Regulation of DNA Topology in Archaea: State of the Art and Perspectives

ABSTRACT

DNA topology is a direct consequence of the double helical nature of DNA and is defined by how the two complementary DNA strands are intertwined. Virtually every reaction involving DNA is influenced by DNA topology or has topological effects. It is therefore of fundamental importance to understand how this phenomenon is controlled in living cells. DNA topoisomerases are the key actors dedicated to the regulation of DNA topology in cells from all domains of life. While significant progress has been made in the last two decades in understanding how these enzymes operate in vivo in Bacteria and Eukaryotes, studies in Archaea have been lagging behind. This review article aims to summarize what is currently known about DNA topology regulation by DNA topoisomerases in main archaeal model organisms. These model archaea exhibit markedly different lifestyles, genome organization and topoisomerase content, thus highlighting the diversity and the complexity of DNA topology regulation mechanisms and their evolution in this domain of life. The recent development of functional genomic assays supported by next‐generation sequencing now allows to delve deeper into this timely and exciting, yet still understudied topic.

DNA topoisomerases are ubiquitous enzymes that play a crucial role in regulating DNA supercoiling which affects fundamental biological processes involving DNA. In this review, we summarize the current understanding of regulation of supercoiling by DNA topoisomerases in the third domain of life, the Archaea, with a particular focus on three key model organisms: Thermococcus kodakarensis, Sulfolobus sp., and Haloferax volcanii.

1. Introduction

Chromosomes are made of a single molecule of DNA that carries the genetic information from the mother cell to the daughter cell. In prokaryotes, Bacteria and Archaea, the chromosome is typically a circular double‐stranded DNA molecule with a size varying from 0.5 to 10 Mbp. The genomic DNA of prokaryotes is folded in a hierarchical and organized manner and is highly compacted into a body called the nucleoid (Verma, Qian, and Adhya ²⁰¹⁹; Lioy, Junier, and Boccard ²⁰²¹). Nucleoid‐associated architectural proteins (NAPs) and DNA supercoiling are important contributors to the condensation and spatial organization of DNA at the smallest scale (1 kbp or less and up to 10 kbp) (Joyeux 2015; Schwab and Dame ²⁰²⁴). We will focus in this review on DNA supercoiling and its regulation by topoisomerases in Archaea. Readers interested in learning more about archaeal NAPs are referred to excellent reviews discussing their characterization and classification (Peeters et al. 2015; Schwab and Dame ²⁰²⁴).

DNA supercoiling can be plectonemic, whereby the DNA helix is wound around itself in a higher‐order helix. The sense of the interwinding is right‐handed (underwound) for negatively supercoiled DNA and left‐handed (overwound) for positively supercoiled DNA (Figure 1). There is also an alternative way to form DNA supercoils by toroidal winding, whereby the DNA helix is coiled around the surface of a protein or a protein complex such as a nucleosome (Bates and Maxwell 2005). In actively replicating cells, plectonemic DNA supercoiling is generated by DNA‐templated processes such as transcription, DNA replication, and repair, and also by topoisomerases DNA gyrase and reverse gyrase (see below for a description of these enzymes). In the case of transcription, according to the twin domain model (Liu and Wang 1987; Wu et al. ¹⁹⁸⁸), the elongating RNA polymerase generates positively supercoiled DNA ahead of the transcribing complex and negatively supercoiled DNA behind. These supercoils must be efficiently removed to allow gene expression to occur. The translocating DNA replication machinery generates positive supercoils ahead, and those can diffuse behind the replication fork by the rotation of the fork, leading to the interwinding of the two daughter duplexes (Postow et al. 2001). These interwound structures have been named precatenanes (as they contain single stranded‐regions on the lagging strand) and they become proper catenanes or hemicatenanes (in which parental strands are still linked to each other and two newly synthesized daughter strands are unlinked to each other) if they are not removed before replication termination (Laurie et al. 1998; Postow et al. ²⁰⁰¹). This leaves the two circular chromosomes (the template and the newly synthesized chromosome) interlinked, a topology that must be resolved to allow the distribution of the newly replicated genome to daughter cells.

FIGURE 1.

Topological properties of closed circular DNA and formation of DNA supercoils. The topology of a closed circular DNA molecule can be explained by the simple formula Lk = Tw + Wr, where Lk (linking number) is equal to half the number of times the two complementary DNA strands cross each other, or, in other words, the number of double helical turns in a plane and a relaxed closed circular DNA molecule such as the one depicted in the center of the figure. In such a molecule Lk is equal to Tw (twist), which describes how the individual strands of DNA coil around one another. Tw can be defined as N/h where N is the size of the molecule in bp and h is the helical repeat, which under standard conditions (0.2 M NaCl, pH 7, 37°C) is taken to be 10.5 bp/turn (Bates and Maxwell ). According to this formula, we can calculate that the molecule in our example is composed of 200 bp. Finally, Wr (writhe) corresponds to the number of times the double helix crosses over itself, or to the number of superhelical turns or supercoils. In our example, we see that the central DNA molecule has a relaxed topology with Lk = Tw = 19. The Lk of a DNA molecule can only be changed by the action of DNA topoisomerases, which do so by introducing a break in one or both DNA strands, rotating the two strands relative to each other, and sealing the break. This can lead to an increase (overtwisting) or decrease (denaturation) in Tw. Since Lk is constant for a closed circular DNA, changes in Tw must be accompanied by an equal and opposite change in Wr. We can see in our example on the right that the relaxed molecule, which underwent an increase of Tw from 19 to 23 (the Lk now being equal to 23), can be converted to a positively supercoiled molecule by the introduction of four positive supercoils and the loss of four Tw (four helical turns) such that the sum Tw + Wr remains 23. Supercoils can be of either handedness: positive supercoils form left‐handed crossovers in the double helix, while negative supercoils form right‐handed crossovers.

