BGPT: Paper Review: Expanded Dataset Reveals the Emergence and Evolution of DNA Gyrase in Archaea

Fuel Your Discoveries

Quick Explanation Copied

What this paper claims (and tests): using a large curated archaeal+bacterial dataset of gyrA/gyrB, the authors infer a late acquisition of DNA gyrase into Archaea—most parsimoniously via a single HGT into Euryarchaeal group II—followed by secondary rare HGT into some DPANN and Asgard lineages, plus a few more recent archaeal↔bacterial transfers. They further argue that the archaeal type II topoisomerase Topo VI shows lineage-specific structural co-adaptation associated with gyrase-encoding lineages, suggesting functional “division of labor”.

Scientific “red flags” to scrutinize: inferences of ancient HGT depend on (a) phylogenetic signal vs rate variation/branch attraction, (b) assumptions about archaeal-tree rooting (which the paper notes remains debated), and (c) reliance on gene presence/sequence similarity as a proxy for in vivo functional integration. The authors explicitly discuss some of these uncertainties (e.g., weak support at some deep nodes, sensitivity to alignment trimming, uncertain DPANN placement/rooting).

Long Explanation

Paper Review (Visual-first): DNA gyrase emergence & evolution in Archaea

Citation: Villain et al., Molecular Biology and Evolution, 10.1093/molbev/msac155 (2022)

Curated dataset + phylogenomics + comparative conservation + synteny + AlphaFold2 modeling + Topo VI structural adaptation claims.

1) What the authors measured (raw-data footprint)

Gyrase subunit dataset sizes: archaeal GyrA & GyrB counts (377 and 331 in the final trees) and a concatenated GyrA+GyrB dataset used for a global tree (799 total; 502 bacterial + 297 archaeal).
Pipeline choices included: intein filtering, removal of sequences with certain contig/annotation issues, trimming using BMGE and a “less strict” Noisy approach, ML trees in IQ-TREE with UFBoot + SH-aLRT, and topology tests.

2) Main evolutionary model (as inferred by the paper)

The authors’ best-supported, most-parsimonious scenario is:

DNA gyrase was introduced into the lineage leading to Euryarchaeal group II via a single bacterial→archaeal HGT event, with donor placement suggested as an ancestor of Gracilicutes and/or Terrabacteria.
Secondary rare HGT disperses gyrase from Euryarchaeal group II into some DPANN and Asgard lineages.
A small number of more recent archaeal↔bacterial transfers can explain gyrase presence in some Methanobrevibacter.

Core phylogenetic reasoning: the global gyrase tree shows tripartite bacterial clades (Terrabacteria and Gracilicutes as two major bacterial groups) while Archaea form a monophyletic clade, and the inter-domain split is supported by short branches, which the authors interpret as evidence against LUCA-origin for gyrase.

3) Visualizing the “transfer gatekeeping”: why gyrase is patchy in Archaea

The paper’s distribution claims are central: systematic presence in Euryarchaeal group II, near-absence in group I Euryarchaeota (with sporadic exceptions), and partial/rare presence in DPANN and Asgard. The authors also emphasize that absence calls may be confounded by incomplete genomes/uncultivated taxa, and that some placements (notably DPANN) can be affected by branch-attraction/fast evolution.

Note: This visualization encodes the qualitative presence/absence language explicitly stated in the paper (not exact genome counts).

4) Co-evolution hypothesis: gyrase vs Topoisomerase VI (Topo VI)

The paper frames Topo VI as ubiquitous ancestral type II topoisomerase in Archaea, then asks why gyrase (which overlaps functionally) isn’t just redundant. Their in silico test compares Top6A/Top6B phylogenies and sequences and reports:

Topo VI sequences in gyrase-encoding vs gyrase-less archaea do not form a single shared clade, suggesting independent lineage-specific adaptations rather than one ancient co-transfer event.
They report a gyrase-associated C-terminal extension pattern in Top6B (Topo VI subunit B), including an immunoglobulin-like fold predicted by AlphaFold2 for gyrase-encoding lineages, and different shorter helical architecture in gyrase-less archaea.

This is a schematic derived from the paper’s descriptive CTD statements; it is not a measurement from the figures, because exact residue-length distributions are not enumerated in the text provided.

5) Rigorous critique: where the inference is strong vs uncertain

Strengths (evidence quality within the paper’s scope)

Large, curated comparative dataset with explicit quality filters (inteins, contig/rRNA congruence, redundancy reduction) and a deposited figshare dataset, supporting reproducibility of downstream phylogenetic analyses.
Robustness testing for phylogenetic signal: they rerun trees with different trimming rigor (BMGE vs Noisy) and report changes in internal resolution and support.
Use of topology tests (IQ-TREE RELL resampling) to evaluate whether forcing archaeal monophyly is significantly worse than maximum likelihood.

Uncertainties / potential failure modes (what could disprove key claims)

Archaeal rooting remains debated, which the paper explicitly flags as limiting interpretation of the timing of domain-level acquisitions.
Fast-evolving lineages (DPANN) can create artefacts (branch attraction). The paper acknowledges cases where DPANN sequences branch within other archaeal clades and offers both recent HGT and artefactual attraction as alternatives.
Inferred “HGT success conditions” are still hypothetical: the paper does a mechanistic plausibility argument (mobile elements rarely encode gyrase; physical proximity may be required via donor-recipient ecology/symbiosis) but does not experimentally test those prerequisites across candidate host backgrounds.
Functional integration in DPANN/Asgard is not directly assayed in the paper’s analysis: sequence motifs and modeled/phylogenetic placement support plausible negative-supercoiling activity, but this remains inferential for many taxa.

6) Data availability & reproducibility checkpoints

The authors state that all datasets used are deposited on figshare under DOI: 10.6084/m9.figshare.19137899.

Score is a qualitative indicator of stated data availability only (not a guarantee of complete method reproducibility).

7) Suggested next falsification experiments (in silico → experimental bridge)

Below are targeted experiments that would most directly challenge the paper’s central inferences.

Functional transfer test across recipient backgrounds: pick archaeal recipients that differ in gyrase presence/Topo VI CTD architecture (gyrase-encoding vs gyrase-less), heterologously express inferred gyrase orthologs and test whether negative supercoiling introduction occurs without destabilizing growth—specifically measuring whether Topo VI CTD architecture predicts compatibility.
Phylogenetic robustness falsification: re-run phylogenies using alternative models, exclusion masks for known problematic sites, and alternative rooting constraints while tracking how many inferred HGT events persist. The goal is to see if the “single transfer into group II” conclusion is stable under more stringent phylogenetic assumptions.

Author reviews (jump deeper)

Feedback:

Updated: March 25, 2026