Involvement of the inner surface residues of bacterial SMC protein MukB in the ssDNA binding in vitro

Abstract

The bacterial condensin MukB facilitates proper chromosome segregation in Escherichia coli. MukB protein localizes at the ori adjacent region by unknown mechanism. The MukB protein entraps the single-stranded DNA (ssDNA) molecule more efficiently than double-stranded DNA (dsDNA). In the bacterial genome, several copies of the rrn genes are encoded near the ori region. The rrn regions are expected to efficiently generate ssDNA due to their high transcriptional activity and the frequent formation of R-loops. In this study, we identified residues involved in DNA binding. The mutations impaired ssDNA binding more severely than dsDNA binding in vitro, and also caused deficiencies in cell growth and nucleoid segregation. These amino acid residues are aligned and are thought to bind DNA when the MukB dimer entraps a DNA molecule within its ring, thereby likely enhancing the DNA-binding activity of MukB. These residues may contribute to the accumulation of MukB on the chromosome, including the rrn regions.

Amino acids of the E. coli condensin MukB were identified to be important for ssDNA binding in vitro and potentially necessary for chromosome segregation. Mutations in key residues causes deficiencies in cell growth and nucleoid segregation.

Introduction

The compaction of replicated chromosomes is crucial for proper chromosome segregation within the confined space of bacterial cells. Nucleoid organization, which is involved in the highly ordered arrangement of chromosomal DNA, plays a critical role in this process. Two key factors contribute to the tight compaction of chromosomal DNA: topoisomerase, which generates DNA supercoiling, and bacterial condensin, which folds newly replicated DNA in a manner that is not yet fully understood^1–4. Bacterial condensin is a member of the structural maintenance of chromosome (SMC) protein family^5–9. Two major types of bacterial condensin have been identified and extensively studied to unravel the molecular mechanisms of DNA compaction: the Smc–ScpAB complex in Bacillus subtilis^10–12 and the MukBEF complex in Escherichia coli^2,13.

In the Smc–ScpAB and MukBEF complexes, Smc and MukB function as their respective core subunits^5,7,8. These proteins are collectively classified as SMC proteins. The N-terminal and the C-terminal parts of the SMC proteins undergo intramolecular assembly, forming the ATPase head domain in an ABC cassette arrangement, whereas the middle region forms the hinge domain. The head and the hinge domains are connected by the coiled-coil arm domain. Homodimerization of the MukB protein leads to the formation of a ring structure, a characteristic feature of SMC proteins^5,6,14. In the case of the MukB dimer, the flexibility of the arm domain allows the MukB dimer to adopt various shapes^15,16. The SMC proteins form holo complexes with the kleisin subunits (MukF and ScpA) and the kite subunits (MukE and ScpB)^7,16–23. MukE and MukF stably form the MukEF complex, which inhibits the DNA binding of MukB in vitro^18,24. Although MukE and MukF proteins are essential for the functional cycle of bacterial condensin, it is demonstrated that MukB protein alone can act on DNA to compact it^4,25,26. In addition to MukEF protein, MukB interacts with other protein factors such as TopoIV and Acyl-carrier protein (ACP)^14,27–32.

MukB binds to DNA in two modes. One is a canonical protein–DNA interaction mediated by the electrostatic force between the positively charged residue of the protein and the negatively charged phosphate backbone in the polynucleotide. The other is the topological binding, where the protein captures a DNA strand inside its proteinaceous ring^{5,7,8,33–35}. In the former mode, MukB binds to single- and double-stranded DNA (ssDNA and dsDNA) by its head domain in a non-sequence-specific manner. Binding to ssDNA and dsDNA is a common feature of the SMC proteins. In general, not all dsDNA-binding proteins possess ssDNA-binding activity and vice versa. Thus, it is probable that the SMC proteins utilize the ssDNA binding ability, whereas their biological significance remains elusive. In B. subtilis, the Smc protein binds to ssDNA through its hinge domain³⁶. The hinge domain of B. subtilis Smc can be replaced with that of Rad50, an SMC-like protein in yeast that is involved in DNA repair³⁷. This suggests that the ssDNA-binding activity of the B. subtilis Smc hinge is not essential for the functional cycle of bacterial condensin, if indeed it plays any role at all. In the case of MukB, the significant DNA-binding site is its head domain^19,38,39. Several multiple mutations in the MukB head domain impair its dsDNA-binding activity¹⁹, but it is unclear whether the resulting mutants also have decreased affinity for ssDNA.

The latter mode, the topological binding mechanism, is conserved in the SMC protein family, including bacterial condensin and plays key roles in their properties^40–43. Understanding the mechanism by which SMC proteins capture chromosomal DNA is essential for elucidating the fundamental and universal principles of their biological functions. The loop extrusion model attempts to explain how SMC proteins fold chromosomal DNA, resulting in an organized loop structure^44–49. Alternatively, it is thought that two SMC proteins, especially when bound topologically to different sites on chromosomes, gather to compact DNA⁵⁰. In any case, it is critical to determine where and how SMC proteins topologically load on chromosome sites. However, consensus nucleotide sequences for topological binding have not been found thus far, either.

MukBEF complexes form clusters at the oriC-adjacent region^51,52, although they can be distributed throughout the entire chromosome region except the ter domain, where MukB is excluded by matS-MatP system^30,53,54. MukB has been shown to play a role in the segregation of oriC regions, suggesting that its localization to these regions is particularly important for its function in chromosome segregation. The mechanism by which MukB achieves the localization at a specific site, the oriC region, on the chromosome without sequence specificity is still unknown. In B. subtilis, Smc-ScpAB complex is loaded at the oriC-adjacent parS site and translocates toward the ter region. Smc-ScpAB also accumulates on the actively transcribed ribosomal RNA (rrn) operon⁴², likely through topological-type DNA binding. More than two copies of the cis-located rrn gene operon are necessary to segregate the sister chromosomes properly⁴². By assembling at independent rrn regions, Smc-ScpAB complexes contribute to the organization of the oriC region. The accumulation of SMC protein at rDNA is observed in yeast, too^55–57. The rrn region forms an R-loop, a DNA-RNA hybrid, during transcription, resulting in the relatively stable presence of an ssDNA region⁵⁸. Many bacteria, including E. coli and B. subtilis, possess multiple rrn operons near the oriC region. Bacterial condensin could achieve oriC-adjacent localization by targeting rrn region. One plausible scenario to explain the localization of bacterial condensin to the rrn region could involve leveraging the property of ssDNA-specific topological binding, as demonstrated with MukB⁴¹.