As we have seen, it is critical to timely and finely regulate DNA topology such that essential cellular processes can occur. DNA topoisomerases are enzymes with dedicated roles in dynamically regulating the topological constraints by introducing transient DNA breaks (Vos et al. 2011; Pommier et al. ²⁰¹⁶; Duprey and Groisman ²⁰²¹). Since the discovery of the first DNA topoisomerase by James C. Wang in 1971 (Wang 1971) (the omega protein, ω, from Escherichia coli, now called Topoisomerase I), much of the research has focused on the biochemical and structural properties of these essential enzymes. This has led to their classification into type I and type II enzymes, according to whether they cleave ssDNA or dsDNA, respectively. Each type is further divided into families that are further classified into subfamilies. Comprehensive reviews on this topic are available, and readers are referred to them for a detailed description of known DNA topoisomerases (Wang 1996; Schoeffler and Berger ²⁰⁰⁸; Chen, Chan, and Hsieh ²⁰¹³; Bush, Evans‐Roberts, and Maxwell ²⁰¹⁵; McKie, Neuman, and Maxwell ²⁰²¹; Sutormin et al. ²⁰²¹; Tan and Tse‐Dinh ²⁰²⁴). For the purposes of the present review, we have summarized the principal characteristics of each family in the Table 1.

TABLE 1.

Main characteristics of known DNA topoisomerases.

Classification			Characteristics				Activities					Distribution
Type	Family	Sub‐family	Enzyme structure	Bond	ATP	Mg2+	Relaxation		Supercoiling		Catenation/Decatenation	Archaea	Bacteria	Eukaryotes
							−SC	+SC	−SC	+SC
I	IA	I	Monomer	5′	No	Yes	+++	−	−	−	+	Yes	Yes	No
		III	Monomer	5′	No	Yes	+	−	−	−	++	Yes	Yes	Yes
		Reverse gyrase	Monomer or hetero‐dimer	5′	Yes	Yes	+++	−	−	+++	ND	Yes	Yes	No
	IB	IB	Monomer	3′	No	No	+	+	−	−	−	Yes	Yes	Yes
	IC	V	Monomer	3′	No	No	+	+	−	−	−	Yes	No	No
II	IIA	II	Homo‐dimer	5′	Yes	Yes	+	+	−	−	+	No	No	Yes
		IV	Hetero‐tetramer	5′	Yes	Yes	+	+	−	−	+	No	Yes	No
		Gyrase	Hetero‐tetramer	5′	Yes	Yes	−	+++	+++	−	+	Yes	Yes	No
		NM$	Hetero‐tetramer	ND	Yes	Yes	+++	+	−	+++	+	ND	Yes	ND
	IIB	VI	Hetero‐tetramer	5′	Yes	Yes	+	++	−	−	++	Yes	No	Yes#
		VIII*	Homo‐dimer	ND	Yes/No	Yes	+	+	−	−	+	Yes	Yes	No

Note: (−SC) DNA negatively supercoiled; (+SC) DNA positively supercoiled; (Bond) covalent bond to the 5′ or 3′ extremity of DNA; (ND) non‐determined; ($) TopoNM is a unique type II topoisomerase found so far only in Mycobacterium smegmatis (Jain and Nagaraja 2005); (#) The homologs of the Topo VI A subunit named Spo11 are widely spread in Eukaryotes; some lineages (including plants) additionally encode the archetypal Topo VI composed of the Topo VI A and Topo VI B proteins (Brinkmeier et al. 2022); (*) homologs of the A‐subunit of Topo VIII, named Mini‐A, were recently identified in extrachromosomal and integrated bacterial and archaeal viruses (Takahashi et al. 2020); Table adapted from Bush, Evans‐Roberts, and Maxwell (2015).

While the extensive biochemical characterization of topoisomerases has been crucial for the development of clinically successful drugs targeting these enzymes to fight against bacterial infections (e.g., fluoroquinolones) or cancers (e.g., camptothecins) (Pommier et al. 2010), our knowledge about the cellular roles of topoisomerases and their regulation by external factors continues to expand. Many key fundamental questions, however, remain. This review aims to provide an overview of what is currently known on this topic in Archaea and present a vision of the future challenges and perspectives for this timely and exciting, yet still understudied, avenue of research.

2. Archaea Encode a Unique Blend of Topoisomerases

Archaea were recognized as a separate domain of life, phylogenetically distinct from the Bacteria and the Eukaryotes in 1977 by the pioneering work of Carl Woese (Woese and Fox 1977), and since those early studies, the knowledge about their diversity, evolution and ecology has been greatly expanded, especially in the last decade where such studies benefited from the wealth of metagenomic data (Baker et al. 2020). Archaea exhibit broad phylogenetic diversity (Figure 2), comparable to that of the Bacteria, are ubiquitously distributed in the environment, including the human gut, and play important roles in global geochemical cycling and climate change through production of methane, a potent greenhouse gas (Hug et al. 2016; Adam et al. ²⁰¹⁷; Baker et al. ²⁰²⁰; van Wolferen et al. ²⁰²²). Beyond their role in the ecology of our planet, Archaea have been recently identified as the direct ancestors of eukaryotes, highlighting these organisms as relevant models for understanding the evolutionary pathway leading to the emergence of complex eukaryotic cells (Spang et al. 2015; Zaremba‐Niedzwiedzka et al. ²⁰¹⁷; Eme et al. ²⁰¹⁷). Remarkably, the Archaea exhibit bacterial‐like morphology and small tightly packed circular genomes but their information processing systems (e. g., DNA replication, transcription and translation) are homologous to eukaryotic counterparts (Cavicchioli 2011). This characteristic combination of features is followed by an equally unique combination of DNA topoisomerases found within the Archaea (Forterre and Gadelle 2009; Garnier et al. ²⁰²¹). As a general rule, every cell encodes at least one type I and one type II topoisomerase, and the Archaea are not an exception. The topoisomerase content in representative species of the main archaeal phyla and the biochemical properties of those topoisomerases were recently reviewed (Garnier et al. 2021). It is worth mentioning that topoisomerases encoded by plasmids and integrated viruses (Topo VIII and Mini‐A, Table 1) can also be found in the Archaea (Takahashi et al. 2020). How these enzymes promote plasmid or virus replication and whether they help viruses to subvert the chromosome organization of the host is still an unexplored topic. For the purposes of this review, we will briefly describe the characteristics of topoisomerases relevant for the understanding of the following sections.

