Transient DNA Occupancy of the SMC Interarm Space in Prokaryotic Condensin

Abstract

Multi-subunit SMC ATPases control chromosome superstructure and DNA topology, presumably by DNA translocation and loop extrusion. Chromosomal DNA is entrapped within the tripartite SMCkleisin ring. Juxtaposed SMC heads ("J heads") or engaged SMC heads ("E heads") split the SMCkleisin ring into "S" and "K" sub-compartments. Here, we map a DNA-binding interface to the S compartment of E heads SmcScpAB and show that head-DNA association is essential for efficient DNA translocation and for traversing highly transcribed genes in Bacillus subtilis. We demonstrate that in J heads, SmcScpAB chromosomal DNA resides in the K compartment but is absent from the S compartment. Our results imply that the DNA occupancy of the S compartment changes during the ATP hydrolysis cycle. We propose that DNA translocation involves DNA entry into and exit out of the S compartment, possibly by DNA transfer between compartments and DNA segment capture.

Keywords: DNA loop extrusion; MukB; Rad50; SMC; Smc5; SmcScpAB; chromosome condensation; cohesin; condensin; kleisin.

Figure 1. Locating DNA duplexes in J heads Smc/ScpAB

(A) Schematic view of J heads (left panel) and E heads Smc/ScpAB (right panel). Cysteine residues are indicated as dots in black colors. The carboxy-terminal HaloTag fusion on Smc is omitted for simplicity. (B) Identification and quantification of cross-linked species of Smc-HaloTag (‘Smc-HT’; labelled with HaloTag-TMR) using cysteine pairs in isolation and combination. Fully cross-linked species of Smc-HT are indicated in magenta colors and denoted as ‘S/K’ for the SMC/kleisin ring, ‘J-S’ for the J-S compartment and ‘J-K’ for the J-K compartment. Species with Smc hinge cross-linked only (‘S’) or J heads cross-linked only (‘J’) are labelled in blue colors, while the species having both Smc/ScpA interfaces cross-linked only (‘K’) is shown in yellow colors. All other species are indicated in black colors: Smc-neck/ScpA-N (‘N’), Smc-cap/ScpA-C, combination of J-Cys and Smc-neck/ScpA-N (‘JN’), combination of J-Cys and Smc-cap/ScpA-C (‘JC’), combination S-Cys and Smc-neck/ScpA-N (‘SN’), combination S-Cys and Smc-cap/ScpA-C (‘SC’). Species ‘X’ and ’Y’ are derived from alternative Smc/ScpAB having ScpA associated with neck and cap of a single Smc protomer. Annotation of X, SN, SC and S/K species with help of published data (Bürmann et al., 2013; Wilhelm et al., 2015). Mean fraction (and standard deviation) of species N, C, K, S and J were quantified from three replicate experiments (Table S2). The fraction of J-S, J-K and S/K species was calculated using efficiencies determined for isolated J-Cys, K-Cys and S-Cys residues as appropriate. (C) Chromosome entrapment assay in agarose plugs with J-Cys strains. Cells harboring Smc-HT were cross-linked with BMOE using cysteine pairs at the indicated protein-protein interfaces and incubated with HaloTag-OG substrate (‘input’). Intact chromosomes were isolated in agarose plugs. Input material and co-isolated proteins (‘eluate’) were analyzed by SDS-PAGE and in-gel fluorescence detection. K37I, Walker A ATP binding mutant of Smc (‘KI’). (D) Chromosome entrapment assay in agarose microbeads with J-Cys strains. As in (C) using agarose microbeads and HaloTag-TMR. A mix of cells carrying wt DnaN (80 %), DnaN-HT (10 %) and DnaN(Cys)-HT (10 %) is included as positive and negative control for entrapment. Note: untagged cells are added to reduce the DnaN-HT signal to the levels of Smc-HT. Arrowheads in magenta colors denote circular protein species. ‘XX’ indicates the double cross-linked, circular DnaN-HT species. The single cross-linked, X-shaped DnaN-HT species exhibits slightly slower mobility (labeled in black colors) (Wilhelm et al., 2015). See also Figure S1 and Table S2.