Purified MukB protein topologically binds to DNA in a manner independent of MukEF protein and ATP, and MukB prefers ssDNA as the substrate⁴¹. Other SMC proteins have also been shown to have the potential to capture ssDNA. Yeast cohesin captures DNA twice sequentially, enclosing two DNA molecules in the ring⁵⁹. While both ssDNA and dsDNA are effectively captured at the first topological-binding reaction, only ssDNA can serve as the effective substrate for the second capturing reaction. As another example, yeast condensin exhibits ssDNA-binding activity and promotes effective renaturation of complementary ssDNA in vitro⁶⁰. These observations strongly suggest that SMC proteins can distinguish ssDNA from dsDNA through an unknown mechanism. In this study, we identified amino acid residues that are important for ssDNA binding in vitro. These residues are candidate determinants responsible for the observed preference for ssDNA. Based on our structural and biochemical analyses, we propose a working hypothesis for the molecular mechanism underlying the preferential establishment of topological binding on ssDNA rather than dsDNA.

Material and methods

Bacterial strains and plasmids

Bacterial strains and plasmids used in this study are listed in Supplementary Tables S1 and S2. LB-broth (10 g/L Bacto tryptone, 5 g/L yeast extract, 5 g/L NaCl: pH adjusted to 7.5 by using NaOH) was used for bacterial culture. Kanamycin (25 µg/mL) was added to the medium if necessary. The ∆mukB::cat mutation was introduced into the MC1061 strain by the λRED recombination method⁶¹, resulting in YAN4081.

To obtain a mutated gene with a single amino acid substitution, the mukB gene was cloned into pET28 and expressed as an N-terminus histidine-tagged product, resulting in pHis6-MUKB. The introduction of a single amino acid substitution in His₆-MukB was achieved through two types of PCR-mediated site-directed mutagenesis. The plasmid DNA of pHis6-MUKB was amplified by PCR using a pair of complementary primers that included the desired mutations. After removing the template DNA of pHis6-MUKB through DpnI treatment, the PCR product was transformed into DH5α cells. Instead of using conventional site-directed mutagenesis, the in vivo E. coli cloning method (iVEC) was employed⁶². DNA fragments having homologous overlap regions at both the 5’ and 3’ ends were amplified through a PCR reaction with pHis6-MUKB as a template DNA and a pair of primers, one of which contained the target mutation. The PCR product was then transformed into the iVEC strain using a method involving PEG 8000, as described in Nozaki and Niki⁶². The circularization of the PCR fragment by enzymes in the iVEC strain resulted in the establishment of a plasmid. For both methods, the established plasmid was extracted from the cell, and the entire coding region of his₆-mukB was sequenced.

Cell viability

Temperature-sensitive growth was determined by assessing cell viability at a specific temperature at which colonies formed on an agar plate. The cells were grown in L-broth at 25 °C until the early log phase. A portion of the culture was then removed and washed with saline. The cells were resuspended in saline and serially diluted. Three μL aliquots of each dilution of the cell suspension were then spotted onto multiple L-plates, and respectively incubated at 25 °C and 37 °C.

Microscopy

Cells were grown in L-broth at 25 °C until the early log phase. An aliquot of the cells was diluted with fresh L-broth followed by further shaking for 2.5 h at 37 °C. An additional aliquot of the cells was removed and washed with saline. The cells were resuspended in saline and spread on a poly-lysine-coated slide glass. After air-drying for 5 min, the cells were treated with 80% methanol for 5 min, followed by a wash with ultrapure water and subsequent air-drying. The cells mounted with DAPI-containing solution were analyzed using a microscope (ZEISS). Images were processed using AxioVision or ZEN.

Immunoblotting

YAN4081 cells harboring a plasmid encoding the his₆-mukB were grown in L-broth at 25 °C. A part of culture was mixed with an equal volume of 10% trichloroacetic acid and incubated on ice. Precipitated proteins were centrifuged, followed by washing with acetone. Pellets were dissolved in 1x SDS sample buffer (62.5 mM Tris-HCl (pH 6.8), 2% SDS, 10% glycerol, 10% 2-Mercaptoethanol) followed by boiling. Samples were analyzed by SDS-PAGE with 8% Laemmli gel and blotted on a PVDF membrane (Merck). Anti-MukB antibody (laboratory stock; #9D-326) and anti-rabbit-IgG-HRP were used as the 1st and the 2nd antibody, respectively. The protein signals were visualized with ECL reagent (PerkinElmer Inc., Western Lightning Plus-ECL) and ATTO image analyzer.

Purification of His-tagged MukB

The His-tagged MukB protein was prepared essentially as described previously⁴¹. BL21(DE3) cells harboring a plasmid encoding a his₆-tagged mukB were grown at 37 °C in LB-broth containing kanamycin. 20 mL of the overnight culture was inoculated into 1 L of LB-broth containing kanamycin and grown till the optical density (OD₆₀₀) reached 0.3. Isopropyl β-D-thiogalactopyranoside (IPTG) was added to a final concentration of 0.1 mM, and the culture was incubated at 37 °C for an additional 2 h. Cells were harvested and washed with 25 mL of HK buffer (50 mM HEPES-KOH (pH 7.6), 100 mM KCl). Cell pellets were frozen with liquid nitrogen and stored at −80 °C until use. For protein purification, the frozen cell pellets were thawed on ice and resuspended in 30 mL of sonication buffer (50 mM HEPES-KOH (pH 7.6), 100 mM KCl, 1 mM phenylmethylsulfonyl fluoride). Cells were sonically disrupted by BRANSON SONIFIER 250 (30 seconds for 20 times with 30 seconds intervals) followed by centrifugation at 10,000 × g for 10 min to remove the cell debris. Supernatant was moved into a glass flask, and ammonium sulfate was added to a final concentration of 30%. The mixture was incubated on ice for 15 minutes to precipitate proteins. The mixture was centrifuged at 9100 × g for 10 min at 4 °C, and discarded the supernatant. The pellet was resuspended in 10 mL of the equilibrium buffer (50 mM HEPES-KOH (pH 7.6), 100 mM KCl, 20 mM imidazole). The suspension was centrifuged again at 9100 × g for 10 min at 4 °C, and the supernatant was applied to an open column filled with 8 mL of TALON metal affinity resin (Clontech) equilibrated with the equilibrium buffer. After washing with 15 mL of wash buffer (30 mM HEPES-KOH (pH 7.6), 750 mM KCl, 10 mM imidazole), the His₆-tagged MukB proteins were eluted by elution buffer (50 mM HEPES-KOH (pH 7.6), 100 mM KCl, 500 mM imidazole). Each 1 mL of the eluates was collected into microtubes. The fractions containing the eluted protein were detected by measuring the absorbance at 280 nm, and the peak fragments were pooled. The pooled fragments containing His₆-tagged MukB proteins were loaded onto a Resource Q column (1 mL; GE) equilibrated with buffer A (50 mM HEPES-KOH (pH 7.6), 100 mM KCl). After washing the column with 25 mL of buffer A, a linear gradient between 100 to 1000 mM KCl was generated in the column by 20 mL of buffer B (50 mM HEPES-KOH (pH 7.6), 1000 mM KCl). Each 1 mL of the fraction was collected in microtubes. Pooled fractions containing His₆-tagged MukB protein were dialyzed with dialysis buffer (20 mM HEPES-KOH (pH 7.6), 20 mM KCl, 0.1 mM DTT, 0.1 mM EDTA, 50%(w/v) glycerol). The concentration of purified proteins was measured with Bradford Protein Assay (Bio-Rad). The integrity of the purified protein was checked by SDS-PAGE followed by CBB staining (Bio-Rad, Bio-Safe Coomassie Stain) (Fig. S4).