FIGURE 2.

Phylogenetic tree of the Archaea. The main lineages within the four supergroups are shown according to the GTDB classification (GTDB v214.1 released 09.06.2023). The lineages and the model species described in more detail in this review are highlighted in red.

Topoisomerase III (Topo III) is the main type I enzyme in the Archaea. It is composed of a single polypeptide and is highly conserved across prokaryotes and eukaryotes (Bizard and Hickson 2020) (Table 1). In bacteria, Topo III has been shown to efficiently resolve precatenanes in vitro and is therefore thought to be mainly functioning in the decatenation pathways in vivo (Nurse et al. 2003). Accordingly, in E. coli, Topo III was reported to be associated with the replication fork in vivo and its deletion leads to severely deficient chromosome segregation (Lee et al. 2019). E. coli Topo III has also been shown to resolve converging replication forks in association with RecQ helicase, and this functional cooperation is mediated by interaction with the SSB proteins (Suski and Marians 2008). In eukaryotes, yeast Topo III was shown to function in the maintenance of genomic integrity and DNA repair by associating with RecQ‐like helicase in a DNA‐repair complex (Gangloff et al. 1994). Intriguingly, about 30 years ago, Topo III from E. coli was reported to cleave RNA and interconvert different topological states of RNA molecules by RNA strand passage, suggesting this enzyme may also be an RNA topoisomerase (DiGate and Marians ¹⁹⁹²; Wang, Di Gate, and Seeman ¹⁹⁹⁶). The same has more recently been suggested for one of the two isoforms of Topo III (isoform beta) found in metazoans and fungi and which has been found to interact with RNA binding proteins and co‐localize with polyribosomes bound mRNA (Stoll et al. 2013; Xu et al. ²⁰¹³; Ahmad et al. ²⁰¹⁷). The best‐studied archaeal Topo III is the one from Saccharolobus (formerly Sulfolobus) solfataricus (hereafter called SsTop3). In vitro SsTop3 was shown to relax negatively supercoiled plasmid DNA whereby the religation step was temperature‐dependent and could only be achieved at close to the native growth temperatures of S. solfataricus (Dai et al. 2003; Chen and Huang ²⁰⁰⁶). SsTop3 has also been found to be very efficient at unlinking DNA catenanes (Bizard et al. 2018). A close homolog from Nanoarchaeum equitans (DPANN clade) was additionally shown to generate or solve hemicatenanes when used at high or low concentrations, respectively (Lee et al. 2013). SsTop3 enzyme can use the breathing of negatively supercoiled DNA to provide a transient single stranded region required for the relaxation reaction while the decatenation activity was strongly stimulated by stabilization of the ssDNA regions by single‐strand binding proteins (Bizard et al. 2018). Interestingly, SsTop3 was found to form a stable complex with the Hel112 helicase a member of the RecQ family helicases. When assayed in combination with Hel112, SsTop3 inhibited the Hel112 activity on Holliday junctions and stimulated formation and stabilization of such structures (Valenti et al. 2012). It is speculated that in vivo Hel112 unwinds DNA to expose ssDNA regions, thus facilitating the binding of SsTop3 and the cleavage reaction. Depending on the substrate, these combined activities may result in DNA catenation/decatenation, dissolution of double Holliday junctions, resolution of replication and recombination intermediates, restart of stalled replication forks, or DNA damage repair.

The gene encoding the Topo III homolog from a closely related Sulfolobus islandicus archaeon could be deleted, showing that, at least in this species, this topoisomerase is nonessential (Li et al. 2011). Still, the cells exhibited a slow growth phenotype whereby a small fraction of cells carried an aberrant number of genome equivalents, suggesting that chromosomal segregation was affected and potentially pointing to the role of this topoisomerase in chromosome decatenation (Li et al. 2011). In distantly related species, Methanococcus maripaludis, the Topo III‐encoding gene was called essential following the analysis of genome‐wide transposon mutagenesis data (Sarmiento, Mrázek, and Whitman ²⁰¹³).

Reverse gyrase (RG) is a particularly intriguing type I topoisomerase discovered forty years ago in a thermophilic archaeon Sulfolobus sp. (Kikuchi and Asai 1984). With few exceptions, RG is a single polypeptide composed of an N‐terminal RecQ‐like helicase domain and a C‐terminal bona fide topoisomerase domain (type IA) that cooperate to change DNA topology in an ATP dependent fashion (Confalonieri et al. 1993; Rodríguez and Stock ²⁰⁰²; Yang et al. ²⁰²⁰; Collin, Weisslocker‐Schaetzel, and Klostermeier ²⁰²⁰) (Table 1). RG is the only topoisomerase capable of generating positive supercoils in DNA (McKie, Neuman, and Maxwell ²⁰²¹). Remarkably, RG is found in many thermophiles and all hyperthermophiles (the organisms with optimal growth temperatures above 80°C), but never in mesophiles indicating a pivotal role for this enzyme in thermoadaptation (Forterre 2002). It was early on suggested that RG ensures genome stability at high temperatures through the introduction of positive supercoils that make DNA less prone to denaturation by increasing the number of topological links (Kikuchi and Asai 1984; Forterre ²⁰⁰²). Indeed, the evidence for the existence of positively supercoiled DNA in vivo was reported in 1986 in Sulfolobus sp. (Nadal et al. 1986), strengthening the idea that RG has positive supercoiling activity in vivo. However, several other studies implicated, albeit indirectly, the RG in DNA repair but the underlying mechanism and whether it requires the topoisomerase activity remained obscure. RG was reported to be recruited to UV‐damaged DNA (Napoli et al. 2004) in vivo and involved in sensing and eliminating unpaired ssDNA regions (Hsieh and Plank 2006) and nicks in DNA independently of its topoisomerase activity in vitro (Kampmann and Stock 2004). In Saccharolobus solfataricus, co‐immunoprecipitation experiments from cell extracts and in vitro interaction studies with purified proteins revealed a stable interaction between the RG, translesion DNA polymerase and the SSB protein, suggesting that RG may associate in DNA‐repair related complexes (Valenti et al. 2009). Other in vitro work on the thermophilic archaeal transcription system from Sulfolobus shibatae showed that RNA polymerase binding and promotor opening was blocked on positively supercoiled DNA templates at lower non‐physiological temperatures, suggesting that RG topoisomerase activity may be important for the regulation of gene expression in (hyper)thermophiles (Bell et al. 1998). In line with this hypothesis, the transcription from a plasmid encoded promotor using an in vitro reconstituted system from a distantly related hyperthermophilic archaeon Pyrococcus furiosus was inhibited when the assay was conducted in the presence of RG, indicating that positive supercoiling of DNA inhibits transcription. Collectively, these results suggest a link between the RG, DNA damage repair, and gene expression regulation but how exactly the RG functions in these processes is not yet understood.

