Interlinked sister chromosomes arise in the absence of condensin during fast replication in B. subtilis

Abstract

Condensin-an SMC-kleisin complex-is essential for efficient segregation of sister chromatids in eukaryotes [1-4]. In Escherichia coli and Bacillus subtilis, deletion of condensin subunits results in severe growth phenotypes and the accumulation of cells lacking nucleoids [5, 6]. In many other bacteria and under slow growth conditions, however, the reported phenotypes are much milder or virtually absent [7-10]. This raises the question of what role prokaryotic condensin might play during chromosome segregation under various growth conditions. In B. subtilis and Streptococcus pneumoniae, condensin complexes are enriched on the circular chromosome near the single origin of replication by ParB proteins bound to parS sequences [11, 12]. Using conditional alleles of condensin in B. subtilis, we demonstrate that depletion of its activity results in an immediate and severe defect in the partitioning of replication origins. Multiple copies of the chromosome remain unsegregated at or near the origin of replication. Surprisingly, the growth and chromosome segregation defects in rich medium are suppressed by a reduction of replication fork velocity but not by partial inhibition of translation or transcription. Prokaryotic condensin likely prevents the formation of sister DNA interconnections at the replication fork or promotes their resolution behind the fork.

Figure 1

Conditional Inactivation of Condensin Blocks Nucleoid Segregation during Spore Germination (A) Colony formation of BSG1002 (smc⁺), BSG1007 (Δsmc), BSG1005 (scpAB⁺), BSG1004 (ΔscpA), and BSG1489 (ΔscpB) on SMG and NA plates at 37°C. Overnight cultures in SMG were diluted 81-fold and 60,000-fold. We spotted 5 μl of each dilution onto an agar plate. (B) Colony formation of BSG165 (scpA(pk)), BSG224 (scpA(pk), P_xyl-tevP), BSG195 (scpA(tev-pk)), and BSG225 (scpA(tev-pk), P_xyl-tevP) on nutrient agar at 37°C in the presence and absence of 1% xylose (left and right panels, respectively). Relevant genotypes of strains are indicated above the images: – and + denote the absence and presence, respectively, of the TEV protease gene (tevP); pk and tev-pk mark respective insertions in scpA. Serial dilutions of an overnight culture in LB were used as in (A). (C) Scheme of spore germination in B. subtilis. (D) TEV cleavage prevents accumulation of ScpA during spore outgrowth. Spores of strains BSG224 and BSG225 were heat activated for 30 min at 70°C and grown at 37°C. Samples were taken every 30 min for the preparation of whole-cell extracts. Proteins were immunoblotted using antibodies directed against the PK epitope tag and MreB protein. (E) Time-lapse microscopy of an equal mixture of germinating spores of strains BSG221 (scpA(tev-pk), hbs-gfp) and BSG222 (scpA(tev-pk), P_xyl-tevP, hbs-gfp). Spores displaying wild-type and mutant phenotypes are labeled as –TEVp and +TEVp, respectively. Cells were incubated on LB agar pads supplemented with glucose and xylose to induce TEVp expression (LBG) at 30°C. Hbs-GFP and phase-contrast signals are shown in green and gray colors, respectively. Outgrowing spores harboring tevP and hbs-gfp display an unusually large diameter. This seems to be caused at least in part by Hbs-GFP, because unlabeled cells or cells labeled otherwise do not show this pronounced shape defect.

Figure 2

Inactivation of Smc-ScpAB Prevents Segregation of oriC (A) Time-lapse microscopy of an equal mixture of germinating spores of strains BSG203 (P_xyl-tevP, parB-gfp) and BSG204 (scpA(tev-pk), P_xyl-tevP, parB-gfp). Spores encoding wild-type ScpA and ScpA containing a TEV site are labeled as WT and TEVs, respectively. LBG agar pads were incubated at 37°C. ParB-GFP and phase-contrast signals are shown in green and red colors, respectively. (B) Germinating spores of strain BSG239 (P_spac-scpA(tev-pk)scpB, P_xyl-tevP, parB::tetOx240, tetR-yfp) grown in liquid LBG media in the presence and absence of 1 mM isopropyl-thiogalactopyranosid (IPTG) at 37°C. YFP images were taken 210 min after heat activation of spores.

Figure 3

Rescue of Lethal Growth Defect of Δsmc by Small Molecules Growth of BSG1002 and BSG1007 in microtiter plates containing LB medium supplemented with different concentrations of the following antibiotics: (A) Chloramphenicol. (B) Streptolydigin. (C) Arginine hydroxamate. (D) Hydroxyurea. Growth at 37°C was monitored by light scattering at a wavelength of 620 nm.

Figure 4

Rescue of the Segregation of oriC at Artificially Low Replication Fork Velocities (A and B) Time-lapse microscopy of strains BSG448 (parB-gfp, Δsmc) grown on LB agar pads at 30°C in the absence (A) and presence (B) of 8 mM hydroxyurea. ParB-GFP and phase-contrast signals are shown in green and red colors, respectively. Corresponding images of cells with a wild-type smc allele are shown in Figure S4. (C) Growth and replication characteristics of BSG1002 and BSG1007 at 37°C. Cells were grown to exponential phase in SMG and diluted 1:11 into SMG medium or LB medium with 8 mM HU, with 1.2 μg/ml chloramphenicol, or without inhibitors. Doubling times (Td) and oriC/ter ratios were determined experimentally. Standard deviation from mean values was calculated from at least three independent replicates. Population average replication periods were estimated on the assumption of uniform replication of the chromosome and balanced growth based on the following equation [26]: C_(min) = ln(oriC/ter)/ln(2) × Td_(min). Estimates derived for cells displaying potentially nonuniform DNA replication and/or unbalanced growth under the given conditions are shown in parenthesis.