Electromobility shift assay (EMSA)

A binding reaction was performed using 10 μL of a reaction mixture comprising a specific quantity of purified His₆-MukB protein and either 5 fmol of circular single-stranded (css) DNA or covalently closed circular (ccc) DNA from pUC119. The reaction mixture contained a buffer with 25 mM HEPES-KOH (pH 7.6), 100 mM KCl, and 1 mM DTT. After incubating the mixture at 37 °C for 10 min, loading dye was added. Subsequently, the reaction mixture was subjected to electrophoresis through a 0.7% agarose gel in Tris-borate buffer at 100 V for 30 min at 4 °C. Prior to loading the reaction mixture, the gel was pre-run at 100 V for 60 min at 4 °C. The DNA gel was stained with SYBRGreen II (Takara), and DNA bands were detected using a LuminoGraph (ATTO) system. The intensity of DNA bands was measured using Fiji software, which can be found at this website: https://imagej.net/software/fiji/.

Purification of DNA substrates for EMSA and topological-binding assay

The DNA substrates were prepared following a procedure similar to that described previously⁴¹. Circular single-stranded (css) DNA of pUC119 was generated using M13 phage. E. coli MV1184 cells containing pUC119 were grown overnight in 3 mL of 2x YT medium (Bacto tryptone 16 g/L, yeast extract 10 g/L, NaCl 10 g/L, pH 7.6) with 0.25% glucose at 37 °C. Then, 1.5 mL of the overnight culture was mixed with the lysate of the M13KO7 helper phage and incubated at 37 °C for 1 h for infection. The mixture was then transferred into 150 mL of 2x YT medium with 0.25% glucose and incubated at 37 °C for 1.5 h. Kanamycin (final concentration 50 μg/mL) was added, and the incubation was continued until full growth was achieved. The culture was centrifuged, and the supernatant was collected. Phage particles in the supernatant were precipitated by ultracentrifugation at 100,000 rpm for 30 min using a TLS-110 rotor (Optima TLX; Beckman Coulter Inc.). The resulting pellet was suspended in 800 μL of Tris-EDTA solution (pH 8.0). The solution was treated with 5 μg/mL RNaseA and 70 unit/ml DNase I (Takara) for 1 h at 37 °C. Circular single-stranded DNA of pUC119 was isolated by phenol and phenol/chloroform extraction followed by ethanol precipitation. Covalently closed circular (ccc) DNA was prepared by using a DNA extraction kit (Promega). The obtained cccDNA was purified by CsCl density gradient ultracentrifugation.

Topological-binding assay

A topological-binding reaction was performed using 10 μL of a reaction mixture comprising 3.6 pmol of His₆-MukB and 100 ng of css pUC119 DNA. The reaction mixture contained a buffer with 25 mM HEPES-KOH (pH 7.6), 25 mM KCl, 1 mM DTT, and 1 mM MgCl₂. After the mixture was incubated at 37 °C for 15 min, it was chilled on ice, and then the reaction was quenched by adding 500 μL of CP buffer (25 mM HEPES-KOH (pH 7.6), 500 mM KCl, 1 mM DTT, 1 mM MgCl₂, 5% glycerol, 0.35% Triton X-100, and 5 mM imidazole). Then 10 μL of His-Tag beads (Dynabeads; Thermo Fischer Scientific) equilibrated with CP buffer were added, and the mixture was gently rotated at 4 °C for 30 min. The beads were collected on a magnetic stand and washed with 1 mL of CW1 buffer (25 mM HEPES-KOH (pH 7.6) 750 mM KCl, 1 mM DTT, 1 mM MgCl₂, 0.35% Triton X-100, and 5 mM imidazole) twice followed by washing with 1 mL of CW2 buffer (25 mM HEPES-KOH; pH 7.6, 100 mM KCl, 1 mM DTT, 1 mM MgCl₂, 0.1% Triton X-100, 5 mM imidazole). The remnant buffer was completely removed. Then the beads were resuspended in 12 μL of elution buffer (25 mM HEPES-KOH (pH 7.6), 100 mM KCl, 1 mM DTT, 1 mM MgCl₂, and 500 mM imidazole) and incubated at 37 °C for 5 min. To a 10 μL aliquot of the supernatant were then added 2 μL of loading dye (NEB, #B7021S) and 1 μL of 2% sodium N-dodecanoylsarcosinate. The samples were electrophoresed (100 V, 20 min, room temperature) and DNAs stained by SYBRGreen II (Takara) were detected by LuminoGraph (ATTO). The intensity of DNA bands was measured using Fiji software, which can be found at this website: https://imagej.net/software/fiji/., and the rate of DNA retrieval was calculated.

Statistics and reproducibility

All graphs were generated using Numbers (Apple Inc. software). Biochemical data were obtained from two or three independent experiments.

Results

Positively charged residues on the inside surface of the MukB head domain are conserved

MukB preferentially entraps ssDNA rather than dsDNA in vitro⁴¹. The topological entrapment of DNA by the MukB ring implies that ssDNA interacts with amino acid residues on the inner surface of the closed proteinaceous ring. Generally, a phosphate residue of ssDNA interacts with a positively charged amino acid residue when a protein specifically binds to ssDNA. The inside surface of the closed MukB dimer carries a predominantly positive charge, while the outer surface is negatively charged (Fig. 1a). Hence, we focused our attention on the positively charged residues, Lys (K) and Arg (R), on the inner surface of the MukB, as a key element for ssDNA binding. We mainly based our estimation of candidate amino acid residues on the crystal structure of the MukB protein encoded in Haemophilus ducreyi¹⁹, given the high similarity in amino acid sequence between H. ducreyi MukB and E. coli MukB: 77% for the entire region and 89% for the head domain (Fig. 1b). Within the head domain of the MukB ring, the region ranging from R61 to K119 appears to be mainly localized at the inner surface of the closed proteinaceous ring, which is where the amino acid residues having the potential to interact with ssDNA lie. This region contains 12 positively charged amino acids that lie either on or in close proximity to the inner surface (Fig. 1a, b). These positively charged residues were found to be well conserved among MukB homologs (Fig. 1b).