The Sulfolobales archaea are unique in that they encode two reverse gyrases, TopR1 and TopR2 (Garnier et al. 2021). Two independent studies (C. Zhang et al. ²⁰¹³; Zhang et al. ²⁰¹⁸) concluded that TopR1 is essential, while in one of those studies the TopR2 encoding gene could be deleted (Zhang et al. 2018), but the resulting strain required two weeks to grow colonies on plates, suggesting that TopR2 is also essential in the natural environment of Sulfolobus. The two enzymes differ markedly in their in vitro activities: TopR2 is highly processive and is able to introduce a very high number of positive supercoils in DNA, while TopR1 is distributive and generates weakly positively supercoiled DNA. The two enzymes do not have the same requirement for ATP, with TopR1 exhibiting ATP‐dependent relaxation activity while TopR2 was active in the absence of ATP. On top of that, the two reverse gyrases behave differently at different temperatures (Bizard, Garnier, and Nadal ²⁰¹¹). For instance, TopR2 is active at a “cold” temperature (45°C) while TopR1 is not. Instead, TopR1 is more active at supraoptimal temperatures above 88°C, while TopR2 is less active at this temperature (Couturier et al. 2014). On the basis of these in vitro data and the quantification of TopR1 and TopR2 in Sulfolobus cells, it was proposed that TopR1 is involved in the homeostatic control of supercoiling (see the chapter devoted to Sulfolobus below), while the in vivo function of TopR2 remains mysterious.

All archaea, with the exception of Thermoplasmatales, encode Topoisomerase VI (Topo VI), a type IIB enzyme that in vitro has robust DNA decatenase activity and relaxase activity of positively supercoiled DNA, both in an ATP‐dependent fashion (Table 1). Topo VI enzymes are active as heterotetramers, composed of two Topo6A and two Topo6B subunits, whereby the former carries the catalytically active tyrosine necessary for inducing the double stranded DNA break (DSB), and the latter contains an ATP binding domain (Bergerat et al. 1997; Corbett, Benedetti, and Berger ²⁰⁰⁷). Using a single‐molecule approach, a recent study demonstrated that Methanosarcina mazei Topo VI preferentially senses DNA crossings with geometries close to 90° indicating that this enzyme is a preferential decatenase with a role in supporting chromosome disentanglement during DNA replication (McKie et al. ²⁰²²). Interestingly, Topo6A homologs named Spo11 are also widespread in eukaryotes (Brinkmeier et al. 2022), including fungi, plants and animals, where they catalyze the induction of DSB to initiate meiotic homologous recombination, a key step in sexual reproduction (Robert et al. 2016; Vrielynck et al. ²⁰¹⁶; Arter and Keeney ²⁰²³). Plants and several other eukaryotic lineages additionally encode Topo6A and Topo6B homologs (Brinkmeier et al. 2022) that in plants were shown to assemble into Topo VI‐like enzymes that are essential for endoreduplication, a common process in eukaryotes that determines cell size and involves DNA amplification without corresponding cell divisions (Sugimoto‐Shirasu et al. ²⁰⁰²). Remarkably, the yeast Spo11 core complex (which in addition to Spo11 includes three other proteins Rec102, Rec104 and Ski8) shows a high degree of structural similarity to the archaeal Topo6AB heterodimer (Claeys Bouuaert et al. 2021), thus highlighting the relevance of the studies of the archaeal Topo VI in particular with regard to the studies of meiosis and the emergence and evolution of this complex process from an archaeal ancestor.

A second type II enzyme found in Archaea is the DNA gyrase (gyrase). The gyrase is the only topoisomerase that is capable of introducing negative supercoils into DNA at the expense of ATP (Gellert et al. 1976; Cozzarelli ¹⁹⁸⁰). It is a type IIA enzyme composed of two GyrA and two GyrB subunits assembling into an A₂B₂ heterotetramer (Table 1). This topoisomerase is ubiquitous in Bacteria and is also found in a few archaeal lineages belonging to the Euryarchaeota, the DPANN and the Asgard phyla (Villain et al. 2022). Recent phylogenetic analyses indicate that this enzyme was introduced into an Euryarchaeal ancestor via a single horizontal gene transfer (HGT) event from a bacterial donor (Raymann et al. 2014; Villain et al. ²⁰²²). Even though bacterial gyrase was shown to have versatile in vitro activities, its main in vivo activity is the relaxation of positive supercoils introduced by replication and transcription (Kreuzer and Cozzarelli 1979; Ahmed et al. ²⁰¹⁷; Sutormin et al. ²⁰¹⁸; Stracy et al. ²⁰¹⁹). Early work revealed that the synthesis of gyrase is itself controlled by DNA supercoiling thus demonstrating a homeostatic control mechanism of supercoiling in E. coli (Menzel and Gellert 1983). Through its ATP‐dependent negative supercoiling activity, gyrase is thought to play a pivotal role in transducing environmental signals to the bacterial chromosome and coordinating its transcriptional response, allowing bacteria to rapidly adapt to changing environmental conditions (Dorman and Dorman 2016; Martis et al. ²⁰¹⁹). The underlying mechanism involves the modulation of gyrase activity by the available energy content of the cell (the ATP/ADP ratio), which in turn affects global DNA supercoiling and is associated with complex transcriptional responses affecting a large fraction of the genome in a variety of distantly related bacterial species (Martis et al. 2019).