Figure 2. Hinge/DNA association is dispensable

(A) Schematic representation of E heads Smc/ScpAB with DNA bound to hinge and heads. Schematic view of the SmcH-CC100 construct carrying the K666A, K667A, K668A (‘3A-hng’) mutation. Surface-charge electrostatic potential is shown for the coiled coil-proximal surface of the Bs Smc hinge (modelled from PDB: 1GXL). Scheme for DNA binding to the coiled coil-proximal surface of the hinge. (B) Coomassie Brilliant Blue (‘CBB’) staining of purified SmcH-CC100 protein. The asterisk denotes an impurity present in the 3A-hng variant. Size of marker proteins (lane M) is given in kilodalton (kD). (C) DNA binding as measured by fluorescence anisotropy with increasing protein concentration using fluorescein-labelled 40 bp dsDNA at 50 nM concentration. Data points from three experiments are shown as dots. The lines correspond to a non-linear (see STAR*Methods). Note: the substoichiometric amounts of a protein impurity in 3A-hng (see in (B)) may contribute to the observed residual DNA binding but is unlikely to strongly hinder DNA binding of the more abundant SmcH-CC100(3A-hng) protein. (D) Colony formation by smc(3E-hng) and smc(3A-hng) mutant strains. Dilutions (81-fold and 59,000-fold) of overnight cultures were spotted on nutrient-rich medium (ONA) and nutrient-poor medium (SMG) and grown at 37 °C for 16 and 24 hours, respectively. See also Figure S2 and Table S3.

Figure 3. Identification of surface-exposed Smc head residues required for Smc function.

(A) Alignment of N-terminal sequences of five bacterial Smc proteins (Bs, Bacillus subtilis; Gs, Geobacillus stearothermophilus; Sp, Streptococcus pneumoniae; Tm, Thermotoga maritima; Dr, Deinococcus radiodurans). The conserved Walker A box motif is indicated for reference. Residues chosen for detailed analysis are denoted. (B) Characterization of smc alleles harboring triple alanine mutations. Top panel: Colony formation by dilution (81-fold and 59,000-fold) spotting as in Figure 2D. Bottom panels: Cellular expression levels of Smc variants determined by immunoblotting with polyclonal antibodies raised against Bs Smc. CBB staining of cell extracts on separate gels is shown as control for uniform protein extraction. (C) Smc alleles harboring selected quadruple mutations and the quintuple alanine mutation. As in (B). (D) Surface representation of the structure of the Gs SmcHd-ATPγS complex (PDB:5H68) (in gray colors) superimposed onto Rad50Hd-ATPγS-DNA (PDB:5DAC) (only DNA is shown - in yellow colors) (Kamada et al., 2017; Seifert et al., 2016). The side chains of putative DNA binding residues are marked in blue colors. ATPγS is shown in stick representation in red colors. Residues mutated in 3A-hd are marked by boxes (left panel only). Surface-charge electrostatic potential is shown for the Gs SmcHd-ATPγS complex (PDB:5H68) (top right panel). See also Figure S3

Figure 4. ATP dependent head/DNA association.

(A) Construction of SmcHd/ScpA-N with and without the ‘3A-hd’ mutations (K60A, R120A, K122A). The amino-terminal His6-tag on ScpA-N is omitted from the scheme for simplicity. (B) CBB staining of purified preparations of SmcHd/ScpA-N constructs with and without EQ and 3A-hd mutations. (C) Gel filtration profiles of SmcHd/ScpA-N preparations eluted from a Superdex 200 10/300 column. Samples were eluted in buffer supplemented with or without 1 mM ATP. Elution curves for proteins without and with the EQ mutation are shown in the left and right panels, respectively. (D) DNA binding measurements using fluorescence anisotropy titrations as in Figure 2C. Data points for proteins without and with the EQ mutation are displayed in the top and bottom panels, respectively. Data points and fits for three experiments are shown. The estimated K_d for head/DNA association is calculated under the assumption of complete dimerization of SmcHd(EQ)/ScpA-N. The actual K_d may thus be somewhat lower. See also Figure S4 and Table S3.