Fig. 1. Conserved positively charged residues and crystal structure in the head domain of MukB.

a The crystal structure of the head domain of the Haemophilus ducreyi MukB dimer (PDB-ID: 3EUK) is depicted using ribbon diagrams in the top panels, while surfaces with electronic charge density are shown in the bottom panels. These crystal structure models were generated using PyMOL (https://pymol.org/2/). In the top panels, each monomer of the head-engaged dimer is represented in green and gray, respectively. Arrows show the regions connected to the arm. The structure between R61 and K119 is displayed in magenta, and the 12 positively charged residues in (b) are indicated as blue spheres in the top panels and as yellow open circles in the bottom right panel. b The phylogenetic tree and the amino acid sequence alignment of MukB homologs indicate nine bacteria closely related to Escherichia coli: Actinobacillus succinogenes, Vibrio cholera, Escherichia coli, Shigella flexneri, Photorhabdus laumondii, Yersinia pestis, Haemophilus ducreyi, Aeromonas dhakensis, and Pasteurella multocida. The alignment, along with the phylogenetic tree, was generated by Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/). The residue numbers of E. coli MukB are indicated at the top of the panel. In the alignment, positively charged residues are highlighted in blue, and negatively charged residues in red. The 12 residues indicated by the open triangles at the bottom with the residue number were substituted in this experiment.

Several head domain mutants of MukB showed temperature-sensitive growth

We introduced amino acid substitution mutations into the head domain of MukB to determine which of these positively charged amino acid residues on the inner surface of the MukB ring is involved in the specific recognition of ssDNA. A plasmid containing the mukB gene with an amino acid substitution mutation was constructed. The plasmid expressed the N-terminal His-tagged gene product under the T7 promoter. Twelve positively charged amino acids were substituted for glutamic acid (E), which is negatively charged, or glutamine (Q), which is neutrally charged. To assess the biological significance of the amino acid substitution in the MukB protein, we examined whether the mutations seriously affect cell growth and chromosome segregation (Table 1 and Figs. 2, 3).

Table 1.

Properties of the MukB head domain mutants

	Anucleate cell production rate (%)a		Nucleoid segregationb	Cell shapec	Cell viabilityd
His6-MukB	25 °C	37 °C	25 °C/37 °C	25 °C/37 °C	37 °C
WT	0.2	0.2 >	+/+	+/+	+
vector	6.5	8.1	−/−	L/F	−
R61E	9.2	9.6	−/−	L/F	−
R61Q	10.4	13.7	−/−	L/F	−
R73E	1.4	1.3	+/+	+/L	+
R73Q	0.5	0.5	+/+	+/L	+
K75E	3.4	5.0	−/−	L/F	−
K75Q	0.5	1.0	+/+	L/L	+
K75D	6.2	12.7	−/−	L/F	−
K75A	0.5	1.1	+/+	+/L	+
K75F	2.5	7.7	+/−	L/L ~ F	+
K75L	2.4	9.3	+/−	L/L ~ F	+
K75R	1.2	2.1	+/+	L/L	+
K80E	4.8	14.8	−/−	L/F	−
K80Q	13.6	16.3	−/−	L/F	−
R96E	0.3	1.3	+/+	+/L	+
R96Q	1.2	0.8	+/+	+/L	+
R99E	8.4	9.7	−/−	L/F	+
R99Q	1.3	1.3	+/+	+/L	+
R105E	1.4	1.6	+/+	+/L	+
R105Q	0.2	0.4	+/+	+/L	+
R112E	8.2	5.1	−/−	L/F	−
R112Q	1.7	1.5	+/+	L/L	+
R112D	7.4	8.8	−/−	L/F	−
R112A	0.5	0.2 >	+/+	L/L	+
R112F	0.8	1.0	+/+	L/L	+
R112L	1.0	0.3	+/+	+/L	+
R112K	0.3	1.2	+/+	+/L	+
R114E	0.7	1.2	+/+	+/L	+
R114Q	0.2	0.5	+/+	+/L	+
K115E	1.1	0.7	+/+	L/L	+
K115Q	0.4	0.3	+/+	L/L	+
K119E	0.2 >	0.2 >	+/+	L/L	+
K119Q	0.7	0.2 >	+/+	L/L	+

^aThe frequency of anucleate cells in relation to the total cell count was determined from photographs as shown in Figs. 3, S3. The total cell count was at least 162 and usually exceeded 300 cells.

^bDefects in nucleoid segregation were assessed from the histogram showing the frequencies of anucleate cells (Fig. S2). In normal nucleoid segregation (+), the percentage is less than 3% at 25 °C and 37 °C.

^cThe cell shape was clarified from the level of cell elongation; cell size in the wild−type strain (+), a few folded cell lengths of WT (L), extremely elongated cells as filamentous (F).

^dCell viability was assessed from the results in Fig. 2.

Fig. 2. Colony formation test of the MukB head domain mutant-expressing cells.

a–c YAN4081 (ΔmukB::cat) harboring the plasmid encoding the indicated His₆-MukB mutant was cultured in L-broth at 25 °C. Cells were serially 10-fold diluted with saline and spotted on L-plates, followed by incubation at 25 °C or 37 °C for the indicated time. Also see Supplementary Fig. 1.

Fig. 3. Nucleoid segregation in cells expressing the MukB head domain mutants.

A series of YAN4081 (ΔmukB::cat) strains harboring the plasmid encoding the indicated His₆-MukB mutant is depicted as merged images of phase-contrast images and DAPI-stained fluorescent images (pseudo color in magenta). The images of cells cultured at 25 °C are displayed in the upper panels, and images at 37 °C are shown in the lower panels. The scale bar indicates 5 µm.

The plasmids encoding the mutated MukB protein with a single amino-acid substitution were introduced into the mukB-deletion mutant to assess the biological functions of the mutated MukB protein. The mukB-deletion mutant shows two representative phenotypes: temperature-sensitive growth and aberrant chromosome separation. Initially, we examined whether the amino acid-substituted mutants could suppress the temperature-sensitive growth of the mukB-deletion mutant cells. Cells harboring a plasmid encoding the his₆-mukB gene under the T7 promoter were incubated at the permissive (25 °C) or the restrictive temperature (37 °C). The plasmid encoding the his₆-mukB^wt enabled the deletion mutant to grow at 37 °C (Fig. 2a), even without addition of an inducer to the agar medium to activate transcription from the T7 promoter, although the accumulation level of the His₆-MukB^WT protein by leaked expression from the plasmid was lower than that of endogenous MukB (Supplementary Fig. 1). Similarly, we tested a total of 24 amino acid substitution mutants for their temperature-sensitive growth at 37 °C (Fig. 2a, b).