In summary, we now have a fairly good understanding of the biochemical and structural properties of archaeal topoisomerases, showing that these enzymes can, to some extent, be active over a wide range of substrates and even have overlapping activities (e.g., both Topo III and Topo VI exhibit robust decatenase activities in vitro). What the preferential activities of these enzymes are in vivo and how they cooperate in cells to regulate DNA topology is still far from being understood. As a primer for delving into what is currently known about this topic, we will describe in the following section what is known about the average levels of unconstrained DNA supercoiling in Archaea based on studies of plasmid DNA topology, as it is likely that this level is mainly controlled by the relative activities of competing topoisomerases.

3. Archaea Harbor the Widest Range of Mean Supercoiling Levels

Unfortunately, no method is currently available to measure finely and directly the absolute value and sense of plectonemic supercoiling across entire chromosomes. It is possible, though, to determine the average chromosomal supercoiling level and handedness using ethidium bromide gradient centrifugation of purified chromosomes (Crawford and Waring 1967; Worcel and Burgi ¹⁹⁷²; Pruss, Manes, and Drlica ¹⁹⁸²). Methods relying on the λ phage site‐specific recombination system or psoralen photobinding have also been used since the 1980s to measure the distribution of supercoiling on the chromosome of bacteria and eukaryotes (Sinden 1980; Bliska and Cozzarelli ¹⁹⁸⁷). These early studies showed that E. coli DNA is packed in chromosomes with net negative torsional tension, whereby it is assumed, based on the topology of plasmids isolated from E. coli, that about 40% of negative supercoiling is unconstrained, that is, able to adopt a plectonemic conformation seen in vitro, while the remaining 60% is constrained and consists of changes in writhe and/or twist stabilized by binding of different kinds of proteins (Pettijohn and Pfenninger 1980; Bliska and Cozzarelli ¹⁹⁸⁷). In marked contrast to bacteria, in eukaryotes, wrapping of initially relaxed DNA around nucleosomes (toroidal winding of DNA) induces compensatory positive supercoiling that is relaxed by topoisomerases, leading to negative supercoiled DNA wholly constrained within nucleosomes (Sinden 1980; Richmond and Davey ²⁰⁰³). Although very informative, site‐specific recombination systems or psoralen photobinding methods are quite laborious to set up, and therefore the analysis of native plasmid DNA topology is frequently used as a proxy for the global steady‐state level of chromosomal DNA supercoiling.

Plasmid DNA topology can be fairly easily studied in the lab by using two‐dimensional gel electrophoresis, a method that allows for the separation of plasmid topoisomers based on their supercoiling level and sense (positive or negative) (Lee, Mizusawa, and Kakefuda ¹⁹⁸¹; Gibson, Oviatt, and Osheroff ²⁰²⁰). From these data, one can calculate the specific linking difference, or the supercoiling density, (σ) that is used as a standardized measure of ΔLk where the linking difference is scaled to the size of the plasmid, thus allowing for comparisons between plasmids of different sizes (Bates and Maxwell 2005). It is important to note, though, that factors such as temperature, pH, salt concentration, and pressure significantly impact the helical repeat, h (number of base pairs per turn of the DNA helix) of DNA and this should be considered when interpreting two dimensional gels. The two well‐studied conditions that have an impact on helical repeat (and hence twist) of DNA are the availability of positively charged ions and temperature. Positively charged ions can neutralize the repulsive forces coming from the negatively charged phosphate backbone of DNA, which tend to unwind DNA (Rybenkov, Vologodskii, and Cozzarelli ¹⁹⁹⁷). Hence, increasing the concentration of positively charged ions weakens those repulsive forces, making DNA more tightly wound, thus decreasing the helical repeat and increasing the twist. That means that a plasmid DNA molecule that has relaxed topology in the cytoplasm of an extreme halophilic organism containing molar concentrations of Na⁺ ions (such as Haloferax volcanii, which we discuss below) will show an increased linking number corresponding to positive supercoiling when migrated on an agarose gel under low Na⁺ conditions. Temperature is another factor that affects the helical repeat of the DNA. With increasing temperature, the motion of molecules increases, and in the case of DNA, this results in the unwinding of the DNA helix. The effect of temperature has been studied over a wide range of temperatures (0 to 83°C), and it shows a linear relationship with the change of the winding angle of DNA per base pair of −0.0105°/°C bp (Duguet 1993). This needs to be considered and corrected for when plasmid DNA isolated from thermophilic or hyperthermophilic organisms (such as Thermococcus kodakarensis and Sulfolobus sp., discussed below) is studied.

There is no centralized database for reported specific linking difference of plasmids, but a recent search across the available literature yielded 26 values for plasmid or circular viral DNA from the three domains of life (Villain et al. 2021). When σ is plotted as a function of the optimal growth temperature (OGT) one can observe that (hyper)thermophilic organisms have, in general, positively supercoiled DNA, that is, exhibit a significant linking excess compared to negatively supercoiled plasmids from mesophilic organisms (Figure 3). Such distribution suggests that positively supercoiled DNA is an adaptation to living at high temperatures; however, this seems not to be an absolute requirement as hyperthermophiles such as the bacterium Thermus thermophilus or the archaeon Archaeoglobus profundus (highlighted with an asterisk, Figure 3) both harbor negatively supercoiled plasmid DNA. One can also immediately see that plasmids isolated from bacteria and eukaryotes are all confined to the left of the plot with, their supercoiling densities ranging from −0.04 down to −0.08. This can be explained by the systematic presence in bacteria of the DNA gyrase with dominant negative supercoiling activity (Cozzarelli 1980) and the wrapping of DNA around nucleosomes in Eukaryotes (Sinden 1980; Richmond and Davey ²⁰⁰³). Archaea uniquely exhibit a much larger span of supercoiling densities, going from strongly negatively supercoiled (σ ~ −0.07) over essentially relaxed DNA (σ ~ ±0.005) up to strongly positively supercoiled DNA (σ ~ +0.04), suggesting the existence in these organisms of a variety of singular mechanisms to achieve supercoiling equilibrium. So far, efforts to understand how such equilibrium is achieved, and more generally, how DNA topoisomerases cooperate to dynamically regulate DNA topology in archaea, have been constrained mostly to non‐physiological in vitro conditions centered upon a particular DNA topoisomerase. In the following sections, we will summarize these data in an organism‐centered way and propose how they can be fitted into an integrative model for the regulation of DNA topology in the three most studied archaeal model organisms.

