Insight into centromere-binding properties of ParB proteins: a secondary binding motif is essential for bacterial genome maintenance

Abstract

ParB proteins are one of the three essential components of partition systems that actively segregate bacterial chromosomes and plasmids. In binding to centromere sequences, ParB assembles as nucleoprotein structures called partition complexes. These assemblies are the substrates for the partitioning process that ensures DNA molecules are segregated to both sides of the cell. We recently identified the sopC centromere nucleotides required for binding to the ParB homologue of plasmid F, SopB. This analysis also suggested a role in sopC binding for an arginine residue, R219, located outside the helix-turn-helix (HTH) DNA-binding motif previously shown to be the only determinant for sopC-specific binding. Here, we demonstrated that the R219 residue is critical for SopB binding to sopC during partition. Mutating R219 to alanine or lysine abolished partition by preventing partition complex assembly. Thus, specificity of SopB binding relies on two distinct motifs, an HTH and an arginine residue, which define a split DNA-binding domain larger than previously thought. Bioinformatic analysis over a broad range of chromosomal ParBs generalized our findings with the identification of a non-HTH positively charged residue essential for partition and centromere binding, present in a newly identified highly conserved motif. We propose that ParB proteins possess two DNA-binding motifs that form an extended centromere-binding domain, providing high specificity.

Figure 1.

Model of SopB–sopC interaction. (A) Schematic representation of SopB. Top; amino acid numbering of SopB to scale. Middle; secondary structure of SopB. Helices and arrows represent α-helix and β-sheet, respectively, as determined by crystal analysis in blue or predicted (predator, JPRED) in green. Bottom; functional domains and motifs of SopB. Thick cylinders (from amino acids 155–272 and 275–323) represent the parts of SopB structure that have been solved, whereas thin cylinders (from amino acids 1–155 and 272–275) have not (26). (B) Representation of SopB–sopC structure drawn from accession number 3EZF (26), with R219 colored in red, R190, K191 and R195 in blue and the guanine 7 in green. Helices 2 and 3 of the HTH motif are colored in dark blue. (C) Predicted interactions between amino acids of SopB with sopC nucleotides, adapted from the study by Pillet et al. (27). R190, R219, R195 and K191 correspond to arginines 190, 195, 219 and lysine 191. Straight and dotted lines represent suggested amino acid–nucleotide interactions. R and Y stand for purine and pyrimidine nucleotides, respectively. Green zones represent the subunits of an SopB dimer.

Figure 2.

Interaction of SopB wild-type and R219 variants with DNA. The ability of wild-type SopB, SopB-R219A and -R219K to interact with an sopC DNA duplex or with non-specific DNA was tested in a gel mobility shift assay. Reaction mixtures, assembled on ice and incubated at 30°C, were analyzed by electrophoresis on polyacrylamide gels. (A) SopBs binding to sopC centromere site. ³²P-labeled 30-bp sopC2 probe was incubated alone (−) or with increasing concentrations (from left to right; 10, 30, 100, 300 and 1000 nM) of wild-type SopB (WT), SopB-R219A or SopB-R219K as indicated on top, in the presence of 100 µg ml⁻¹ competitor DNA. Positions of DNA probes and protein complexes are indicated on the left: B1 denotes the specific complex, whereas asterisk corresponds to a specific shift from a secondary form of the duplex DNA. (B) SopBs binding to non-specific DNA probe. ³²P-labeled 114-bp repE probe was incubated alone (−) or with increasing concentrations (from left to right; 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000 and 2000 nM) of wild-type SopB (WT) or SopB-R219A as indicated on top, in the absence of competitor DNA. NS1 and NS2 denote discrete complexes, and HMW corresponds to high molecular weight complexes.

Figure 3.

SopB-R129A allele does not form foci cells. Left; localization of SopB–Venus fusion. Fluorescent foci from the mini-F expressing SopB–Venus display the expected pattern for partition complexes faithfully partitioned. Right; localization of SopB-R219A–Venus fusion. All cells displayed diffuse fluorescence throughout the cytoplasm. Cells without fluorescence are presumed to lack the mF carrying SopB-R219A–Venus as expected from the loss rate (see Supplementary Figure S2A and main text).

