henhaobio
Sign in

bioRxivBiophysicsPreprint

DNA gyrase in live bacteria forms liquid condensates through weak multivalent bonding of excess GyrB

Aisha H. SyedaiD, Katy Hollands, Lewis Frame, Jack W ShepherdiD, Alex L. Payne-DwyeriD, Eleanor Goffee, Nicolas P. BurtoniD, Aakash BasuiD, Agnes NoyiD, Tony Maxwell, Mark C. LeakeiD

PDFDOICode
In-depth full-text digest available · Jump to digest ↓

Abstract

Type IIA bacterial topoisomerase DNA gyrase, a GyrA/GyrB heterotetramer, has crucial roles maintaining transcription and DNA replication by relaxing positive DNA supercoils through introducing negative supercoils. However, rates of gyrase-catalysed supercoiling in vitro cannot explain much higher rates required in vivo. To address this puzzle, we used high-speed single-molecule fluorescence imaging of GyrA/GyrB reporters in live Escherichia coli, indicating that cells contain [~]40% more GyrB than GyrA expressed in a diffuse pool or in clusters whose mobility depends on whether they are bound to DNA. Unexpectedly, we discovered that clusters are non-stoichiometric containing [~]150% more GyrB than GyrA, significantly greater than the cellular average, with fluorescence recovery after photobleaching revealing that clustered GyrA and GyrB behave as a liquid whose abundance can be increased by applying gyrase-targeting antibiotics. Structural docking indicates that the liquid state is stabilised through excess GyrB progressively binding to existing clusters via weak, multivalent interactions. By operating in liquid condensates, A2B2 that dissociates from DNA can rebind rapidly instead of diffusing away, increasing enzyme processivity to enable multiple rounds of catalysis that can keep pace with transcription and DNA replication in vivo. This demonstrates a new role for condensates of overcoming kinetic limitations imposed by diffusion. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=158 SRC="FIGDIR/small/744598v1_ufig1.gif" ALT="Figure 1"> View larger version (53K): org.highwire.dtl.DTLVardef@1514d4aorg.highwire.dtl.DTLVardef@a95ceeorg.highwire.dtl.DTLVardef@11417daorg.highwire.dtl.DTLVardef@2a2856_HPS_FORMAT_FIGEXP M_FIG C_FIG

In-depth summary

Claude Sonnet 4.6 · max effort · 8 Nov 2026 · from the full text of version 2 (19 Aug 2026) · CC BY 4.0

Live E. coli gyrase clusters hold excess GyrB (B:A 2.5:1), show hexanediol-sensitive liquid-like turnover, and docking suggests GyrB bridges heterotetramers.

Replication generates positive supercoils far faster than one gyrase can remove them at about 1 per second, implying many enzymes per fork. This work proposes that liquid condensates stabilised by excess GyrB could explain how gyrase keeps up in live cells.

Key results

  1. ShownLive E. coli cells carry roughly 40% more GyrB than GyrA when both are counted by stepwise photobleaching in a dual-labelled strain.

    Mean 471 ± 16 GyrA vs 656 ± 26 GyrB per cell (s.e.m.; n = 81 cells), Student's t-test P = 1.30 x 10^-5. · E. coli cells express 40% more GyrB than GyrA; Figure legends

  2. ShownGyrase clusters are non-stoichiometric and enriched in GyrB, with a B:A ratio well above the whole-cell ratio.

    Mean 19.9 ± 1.1 GyrA per cluster (n = 217) vs 50.6 ± 1.8 GyrB (n = 223), P = 1.4 x 10^-38; cluster B:A 2.5:1 vs 1.4:1 for the cell overall. · Introduction; Gyrase clusters are significantly enriched with…

  3. SuggestedHexanediol partly dissolves clusters and slows turnover, pointing to a weakly bound, liquid-like subpopulation alongside a DNA-bound solid-like fraction.

    HEX lowered stoichiometry to 13.5 ± 0.7 GyrA and 41.7 ± 1.7 GyrB (vs 19.9 and 50.6 untreated). FRAP recovery fell from 42 ± 3% to 25 ± 3% (GyrA) and from 57 ± 3% to 42 ± 5% (GyrB); the effect is only partial. · Gyrase clusters are significantly enriched with…; Gyrase clusters are liquid condensates

  4. ShownGyrA clusters diffuse faster than GyrB clusters. Gyrase-targeting antibiotics lower GyrA cluster diffusivity and raise GyrA copy number, with no significant change in GyrB copy number.