FIGURE 3.

Supercoiling densities of plasmid DNA among living organisms. Plasmid superhelical densities from various organisms and viruses are plotted against the optimal growth temperature (OGT) of their hosts. The highlighted densities correspond to those of the three model organisms discussed in this review. H. volcanii is an obligate halophile; T. kodakarensis, a hyperthermophile; and S. islandicus and related species, which are thermophiles and acidophiles. The values corresponding to plasmids from the bacterium Thermus thermophilus and the archaeon Archaeoglobus profundus are marked by an asterisk. Figure adapted from Villain et al. (2021).

4. Regulation of DNA Topology in Sulfolobus Model Species

Sulfolobales are thermoacidophiles that thrive in some of the most inhospitable environments on Earth, such as acidic hot springs and volcanic solfataras, thus literally bathing in hot acid (Lewis et al. 2021). Based on their position in the phylogenetic tree of Archaea, the Sulfolobales have been classified within the TACK clade, which forms a sister group to the Asgard archaea (Adam et al. 2017; Baker et al. ²⁰²⁰) (Figure 2). The first Sulfolobus species, Sulfolobus acidocaldarius, was isolated by Thomas Brock in 1972 and shown to grow optimally at 70°C–75°C and pH 2–3 under aerobic conditions (Brock et al. 1972). This organism and the two related species Saccharolobus solfataricus and Saccharolobus islandicus (Sakai and Kurosawa 2018), are the only Crenarchaeota with established genetic tools and have emerged as model systems for the study of chromosome organization, DNA replication, transcription and translation processes (She et al. 2009; Wagner et al. ²⁰¹²; Lewis et al. ²⁰²¹). In terms of topoisomerases, the Sulfolobales are endowed, as almost all archaea, with one copy of Topo III and one copy of Topo VI. However, different from other (hyper)thermophiles, the Sulfolobales encode two copies of reverse gyrase, TopR1 and TopR2, that share about 35% of protein sequence identity (Catchpole and Forterre 2019).

Since the topology of DNA has profound effects on genome function and structure, we will first describe the main points of relevance regarding these features in Sulfolobales and then relate what is known about the dynamics of DNA topology in response to environmental cues and its regulation by topoisomerases (Figure 4).

FIGURE 4.

Genome characteristics and DNA topology regulation in Sulfolobus species. (A) Schematic representation of the genome of Sulfolobus sp. The genome is structured into CIDs that are further organized into a transcriptionally active A compartment and a transcriptionally silent B compartment. The latter is more condensed and maintained by a noncanonical SMC‐family protein, coalescin (CslN), which is more abundant in the B compartment. The main NAPs involved in the local spatial organization of the genome are indicated. For simplicity, the binding mode and the relative abundance of these NAPs are not accounted for in the scheme. Bidirectional genomic DNA replication is initiated from three origins of replication (ori). (B) Proposed model for solving DNA supercoiling generated by a transcribing RNA polymerase. The arrow indicates the direction of transcription. Messenger RNA synthesized by the RNA polymerase is depicted in violet. Based on the in vitro activities of the four DNA topoisomerases encoded by Sulfolobus sp., the negative supercoils accumulating behind the RNA polymerase could be relaxed by all four DNA topoisomerases, while the Topo VI is the only topoisomerase predicted to relax the positive supercoils accumulating ahead of the polymerase. (C) A proposed model for solving topological constraints building up during DNA replication. The arrow indicates the direction of replication. For clarity, only the replicative helicase (MCM) and the DNA polymerase are depicted. The arrows indicate the direction of the replication. The two neosynthesized DNA molecules get entangled behind the rotating replisome, forming precatenanes. As precatenanes contain nicks, they could presumably be solved by the action of Topo III, while Topo VI could relax the positive supercoils accumulating ahead of the replisome. (D) Observed changes in DNA supercoiling of plasmids in response to temperature stress. Upon heat shock, the relaxed plasmid DNA becomes positively supercoiled, presumably through the action of the reverse gyrase (mainly TopR1). In response to cold shock, the plasmid DNA becomes negatively supercoiled due to the putative joint action of NAPs and Topo VI. The corresponding mean supercoiling densities (López‐García and Forterre ¹⁹⁹⁷) are indicated below each plasmid molecule. (E) Topo VI is predicted to function as the main decatenase needed for the separation of the two catenated daughter chromosomes upon the end of DNA replication (right part of the panel). The end of replication can also lead to the formation of hemicatenanes (left part of the panel), where the parental strands of the two daughter chromosomes are still linked. Such structures contain single‐stranded regions and can be solved by the action of a type I enzyme. A good candidate is the Topo III of S. solfataricus, which has been shown to exhibit robust decatenase activity in vitro. Created with BioRender.com.

As all archaea studied so far, the Sulfolobales share their genetic organization with bacteria, with an operonic transcription unit structure that is dense and characterized by short intergenic regions, but they also have specific features. One striking feature is the existence, similar to eukaryotes, of multiple replication origins (ori), genomic sites at which chromosomal DNA replication is initiated. Sulfolobales, typically encode three ori sites, and they have been shown to accommodate a single replication initiation event during the cell cycle (Robinson et al. 2004; Lundgren et al. ²⁰⁰⁴; Duggin, McCallum, and Bell ²⁰⁰⁸). The cell cycle is precisely organized, whereby the single chromosome is replicated during the S‐phase, followed by a post‐replicative period (the G2‐phase) in which the sister chromatids remain entangled, forming hemicatenane structures (Robinson et al. 2007).