Figure 4.

Interaction of wild-type SopB with mutated sopC centromeric DNA. Top; the ability of wild-type SopB to interact with different sopC DNA duplexes (indicated on top) was assayed as in Figure 2. ³²P-labeled 30-bp sopC probes were incubated alone (−) or with increasing concentrations of wild-type SopB (from left to right; 10, 30, 100, 300 and 1000 nM), except for sopC2 where the 1000 nM SopB reaction was omitted. Legend is as in Figure 2. Bottom; bases at positions 7 (bottom strand) and 10 (top strand) are indicated in grey. Numbers indicate positions of the bases in the palindromic SopB-binding site according to the study by Pillet et al. (27).

Figure 5.

Chromosomal ParBs binding to parS sites. DNA binding interaction between Pput or Bcen parS, left and right panel, respectively, and wild-type ParB from Pput (ParB_Pput), Bcen (ParB_Bcen) or variant ParB-R192A_Bcen and ParB-R194A_Pput proteins were assayed by EMSA as in Figure 2. ³²P-labeled 30-bp parS probes were incubated alone (−) or with increasing concentrations of purified ParB_Pput proteins (left panel; from left to right: 0, 10, 30, 100 and 300 nM) or ParB_Bcen proteins in crude extracts (right panel; from left to right: 0, 2.8, 8.4, 28 and 84 mg ml⁻¹). Legend is as in Figure 2.

Figure 6.

ParB conserved motifs and residues involved in centromere binding. (A) Amino acid alignment of ParB homologues. Sequence alignments of SopB (F) with KorB (RP4) and Spo0J (T. thermophilus) are based on the study by Leonard et al. (34). Identical residues and highly conserved residues are indicated by asterisks and dots, respectively. Helices H1 to H5 of SopB, based on the crystal structure analysis (26), are schematically represented on top; Helices H2 and H3, corresponding to the HTH motif, are drawn in yellow. Arginine residues R219 (SopB), R240 (KorB) and R179 (Spo0J) are colored in red. (B) Ribbon diagrams of a ParB dimer. Structure is from T. thermophilus Spo0J (34). One monomer is colored in cyan and the other in green. The right diagram is a 90° horizontal rotation of the left diagram. HTH motifs and R179 residues of each monomer are colored in yellow and red, respectively. Each arm of the 16-bp parS DNA is expected to bind one HTH motifs of the Spo0J dimer. Orientations of R179 residues are compatible for providing additional specificity contacts in parS DNA. (C) Amino acid alignment of chromosomal ParBs. ParBs from a large spectrum of bacterial species were aligned with ClustalW; in the case of multichromosomal bacteria, ParBs are from the main chromosome. These ParBs are expected to bind the parS consensus sequence 5′-tGTTNCACGTGNAACa-3′ (30,48). Sequences highlighted in yellow and orange correspond to the HTH motif and the CBM2 (centromere-binding motif 2) box, respectively. Helices are represented as in (A) based on the crystal structure of Spo0J of T. thermophilus (34); amino acid sequence are numbered according to Tthe Spo0J. The red plus sign (+) schematically represents the conserved R or K positively charged residues. ParBs are from Borrelia burgdorferi B31(Bbur), B. cenocepacia (Bcen), B. subtilis (Bsub), Campylobacter jejeuni (Cjej), Caulobacter crescentus (Ccre), Chlamydia trachomatis (Ctra), Helicobacter pylori (Hpyl), Mycobacterium tuberculosis (Mtub), Myxococcus xanthus (Mxan), Neisseria meningitidis (Nme), P. putida (Pput), Porphyromonas uenonis (Puen), Rickettsia typhi (Rtyp), Staphylococcus aureus (Saur), Streptomyces coelicolor (Scoe), Streptococcus pneumonia (Spne), T. thermophilus (Tthe) and Vibrio cholera (Vcho).