    Mean D 0.50 ± 0.05 vs 0.27 ± 0.04 µm^2/s (P = 4.0 x 10^-4). Coumermycin A1: GyrA D 0.29 ± 0.03 µm^2/s, GyrA 901 ± 39 per cell, GyrB 674 ± 32 (P = 0.66). · GyrA clusters have a significantly higher…; The mobility of gyrase clusters is lower in the…; ATP-binding inhibition increases GyrA and GyrB…

  5. SuggestedDocking assembles a cluster of two A2B2 heterotetramers plus four GyrB monomers, with excess GyrB forming weak multivalent bridges. Adding both GyrA and GyrB subunits gave filamentous structures not seen experimentally.

    HDOCK confidence 0.9998 for GyrA:GyrB, 0.996 for heterotetramer, 0.859 for two GyrB vs 0.998 for two GyrA; 0.7 or higher means very high binding probability. · Excess GyrB in clusters forms multiple weak…; Structural docking

Methods, data and reproducibility

Dual-colour Slimfield single-molecule imaging of genomic GyrA-mCherry and mGFP-GyrB in live E. coli, with stepwise photobleaching counting, cluster tracking, FRAP and hexanediol disruption, plus HDOCK structural docking of gyrase subunits.

  • Dual-labelled GyrA-mCherry/mGFP-GyrB E. coli (TB28 derivative) in M9 glycerol; copy numbers per cell · Escherichia coli · n = 81 cells · 10.5281/zenodo.20559106
  • Untreated cluster tracking and stoichiometry (GyrA and GyrB clusters) · Escherichia coli · n = 217 GyrA clusters; 223 GyrB clusters
  • Antibiotic-treated clusters: coumermycin A1, ciprofloxacin, rifampicin (at MIC) · Escherichia coli · n = Coumermycin A1 366/254; ciprofloxacin 400/375; rifampicin 140/137 (GyrA/GyrB)
  • FRAP traces for GyrA and GyrB, untreated and hexanediol-treated · Escherichia coli · n = 20-24 FRAP traces
  • In vitro EMSA of gyrase at 1:1 vs 1:5 GyrA:GyrB with Cy5-labelled 167 bp DNA
  • Structural docking of gyrase subunits and DNA (339 bp minicircle simulation average) · PDB 9GBV; DOI 10.15124/35e4e340-5aee-4fe0-b49b-eae2722a5914
Methods & tools
Slimfield microscopy, Stepwise photobleaching molecule counting, FRAP, MATLAB, Python, ImageJ, SciPy curve_fit, HDOCK server, UCSF ChimeraX, PDB2PQR, Welch's t-test, Fisher F-test (gamma-mixture diffusion fits)
Software & environment
MATLAB, LabVIEW, ImageJ and Python; SciPy curve fitting for FRAP; HDOCK server and UCSF ChimeraX for docking. No software versions stated.
Reproducibility gaps
Plasmids available on request only, subject to materials transfer agreements; code under CC-BY-NC-SA 4.0 licence; data given as DOIs only; no software versions or compute stated.
Limitations (4)
  • Whether clustering raises effective gyrase catalytic rate by the roughly 3 orders of magnitude predicted was not measured and is left to future work. · stated by the authors
  • The mechanism by which coumermycin A1 stalls the gyrase cycle, given an inhibition mode that differs from ciprofloxacin, is described by the authors as an open question. · stated by the authors
  • In vitro EMSA with 1:5 GyrA:GyrB showed no binding change (within ~5%), so no direct biochemical evidence yet shows that excess GyrB enhances function; the processivity link is inferred. · noted in this summary
  • The GyrB bridging model rests on computational docking with no mutational or direct interaction tests, and the hexanediol effects on stoichiometry and FRAP recovery were only partial. · noted in this summary

Read this if you study bacterial topoisomerases, biomolecular condensates, or in vivo single-molecule stoichiometry counting with Slimfield and FRAP.

Citations
0

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISStracy '18Syeda '26Germe '22Soczek '18Pozdeev '20Broeck '19
72% semantic matchBiophysics
M. Stracy, A. Wollman, E. Kaja+9 more·bioRxiv·Nov 1, 2018