Recent studies used chromosome conformation capture (3C) assays showed that the Sulfolobus chromosome is organized into eukaryotic‐like A and B compartments and, within those, local bacterial‐like chromosome interacting domains (CIDs), the boundaries of which are formed by strong transcription (Takemata, Samson, and Bell ²⁰¹⁹; Takemata and Bell ²⁰²⁰). These studies also show that, similar to eukaryotes, the transcriptionally active genes and DNA replication origins are enriched in the A compartment, while genes with low transcription levels are mostly confined to the B compartment, which is enriched in a new class of SMC‐like protein (ClsN), specific to Sulfolobales and named coalescin.

While most archaea harbor histones homologous to the H3 and H4 core histones of eukaryotes, these are completely absent in Crenarchaeota (Peeters et al. 2015; Hocher et al. ^{2019, 2022}). Instead, Sulfolobales rely on several NAPs, that are usually small (between 7 and 10 kDa), basic, and highly abundant, to structure their DNA (Peeters et al. 2015; Hocher et al. ²⁰²²). Among these, the Cren7 is highly conserved in Crenarchaeota and was shown to be a versatile architectural protein, bending and also bridging DNA, thereby forming highly condensed chromatin filaments (Guo et al. 2008; Zhang et al. ²⁰²⁰) (Figure 4A).

The basal transcription machinery in the Sulfolobales, composed of RNA polymerase, the key enzyme of the transcriptional apparatus, and general transcription factors needed for transcription initiation, is homologous to the core components of the eukaryal RNA polymerase II apparatus (Zillig, Stetter, and Janeković ¹⁹⁷⁹; Bell and Jackson ²⁰⁰¹; Blombach et al. ²⁰¹⁹). Interestingly, in vitro reconstitution of the Sulfolobus machinery showed, in contrast to characterized bacterial and eukaryal systems, that the DNA template topology has a negligible impact on transcription levels at high, physiological temperatures (Bell et al. 1998). However, at lower temperatures, negatively supercoiled templates were shown to be more highly transcribed than those that are positively supercoiled. Notably, the authors established that transcription was blocked at the level of promoter opening in positively supercoiled templates, suggesting that modulation of promoter topology is a mechanism to maintain gene expression under non‐physiological temperature conditions.

Several Sulfolobales species (and Thermococcales species, see the next chapter) containing plasmids were isolated in the 1990s, thus opening a possibility of broadening the study of DNA topology dynamics from a comparative point of view. These early studies from the Forterre group provided a wealth of reference data by analyzing the topological variation of Sulfolobus plasmids as a function of growth phase and growth temperature, as well as the effect of temperature stresses (cold and heat shock) (López‐García and Forterre ¹⁹⁹⁷). This revealed that, depending on the species studies, plasmid DNA topology can remain relatively unchanged throughout the growth curve, or the linking number (Lk) (Figure 1) can increase as cells approach stationary phase at the optimal growth temperature. The kinetics of topological changes was followed by exposing cultures to sudden temperature upshifts from 80°C to 85°C or downshifts from 80°C to 65°C. Since the studied Sulfolobus species could grow at these non‐optimal temperatures, this allowed the study of reversible changes in topology over time. The heat shock induced a fast (15–30 min) and quite spectacular increase in Lk (from typically +0.01 to +0.025, e.g., for a 5 kb plasmid, this would correspond to increasing the number of positive supercoils from 5 to 13). Subsequently, the plasmids slowly recovered lower Lk numbers down to those found normally when growing at 85°C (López‐García and Forterre ¹⁹⁹⁷).

Cold shock produced the opposite effect, whereby the plasmid Lk experienced a sharp decrease (from typically +0.010 to negative values down to −0.020), followed by a slow recovery until reaching a Lk typical for growth at 65°C (López‐García and Forterre ¹⁹⁹⁷). Since there is no known enzyme that actively introduces negative supercoiling in Sulfolobus, a plausible explanation for this phenomenon is that such topology results from the binding of NAPs in combination with topoisomerase activity. A good candidate is Sso7, one of the most abundant NAPs (~5% of total protein) that was shown to intercalate into the minor groove, thereby increasing the helical repeat (and thus decreasing the twist) of DNA and inducing positive supercoiling that could be relaxed by a topoisomerase. Once the Sso7 is released from the DNA, the twist increases and is compensated by the introduction of negative supercoils (Robinson et al. 1998; López‐García et al. ¹⁹⁹⁸). A model recapitulating these findings is shown in Figure 4D.

Following these results, a model emerged in which variation in DNA topology (specifically linking number) is a modulable tool used by Sulfolobales for counteracting the effect of temperature on DNA structure and further pointed to the existence of homeostatic regulatory systems controlling DNA topology as part of the stress response. Since only topoisomerase activity can explain changes in plasmid Lk numbers (even if it is indirectly induced by, for example, NAP binding), the follow‐up studies aimed at paralleling these changes with the quantities and activities of Sulfolobus topoisomerases. Lopez‐Garcia and Forterre quantified reverse gyrase and Topo VI in cells exposed to heat or cold shock and found evidence suggesting that an increase in topoisomerase levels is not needed for control of DNA topology during thermal stress. Rather, the reverse gyrase activity detected in crude extracts was strongly dependent on assay temperature, with its activity being greatly enhanced at 85°C (Lopez‐Garcia and Forterre ¹⁹⁹⁹). Intriguingly, reverse gyrase activity in crude extracts from heat‐shocked cells was shown to resist significantly longer to supraoptimal temperatures as compared to extracts from cells that grew at optimal temperature, suggesting that reverse gyrase can become thermoresistant. Using puromycin to poison translation, the authors further showed that thermoresistance required protein synthesis to be functional, but the responsible factor was not identified (Lopez‐Garcia and Forterre ¹⁹⁹⁹). A general mechanism, such as the synthesis of heat shock proteins, or a more specific one, such as posttranslational modification of reverse gyrase, might be behind the observed thermoresistance; however, this remains to be investigated. In more recent studies from Nadal's group, the transcriptional and protein levels of TopR1 and TopR2 were followed separately. This showed that only TopR1 exhibited small quantitative variations following temperature shifts and higher activity at supraoptimal temperatures (Garnier and Nadal 2008; Couturier et al. ²⁰²⁰). In a recent study, the global response of S. acidocaldarius to heat shock was measured both at the transcriptional and proteomic level, showing that, in terms of protein levels, only TopR1 and Topo VI reacted, albeit at small scale. Interestingly, significant variations were observed for various NAPs, with the increase in abundance of three Lrs14‐type NAP homologs and the decrease in abundance of Sul12a being the most drastic changes (Baes et al. 2022).