All the substitution mutants grew at 25 °C. Six mutants, R61E, R61Q, K75E, K80E, K80Q, and R112E were unable to compensate for the growth of the mukB-deletion mutation at 37 °C (Fig. 2a, b). These mutants were accumulated as comparable level as to that of His₆-MukB^WT (Supplementary Fig. 1). Neither the Glu substitution nor the Gln substitution for R61 and K80 could restore the temperature-sensitive growth. These findings suggest that the arginine residue at R61 and the lysine residue at K80 are absolutely necessary for the biological function of the MukB protein. Previously reported structural information suggests that these residues directly participate in and are essential for the ABC transporter-type ATPase activity^19,63. Therefore, the presence of positively charged residues at these positions is required for the proper functioning of the MukB protein.

On the other hand, the amino acid-substitution mutations K75Q and R112Q were able to compensate for the temperature-sensitive growth, while the substitution mutations K75E and R112E could not (Fig. 2a, b). To investigate whether the defects observed in the K75E- and R112E-substitution mutants were specifically due to the inversion of charge from positive to negative, we additionally constructed amino acid-substitution mutants of K75 and R112 using various residues: the negatively charged residue Asp (D), the non-charged residues Ala (A), Phe (F), and Leu (L), and the positively charged residues Arg (R) and Lys (K). The results revealed that the amino acid-substitution mutations K75D and R112D, which change the positive charge to a negative charge, failed to suppress the growth defect at 37 °C (Fig. 2c). Conversely, when the positive charge was maintained, as in the case of the amino acid substitutions K75R and R112K, the complementation activity of the MukB protein for temperature-sensitive growth remained unaffected. Additionally, substitutions with non-charged residues did not impact the activity. Therefore, it is concluded that only the presence of negatively charged amino acid residues at positions K75 and R112 had a detrimental effect on the biological function of the MukB protein with respect to temperature-sensitive growth.

The defect in chromosome separation by the amino acid substitution mutants

We next examined whether the amino acid substitution mutations could reduce the production of anucleate cells in the mukB gene-deletion strain. The mukB-deletion mutant displayed a high frequency of anucleate cell production, as well as cells with aberrant nucleoids, even when cultured at the permissive temperature. The frequency of anucleate cell production increased, and the cells became filamentous when cultured at the non-permissive temperature of 37 °C. These phenotypes of the mukB gene deletion were effectively suppressed by the plasmid encoding the N-terminal his-tagged mukB gene (Fig. 3). When the mukB-deletion mutant carried the vector plasmid, the frequency of anucleate cell production was 8.1% at 25 °C (Table 1, Fig. 3, Supplementary Fig. 2). In the case of the amino acid-substitution mutations R61E, R61Q, K75E, K80E, K80Q, R99E and R112E, these were unable to suppress the production of anucleate cells (Fig. 3 and Supplementary Fig. 3). These results were consistent with the findings from the temperature-sensitive growth analysis, except for the amino acid-substitution mutation R99E (Fig. 2a, b). The amino acid-substitution mutation R99E resulted in the production of anucleate cells regardless of the incubation temperature; however, filamentation at the restrictive temperature was remarkably suppressed by MukB^R99E protein. This indicates an improvement in only temperature-sensitive growth, as observed in Fig. 2a.

The DNA-binding activities of the amino acid substitution mutants

We next assessed the DNA-binding activities of the amino acid substitution mutants by means of an EMSA. The EMSA allows for the quantitative analysis of DNA-binding activities by measuring the delay in migration of DNA molecules caused by protein binding; purified MukB proteins bind to dsDNA and ssDNA. We quantified the intensity of DNA bands that did not bind to the purified MukB protein, and the DNA-binding activity was defined as the concentration of MukB protein required to shift half of the given amount of DNA, denoted as [MukB]₅₀. The [MukB]₅₀ values were calculated based on the binding kinetics of each mutated protein (Table 2 and Fig. 4 and Supplementary Figs. 4, 5). The wild-type His₆-MukB^WT protein exhibited [MukB]₅₀ values of 13.3 ± 4.1 nM for ssDNA and 12.0 ± 3.0 nM for dsDNA (Table 2 and Fig. 4).

Table 2.

The DNA-binding activities of the MukB head domain mutants to ssDNA and dsDNA

	Binding to ssDNA				Binding to dsDNA
	1st	2nd	3rd	mean ± SD	1st	2nd	3rd	Mean ± SD
WT	14	18	8	13.3 ± 4.1	15	9		12.0 ± 3.0
R61E	48	46		47.0 ± 1.0	27	26		26.5 ± 0.5
R61Q	21	18		19.5 ± 1.5	12	13		12.5 ± 0.5
R73E	23	24		23.5 ± 0.5	14	12		13.0 ± 1.0
R73Q	32	30		31.0 ± 1.0	26	23		24.5 ± 1.5
K75E	53	38		45.5 ± 7.5	18	16	16	16.7 ± 0.9
K75Q	9	12	8	9.7 ± 1.7	5	6	5	5.3 ± 0.5
K80E	12	21		16.5 ± 4.5	6	11		8.5 ± 2.5
K80Q	16	16		16.0 ± 0	12	14		13.0 ± 1.0
R96E	29	26		27.5 ± 1.5	46	22		34.0 ± 12.0
R96Q	14	18		16.0 ± 2.0	15	14		14.5 ± 0.5
R99E	78	93		85.5 ± 7.5	92	91		91.5 ± 0.5
R99Q	71	54		62.5 ± 8.5	420	295		357.5 ± 62.5
R105E	22	27		24.5 ± 2.5	26	16		21.0 ± 5.0
R105Q	14	17		15.5 ± 1.5	12	12		12.0 ± 0
R112E	72	88	69	76.3 ± 8.3	19	45	57	40.3 ± 15.9
R112Q	17	18	18	17.7 ± 0.5	8	16	11	11.7 ± 3.3
R114E	55	45		50.0 ± 5.0	26	30	25	27.0 ± 2.2
R114Q	22	24	15	20.3 ± 3.9	13	14	8	11.7 ± 2.6
K115E	110	105		107.5 ± 2.5	63	61		62.0 ± 1.0
K115Q	19	20		19.5 ± 0.5	11	22		16.5 ± 5.5
K119E	60	35		47.5 ± 12.5	70	49		59.5 ± 10.5
K119Q	8	14		11.0 ± 3.0	13	8		10.5 ± 2.5

[MukB]₅₀ values in two or three independent experiences are shown with the mean value ± standard deviations.

