is reported. This structure reveals that an inserted segment in the S1 domain adopts an α-helical conformation, despite being predicted to be natively unstructured. We discuss the implications of these findings for the organization and mechanisms of the RNA degradosome.

Figure 1.

Small recognition motifs in Caulobacter crescentus RNase E. (A) ANCHOR prediction (blue line) of protein–protein interaction sites within RNase E (labelled A–E). The red line corresponds to the IUPred prediction of intrinsically unstructured regions. Bottom, a table summarizing the ANCHOR predicted protein-binding sites and known protein binding partners. The ANCHOR maxima at residue 180 and 550 correspond to regions that are not solvent exposed and are involved in intra-domain interactions respectively in the Escherichia coli crystal structure. (B) Co-immunopurification of C. crescentus RNA degradosome via anti-RNase E antibody, or N-terminally FLAG tagged RNase E. (C) Pull down experiment with GST fusions of ANCHOR predicted protein binding segments, or GST alone. RNase D is marked with an asterisk. (D) Reciprocal pull down experiment using 6xHis-tagged RNase D as bait. Controls are shown in Supplementary Figure S3.

Figure 2.

Interactions of the catalytic domain of Caulobacter crescentus RNase E with the DEAD-box helicase RhlB. (A) Size exclusion chromatography of the CcNTD_1–575–RhlB complexes. Chromatograms from three experiments (CcNTD_1–575, RhlB and a 1:1.3 molar ratio respectively) on a 16/60 S200 size exclusion column are overlaid. The SDS-PAGE gel (inset) contains fractions from the peak corresponding to the CcNTD_1–575–RhlB complex (*), with CcNTD_1–575 and RhlB labelled ‘N’ and ‘R’ respectively. The molecular weights of the standards in lane 1 are indicated (kDa). (B) A proteolytically liberated fragment of CcNTD_1–575 containing the S1 and 5′-sensor domains (CcNTD_1–274) forms a complex with RhlB. As for (A), chromatograms from the three experiments are overlaid. The peak corresponding to the CcNTD_1–274–RhlB complex (*) resulted from co-expressing and co-purifying the two components. The corresponding peak fractions are shown in the SDS-PAGE gel (inset), with CcNTD_1–274 and RhlB labelled ‘F’ and ‘R’ respectively. (C) A continuous molecular mass distribution from analytical ultra-centrifugation sedimentation velocity analysis of the CcNTD_1–575–RhlB complex in (A). The peak value corresponds to a mass of 380 kDa. (D) A continuous molecular mass distribution from sedimentation velocity analysis of the CcNTD_1–274–RhlB complex in (B). The peak value corresponds to mass of 93 kDa.

Figure 3.

The C-terminal extension of RhlB mediates the interaction with RNase E catalytic domain. (A) Limited trypsin proteolysis of RhlB liberates stable fragments characterized by MALDI mass spectrometry as residues 1–390 and 416–517. (B) Pull down experiment with GST-RhlB_416–517, co-purifying degradosome components are indicated. (C) CcNTD_1–575 was mixed with either N-RhlB (1–390) or (D) GST-RhlB_416–517 with a 1:1.3 molar ratio of CcNTD_1–575 and RhlB respectively. The individual mixtures were then run on a superdex 200 16/60 gel filtration column and elution was monitored by UV absorbance at 280 nm. The numbered peaks from the elution profiles were analysed by SDS-PAGE (inset), with numbers above the gel corresponding to the peak from the profile and numbers to the left of the gel are the molecular weights of the standards (kDa).

Figure 4.

X-ray crystal structure of the CcNTD_1–274 fragment from the catalytic domain of RNase E. (A) A linear schematic of the CcNTD_1–575 domain architecture. RNase H, S1, 5′-sensor, DNase I and small domains are indicated along with the insertion within the S1 domain in Caulobacter crescentus RNase E. S1, S1 insert and 5′-sensor domains as resolved in the crystal structure of CcNTD_1–274 are coloured green, yellow and blue respectively. The regions not resolved in the crystal structure are shown in dashed boxes coloured grey. (B) The crystal structure of the proteolytically liberated fragment of CcNTD_1–575 (CcNTD_1–274) is shown as cartoon representation, with the S1 domain, the S1 insert and the 5′-sensor domain coloured as in (A). (C) The structure of CcNTD_1–274 is aligned to the equivalent portion of the crystal structure of the EcNTD apo-protein (PDB ID: 2VMK) at the C-α atoms using PyMol. Both structures are shown in ribbon representation, with CcNTD_1–274 coloured as in (A), and the EcNTD structure is coloured grey. The rmsd for the alignment was 0.82 Å. (D) CcNTD_1–274 is aligned to EcNTD (PDB ID: 2C0B, rmsd at Cα = 0.82 Å) to reveal the predicted location of the former in the tetramer of CcNTD_1–575. Two views are shown after rotating 90° about the indicated axis. For CcNTD_1–274, the S1 domain, S1 insert and 5′-sensor domains are coloured as in (A). The EcNTD tetramer is coloured grey.

Figure 5.

Comparison of the CcNTD_1–274 and the RNA-bound EcNTD crystal structures. (A) Close up view of the RNA binding site in the EcNTD crystal structure (PDB ID: 2C0B). The RNA backbone is coloured in orange, the S1 domain (36–118) in green and the 5′-sensor domain (119–214) in blue. (B) The equivalent view of the CcNTD_1–274 crystal structure aligned onto EcNTD, showing the S1 insert helices in yellow (residues 97–145) protruding into the putative RNA binding site. All colouring is consistent with previous figures.

Figure 6.

A schematic representation of the Caulobacter crescentus RNA degradosome in comparison to the paradigm degradosome assembly from Escherichia coli. Top: C. crescentus degradosome. The N-terminal domain of RNase E is shown as a solid blue bar, with the binding site for RhlB, and the S1 insert indicated. The disordered C-terminal domain is shown as a thin wavy blue line, with interaction sites for aconitase, RNase D and PNPase indicated. RhlB, aconitase, RNase D and PNPase are depicted as red, orange, grey and brown filled blocks respectively. Bottom: E. coli degradosome. As above, but with membrane attachment motif and enolase shown as green and yellow blocks respectively.