Figure 5. Rates of ATP hydrolysis by Smc heads and Smc holo-complexes

(A) ATP hydrolysis rate of SmcHd/ScpA-N at 10 μM protein concentration and increasing concentrations of ATP as measured by an enzyme-coupled assay. The lines represent a non-linear regression fit of the Hill equation (see STAR*Methods). Data points and fits for two experiments are shown. (B) CBB staining of purified preparations of untagged, full-length Smc, ScpA and ScpB proteins. (C) ATP hydrolysis rates of full-length Smc proteins and holo-complexes at 0.15 μM Smc dimer concentration and equivalent concentrations of ScpB dimers and ScpA monomers. As in (A) using protein preparations shown in (B). DNA stimulation was measured by addition of 3 μM 40 bp dsDNA. Data points and fits from three experiments are displayed. See also Figure S5, and Table S5.

Figure 6. Chromosomal association and distribution of Smc(3A-hd).

(A) Chromosome entrapment assay using agarose microbeads. Smc-HT and DnaN-HT species labelled by HaloTag-TMR were analyzed by SDS-PAGE and in-gel fluorescence detection. As in Figure 1D. All Smc-HT samples harbor K-Cys and S-Cys for cross-linking of the two Smc/ScpA interfaces and the Smc hinge, respectively. The circular S/K species is indicated in magenta colors. All other labeling as in Figure 1B. (B) ChIP-Seq profiles for the replication origin region represented as normalized reads per thousand total reads. ChIP was performed with a polyclonal antiserum raised against the ScpB protein. Gene organization in the origin region is shown in the ΔscpB panel. Genes in highly-transcribed operons are displayed in green colors. Gray lines indicate the 3’ end of highly transcribed operons. The position of parS-359 is given by a dashed line. (C) Ratiometric analysis of wild-type and mutant ChIP-Seq profiles shown in (B). For bins with read counts greater than the wild-type sample, the ratio was plotted above the genome coordinate axis (in blue colors). Otherwise the inverse ratio was plotted below the axis (in yellow colors). (D) α-ScpB ChIP-qPCR. Selected loci in the replication origin region and at the terminus were analyzed by quantitative PCR. Each dot represents a data point from one out of three experiments. The solid boxes span from the lowest to the highest value obtained; the horizontal line corresponds to the mean value calculated from the three biological replicates. See also Figure S6.

Figure 7. Hinge and head mutations promote S compartment opening and DNA entrapment.

(A) Dilution (81-fold and 59000-fold) spotting of 3A-hng, 3A-hd and 6A mutants. As in Figure 2D. A corresponding ONA image after 12 h incubation is shown in Figure S7A. (B) Cysteine cross-linking of purified SmcH-CC100(A715C) protein in presence and absence of 10 μM 40 bp dsDNA. The asterisk marks an impurity; see also Figure 2B. (C) Identification and quantification of cross-linked species of E heads Smc/ScpAB. using cysteine pairs in isolation and combination. As in Figure 1B. Circular species are indicated in magenta colors and denoted as ‘S/K’ for the SMC/kleisin ring and ‘E-S’ for the E-S compartment. Species with E heads cross-linked only (‘E’) are labelled in blue colors. All other labeling as in Figure 1B. Note that species S and E-S co-migrate in the gel. Mean fraction (and standard deviation) of species E in Smc(EQ) was quantified from three replicate experiments. Quantification of all other species as in Figure 1B (Table S2). The fraction of E-S species was calculated under the assumption of normal hinge dimerization in E heads Smc(EQ)/ScpAB (Bürmann et al., 2017). (D) DNA entrapment in the E-S compartment of Smc(EQ)/ScpAB. Chromosome entrapment assay with agarose microbeads. As in Figure 1D. Arrowheads in magenta color denote circular protein species, S/K for the SMC/kleisin ring and E-S for the E-S compartment. (E) Summary of main findings. (F) Models for DNA transport cycles: DNA entry into the S compartment by passage between separated Smc heads is shown on the left-hand side (‘Back-and-forth’). DNA entry into the S compartment during head engagement by loop capture displayed on the right-hand side (‘Roundabout’). See also Figure S7 and Table S2.