Fig. 4. The ssDNA-binding and dsDNA-binding activities of the MukB head domain mutants.

a Gel images of the electromobility shift assay (EMSA). The amount of purified His₆-MukB mutant proteins in each reaction is indicated in the top row. The left and the right panels show EMSA with cssDNA and cccDNA, respectively. Arrows indicate the positions of DNA to which the MukB protein does not bind. b–i Kinetics of the ssDNA-binding and dsDNA-binding activities of the MukB head domain mutants were investigated. A single amino-acid substitution of the MukB head domain mutants is indicated at the top of each graph. The intensities of the DNA bands to which the MukB protein does not bind in the EMSA gels in (a) are graphically shown according to inputs of the MukB protein, ranging from 0 to 150 nM. The DNA-binding activities of the wild-type MukB protein are shown as the common results for each panel. The means of the two or three independent experiments are graphically indicated as lines with individual plots shown as open circles. j, k The [MukB]₅₀ values are calculated based on the kinetics of the DNA-binding activities of the MukB head domain mutants in Fig. 4 and Supplementary Fig. 5. The ssDNA-binding and dsDNA-binding activities of the MukB head domain mutants, which are substituted to Glu, are indicated in (j). Those of the Gln-substituted mutants are indicated in (k). The mean values of the two or three independent experiments are indicated as bars with individual plots shown as squares.

The results for the Glu mutants revealed that the ssDNA-binding activities of the MukB proteins were decreased by the specific mutations; this was observed for each of the mutants, such as MukB^R61E, MukB^K75E, MukB^R99E, MukB^R112E, MukB^R114E, MukB^K115E, and MukB^K119E (Fig. 4). The [MukB]₅₀ values for these mutants were more than two-fold higher compared to that of MukB^WT. In the cases of MukB^R99E and MukB^K119E, the dsDNA-binding activities were also reduced to the same level as the ssDNA-binding activities (Fig. 4d, i, j). Hence, the substitution of Glu for these amino acid residues equally affected both the ssDNA and dsDNA-binding abilities.

On the other hand, for the other mutants, the ssDNA-binding activities were markedly decreased compared to the dsDNA-binding activities. For example, the [MukB]₅₀ value of MukB^K75E for ssDNA was 45.5 ± 7.5 nM, whereas it was 16.7 ± 0.9 nM for dsDNA, which was comparable to the [MukB]₅₀ value of MukB^WT (12.0 ± 3.0 nM) (Table 2 and Fig. 4c, j). In the case of MukB^K75E, the ssDNA-binding activity was remarkably decreased, while the dsDNA-binding activity was not distinctly affected. Similar trends were observed for MukB^R61E and MukB^R114E, although their dsDNA-binding activities were slightly worse than that for MukB^WT (Fig. 4b, g, j). Additionally, MukB^K112E and MukB^K115E exhibited distinct differences in DNA-binding activities between ssDNA and dsDNA. Their ssDNA-binding activities were considerably lower than that of MukB^WT, and the dsDNA-binding activities were also severely decreased (Fig. 4f, h, j).

In contrast to the results observed with the Glu substitution mutants, the [MukB]₅₀ values of the Gln substitution mutants were comparable to that of MukB^WT, except for those of MukB^R73Q and MukB^R99Q (Table 2and Fig. 4e, k and Supplementary Fig. 5d). Notably, the dsDNA-binding activities of MukB^R99Q were significantly reduced. The [MukB]₅₀ value of MukB^R99Q for dsDNA, 357.5 ± 62.5, was the highest among both the Glu and Gln substitution mutants. Furthermore, the ssDNA-binding activity of MukB^R99Q was also remarkably reduced, with a value of 62.5 ± 8.5 nM for ssDNA.

The topological DNA-binding activities of the amino acid substitution mutants

To assess the topological binding of MukB on DNA, we utilized the MU assay, which measures the amount of recovered DNA from a DNA-SMC-binding complex after eliminating the salt-sensitive binding of SMC protein to DNA⁴¹. In this assay, a circular single-stranded DNA substrate was incubated with purified histidine-tagged MukB under a low-salt condition (25 mM KCl). The DNA substrates bound to histidine-tagged MukB were then recovered using affinity beads that bind to the histidine-tag, followed by washing with a high salt buffer solution (including 750 mM KCl). The recovered DNA substrates were analyzed by agarose gel electrophoresis, specifically by pulling down with histidine-tagged MukB (Fig. 5). By employing this method, we were able to evaluate the topological binding activity of the mutated MukB protein.

Fig. 5. Topological DNA-binding activities of MukB mutants.

DNA retrieval rates were measured from gel images of the topological DNA-binding assay. The mean values of calculated DNA retrieval rates of two or three independent experiments are indicated as bars, with individual plots shown as squares. Circular single-stranded (css) DNA of pUC119 was used as substrate DNA in (a). Circular single-stranded or linear single-stranded (lss) DNA of pUC119 was used as substrate DNA in (b, c). The enhanced image was shown in the bottom panel in (b) to indicate the detection limit level bands. The mean values of the two or three independent experiments are indicated as bars with individual plots shown as squares.

We investigated whether the defects in the ssDNA-binding activities observed in the Glu substitutions (MukB^R61E, MukB^K75E, MukB^R112E, MukB^R114E, and MukB^R115E) also affected their topological DNA-binding activities. Our experiments confirmed that all these Glu-substituted mutants exhibited decreased DNA recovery rates (Fig. 5a), indicating that these mutations weakened the topological binding activities. In contrast, the topological-binding activities of the Gln-substituted mutants remained equivalent to those of the WT protein. When linear ssDNA was used instead of circular DNA, no DNA molecule was pulled down (Fig. 5b). This result strongly suggests that the recovered cssDNA in this assay was a topologically entrapped product. Collectively, the results indicate that the presence of a negative charge of the amino acid residues affects the topological-binding activity.

The mutants MukB^R99E and MukB^R99Q also affected the ssDNA-binding affinity in the EMSA assay. These substitutions differed from those of MukB^R61E, MukB^K75E, MukB^R112E, MukB^R114E, and MukB^R115E in that they affected dsDNA-binding as much or more than they affected ssDNA-binding. Topological-binding activities of MukB^R99E and MukB^R99Q for the circular ssDNA were slightly lower than that of MukB^WT but were maintained significantly (Fig. 5c). The finding that no linear ssDNAs were retrieved showed that the DNAs were topologically entrapped (Fig. 5c).

Discussion

Although some SMC proteins exhibit ssDNA-binding ability, the biological significance of their ssDNA-binding ability remains poorly understood. In this study, our objective was to identify the specific amino acid residues responsible for the ssDNA-binding ability of the MukB protein. We successfully identified several candidate residues, namely R61, K75, R112, R114, and K115, located in the head domain of MukB, which appear to contribute significantly to the preferential ssDNA-binding activity.

Substituting Glu for these amino acid residues resulted in noticeable differences in the DNA-binding abilities between ssDNA and dsDNA. Notably, the K75E mutation alone reduced the ssDNA-binding ability while preserving the dsDNA-binding ability in vitro. The substitution of Glu for three of these crucial amino acid residues, R61, K75, and R112, significantly impacted all the phenotypes of cell growth, nucleoid separation, and topological binding.

In contrast, the Glu substitutions for the amino acid residues R114 and K115 did not affect cell growth or nucleoid separation. Nevertheless, the introduction of a negative charge in these amino acid residues resulted in a reduction in the ssDNA-binding ability, rather than the dsDNA-binding ability, as well as a reduction in the topological binding ability onto ssDNA in vitro. The amino acids residues R114 and R115 locate at the basement of R112, and they support the protrusion of the charged residue R112 from the MukB head (Fig. 6). Hence, it is reasonable to hypothesize that the amino acid residues R61, K75, and R112 collectively form the primary site for the ssDNA binding, while the two amino acid residues R114 and R115 assist in the efficient binding of ssDNA to the primary site.

Fig. 6. Model of the stable topological binding of MukB using the residues R61, K75 and R112.

a The MukB protein exists in a dynamic equilibrium between open and closed conformations, driven by thermal fluctuations. In the open state, ssDNA regions—melted from chromosomal dsDNA—can incidentally enter the ring structure of the MukB dimer. Upon closure of the head domains, a composite ssDNA-binding site, comprising six residues contributed by both subunits, is formed, allowing stable retention of ssDNA within the ring. Subsequently, MukEF proteins and ATP are incorporated to complete the functional cycle of the MukBEF complex. Ultimately, MukBEF entraps chromosomal dsDNA in a topologically stable manner. Magenta; ssDNA binding region including the residues R61, K75 and R112. b–g A 3D model of the head domain of Haemophilus ducreyi MukB (PDB-ID: 3EUK) was made using a 3D printer (KawakamiModel; http://studio-midas.com/kawakamimodel/). b In two monomers, the 6 residues R61 (orange), K75 (magenta) and R112 (yellow) in each head are represented from the top view. c When the MukB dimer closes, these residues align within the inner surface of the closed MukB ring. d After MukB captures ssDNA inside its ring, ssDNA binds to the inner surface through the ssDNA-binding domain formed by aligned axis residues. The ssDNA crossing the MukB heads stabilizes the state of the engaged MukB head. The twisted strands (red and white) represent DNA with a melted part. e The residues R99 (blue) and R119 (green) are represented along with the six residues R61 (orange), K75 (magenta) and R112 (yellow). f The residues R99 (blue) and R119 (green) are represented from the side view. g The residues R99 and R119 hold ssDNA or dsDNA.

The substitutions of the negatively charged Glu for R99 and K119 resulted in an equal reduction of binding ability to both dsDNA and ssDNA in vitro. Furthermore, substitution of the neutrally charged Gln for R99 led to the loss of these DNA-binding abilities, with the dsDNA-binding ability being the least weakened among all the amino acid substitutions in this experiment. However, despite the decrease in DNA-binding abilities, the Glu and Gln substitutions for R99 and K119 never showed defects in cell growth at high temperatures. These results suggest that DNA binding at residues R99 and K119 is not essential for the fundamental biological functions of MukB. An alternative possible explanation for the markedly reduced DNA-binding activity observed in the R99E and R99Q mutant proteins is that they may be partially misfolded due to the absence of chaperones during purification.

Taken together, these results suggest the likelihood that there are at least two DNA-binding domains within the MukB head. One involves the alignment of R61, K75, and R112 for an essential DNA binding, possibly more important for ssDNA binding. The other domain consists of R99 and K119, which are involved in a non-essential DNA binding to ssDNA and dsDNA, especially to dsDNA. This non-essential DNA binding to dsDNA could potentially facilitate maintaining melted DNA onto the MukB dimer as shown below. Alternatively, it could contribute to the distribution of MukB to the whole chromosome region except the ter domain. Our results suggest that R61, K75, and R112 are involved in ssDNA binding. However, it should be noted that these residues may also affect dsDNA binding, which is functionally important.

As to why the MukB dimer exhibits a preference for entrapping ssDNA over dsDNA in vitro, we propose the following scenario based on the spatial position of the above-described residues within the protein structure model of the MukB head. The in vitro assay showed that a purified MukB dimer repeatedly opens and closes its ring structure through thermal motion rather than ATP hydrolysis⁴¹ (Fig. 6a). The residues R61, K75 and R112 are positioned on the inner surface of the MukB head (Fig. 6b). When the head domains of the MukB dimer bind with each other and form a closed form of the head, these six amino acid residues align within the inner surface of the closed MukB head (Fig. 6c). Consequently, it is plausible that the alignment of R61, K75, and R112 forms an axis of positively charged amino acid residues, functioning as a preferential binding site for ssDNA only when the MukB ring is closed.

Consider the case in which the axis amino acid residues bind to entrapped ssDNA within the ring of the MukB dimer. In that case, the ssDNA can span across the closed dimer head, thereby stabilizing the closed form while the ssDNA interacts with the axis amino acid residues (Fig. 6d). Furthermore, R99, which is located diagonally below R112 (Fig. 6e, f), may assist in additionally binding the extended ssDNA in the appropriate orientation against the MukB head (Fig. 6g). Overall, in the ssDNA-bound MukB dimer, it appears that the two heads are connected by ssDNA, which acts as a linking element between them (Fig. 6d, g). This working model is based on in vitro experiments, and the functional relevance of ssDNA binding in vivo has not yet been established. Although we attempted to obtain supporting in vivo data through ChIP analysis, these efforts were unsuccessful. Further investigation is required to evaluate the biological significance of ssDNA binding in living cells.

Regarding the DNA-binding behavior of MukB, two possibilities should be considered: MukB may bind DNA either independently or as part of the MukBEF complex. In earlier studies, it was thought that only the MukB core unit within the MukBEF complex possesses DNA-binding activity and MukB can localize to chromosomes on its own^4,25,26. However, a recent study⁶⁴ demonstrated that MukE also directly binds to DNA and that this interaction is important for the loading reaction. In the model proposed by Burmann et al.⁶⁴, MukE binds to DNA to form a “capture state,” after which the DNA is topologically loaded through the neck gate formed between the neck of MukB and MukF. In contrast, our previous work showed that purified MukB protein alone can topologically entrap DNA—particularly single-stranded DNA (ssDNA)—in the absence of MukEF and ATP⁴¹. This ssDNA-specific topological binding by MukB alone may reflect an alternative mechanism to that proposed by Burmann et al.⁶⁴. In the model we propose, MukB first binds to ssDNA through residues located in its head domain during the initial step of the MukBEF functional cycle, followed by the recruitment of MukEF and ATP. Given the complex activities of MukB, it is plausible that its mode of DNA binding varies depending on its functional context. The ability of MukB to bind ssDNA, as demonstrated in our study, may thus represent one facet of its multifaceted DNA-binding properties. A parallel can be drawn with the Bacillus subtilis Smc–ScpAB complex, for which two independent chromosome-loading mechanisms have been proposed: one involves loading at the parS site near the oriC region via the parS–ParB system^65,66, followed by translocation toward the ter region while aligning the chromosome arms^47,67, with unloading mediated by XerD⁶⁸. The other involves loading at ssDNA regions generated at highly transcribed rDNA loci^42,58. These findings raise the possibility that E. coli may also utilize multiple mechanisms for chromosomal localization of MukBEF.

The substitutions R61Q, K75Q, and R112Q maintained the DNA-binding activities and topological binding activities at the wild-type level. However, the R61Q substitution alone caused defects in cell growth and nucleoid segregation. The structural information suggests that the amino acid residue equivalent to R61 is directly involved in the hydrolysis of ATP of the ATP-binding cassette transporters^19,63. The R61 amino acid residue corresponds to the residue of the Arg finger, which is the motif involved in ATPase hydrolysis⁶³. In the case of the R61Q substitution, it is conceivable that the residue is involved not only in DNA-binding ability but also in the ATPase activity, which is necessary for the complete biological function of MukB. Furthermore, the K80 amino acid residue is also involved in the ATPase activity since it is positioned in such a way that its side chain faces an ATP molecule, as indicated by the crystal structure of the ATP-binding cassette transporter¹⁹. Therefore, it is reasonable that both the K80E and K80Q substitutions caused defects in cell growth and nucleoid segregation, although they did not affect the DNA-binding activities.

In Vazquez-Nunez et al.⁶⁹, it was shown that combinations of three or more alanine substitutions in the inner surface–exposed region of the B. subtilis Smc head domain reduced dsDNA-binding activity and abolished function. Although those mutants were still able to localize on the chromosome, their DNA translocation activity was impaired, leading to the accumulation of several kilobases away from the loading site. The mutagenized region examined by Vazquez-Nunez et al.⁶⁹ was almost identical to that analyzed in the present study. Our results are consistent with theirs in showing that the region inside the ring of the SMC core protein is critical for DNA binding. The main distinctions are that we used circular ssDNA or dsDNA substrates rather than short 40-mer double-stranded oligonucleotides, we introduced single amino acid substitutions to Glu or Gln rather than multiple alanine substitutions, and, whereas DNA binding in the head domain of B. subtilis Smc-ScpAB is required for DNA translocation from parS sites near oriC toward the ter region, no parS sequences are found in E. coli. It is therefore likely that DNA binding at the inner surface of the head domain is a common property of bacterial condensin Smc core proteins, although the specific roles of this activity may not be entirely identical, even if they are similar in certain respects. In Bacillus subtilis, it has been reported that the phenotype associated with deletion of the smc gene correlates more strongly with growth rate—particularly the velocity of the replication fork—than with temperature⁷⁰. When replication fork progression is slow, the phenotype resulting from smc deletion is less severe. Similarly, in Escherichia coli, mukB-deletion mutants exhibit filamentation and defects in chromosome segregation under high-temperature conditions in rich medium. However, the underlying mechanisms by which temperature or growth rate influences the phenotypes of mukB deletion mutants remain poorly understood. Interestingly, the R99E amino-acid substitution mutant was found to suppress the temperature-dependent cell growth observed in the mukB-deletion mutant. However, this mutant still produced filamentous cells and exhibited errors in chromosome segregation at a high temperature. Therefore, cell elongation was not the sole cause of the temperature-dependent cell growth. It appears that the R99 amino acid residue is involved in a latent and unsuspected function of MukB that contributed to this phenomenon.

The B. subtilis Smc-ScpAB complex has been observed to accumulate on the actively transcribed ribosomal RNA (rrn) operon⁴². Within this operon, an ssDNA segment that facilitates the accumulation of Smc-ScpAB has been identified⁵⁸. Hence, it is plausible that preferential topological loading onto ssDNA could be utilized to promote the accumulation of bacterial condensin in specific regions, such as the rrn operon. This is consistent with previous ChIP analysis by Gruber and Errington⁶⁵ that showed the enrichment of the Bs Smc on rrn regions. Another ChIP analysis showed that yeast condensin accumulated at the highly transcribed regions, including rDNA⁵⁷. The ssDNA-targeted loading is a common feature for both the prokaryotic and eukaryotic condensins.

A recent study⁷¹ demonstrated that MukBEF is associated with newly replicated DNA, although it can also bind chromosomes independently of DNA replication. These findings suggest that MukBEF binds to chromosomes through two distinct mechanisms: one that is coordinated with replication fork progression and another that is independent of it. During DNA replication, ssDNA regions are transiently exposed on the lagging strand. The ssDNA-specific binding activity of MukB observed in our experiments may therefore play a role in chromosome association events that are coupled to replication.

The ssDNA-binding activity of MukB demonstrated in this report represents an important initial step in the functional cycle of bacterial condensing (Fig. 6a). To gain deeper insight into the mechanism of nucleoid compaction, it is essential to investigate the subsequent behavior of MukB following its loading onto ssDNA regions. A comprehensive understanding of this cycle further requires elucidation of the roles of MukEF, the non-SMC subunits, and ATP hydrolysis.

While our biochemical assays demonstrate residue-specific differences in ssDNA versus dsDNA binding, it remains to be determined whether these differences translate into functionally distinct roles in vivo. Further studies, including in vivo binding assays or chromosomal localization analysis, are required to fully elucidate the biological relevance of these findings.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Conceptualization, K.A. and H.N.; investigation, K.A. and H.N.; resources, K.Y.; writing–original draft, K.A. and H.N.; writing–review & editing, K.A. and H.N.; supervision, H.N.; funding acquisition, K.A. and H.N.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Joanna Timmins and Tobias Goris. A peer review file is available.

Data availability

All data supporting the findings of this study are available within the paper and its Supplementary information. Source data underlying Figs. 4, 5 and Supplementary Fig. 5 can be found in Supplementary Data 1. The unedited gel images of Fig. 4 and Supplementary Fig. 5 are shown in Supplementary Fig. 6. The unedited gel images of Fig. 5 are shown in Supplementary Fig. 7. Additional information is available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-025-09345-5.