In summary, the current evidence points to differential activation and/or inhibition mainly of TopR1 activity and changes in the levels or binding capacities of DNA architectural proteins as pivotal factors in generating opposite topological responses upon heat shock and cold shock in Sulfolobus species. This could be an adaptive mechanism in the natural environments of Sulfolobales, where temperatures may greatly vary, helping to maintain genome stability and expression under non‐optimal temperature conditions.

Beyond this stress response mechanism, the role of TopR1 and the remaining Sulfolobus topoisomerases in solving topological problems arising from genome expression (Figure 4B), DNA replication (Figure 4C), and decatenation of replicated chromosomes (Figure 4E) is a fundamental question that remains to be answered.

5. Regulation of DNA Topology in Thermococcales Model Species

Thermococcales are hyperthermophilic archaea that belong to the phylum Methanobacteriota (formerly Euryarchaeota) (Rinke et al. 2021; Parks et al. ²⁰²²), with three genera Thermococcus, Pyrococcus and Paleococcus whereby within the last genus there are no established model species yet. Thermococcus and Pyrococcus model species are all hyperthermophilic sulfur‐dependent heterotrophs that grow optimally between 85°C and 100°C with a doubling time of about 30–40 min (Fiala and Stetter 1986; Erauso et al. ¹⁹⁹³; Atomi et al. ²⁰⁰⁴).

Historically, the best studied Pyrococcus species are Pyrococcus abyssii GE5 and P. furiosus DSM3638 with the former one being isolated at a depth of 2000 m from a hydrothermal vent in the southwest Pacific (Marteinsson et al. 1995) and described in 1993 (Erauso et al. 1993) while the latter was isolated by Karl Stetter and a co‐worker in 1986 from geothermally heated marine sediments at the beach of the Vulcano island in Italy (Fiala and Stetter 1986). The genome of P. abyssi GE5 was the first archaeal genome to be completely sequenced and published in 2003 (Cohen et al. 2003) and this opened the way for studies of archaeal DNA replication, gene expression and development of genetic tools for hyperthermophiles. While no such tools are available for P. abyssi GE5, a naturally competent derivative strain COM1 of P. furiosus emerged as the main genetically modifiable model for this genus (Lipscomb et al. 2011). T. kodakarensis KOD1 is also naturally competent for genetic transformation and is a widely studied experimental model species with the first tools for chromosomal manipulations being developed as early as 2003 (Sato et al. 2003). The genetic toolbox for Thermococcales and in particular T. kodakarensis was greatly expanded in the following years (Hileman and Santangelo 2012; Farkas, Picking, and Santangelo ²⁰¹³) thus pushing forward this organism as one of the best studied archaeal model species.

Thermococcales are phylogenetically distinct from the Sulfolobales (Figure 2) we described above and this is reflected in fundamental differences in terms of DNA metabolism and, in particular, genome structure (Figure 5A). To begin with, Thermococcales are all polyploid organisms, with, in the case of T. kodakarensis, the chromosome copy number varying between 7 and 19 copies, depending on the growth phase (Spaans, van der Oost, and Kengen ²⁰¹⁵). Genome replication is initiated at a single replication origin and not multiple origins as in Sulfolobales, and, remarkably, the replication origin can be removed without deleterious impact on the growth rate (Gehring et al. 2017), with cells in this case most likely relying upon RadA‐supported recombination‐dependent replication (Mc Teer et al. 2024).

FIGURE 5.

Genome characteristics and DNA topology regulation in Thermococcales. (A) Schematic representation of the genome of T. kodakarensis. Cells of T. kodakarensis contain between 7 and 19 chromosomes. The genome is organized into CIDs and is densely covered by histones (HTkA, HtKB) that form dimers and multiples of dimers, wrapping 30 nucleotides per dimer. In addition, DNA is bound by NAPs, among which Alba and TrmBL2 are the most abundant. For simplicity, the binding mode and the relative abundance of these NAPs are not accounted for in the scheme. DNA replication is initiated from a single origin of replication (ori) or by the RadA‐driven recombination‐dependent process. (B) Proposed model for the topoisomerase‐driven resolution of topological constraints generated by transcription. The arrow indicates the direction of transcription and the messenger RNA is depicted in violet. All three DNA topoisomerases encoded by T. kodakarensis could be involved in relaxing negative DNA supercoils building up behind a transcribing RNA polymerase. Topo VI is predicted to relax positive supercoils accumulating in front of the RNA polymerase. (C) Proposed model for solving DNA replication‐induced topological constraints. For clarity, only the replicative helicase (MCM) and the DNA polymerases are shown. As the replisome progresses, the two newly copied molecules become entangled and form precatenanes that could presumably be solved by the decatenase activity of Topo III. Topo VI would act in front of the replisome to relax positive supercoils. The arrows indicate the direction of progression of the replication machineries. (D) Observed changes in DNA topology of plasmids in response to temperature stress. When cells of T. sp. GE 31 (a species closely related to T. kodakarensis) are exposed to heat shock, the plasmid DNA rapidly becomes more positively supercoiled, presumably through the action of the reverse gyrase. Under cold shock, the observed negative supercoiling topology could be established through the joint action of Topo VI, histones and perhaps NAPs. The indicated supercoiling densities have been reported by López‐García and Forterre (1997). (E) Proposed model for the decatenation of two fully replicated chromosomes. Topo VI is the enzyme that likely can induce double‐stranded DNA break needed to decatenate the two catenated chromosomes (right part of the panel). Upon the end of replication, the two daughter chromosomes could still be linked by the parental strands, leading to the generation of hemi‐catenanes. Such structures contain single‐stranded regions and could be solved by a type I enzyme such as Topo III that has shown a robust decatenase activity in vitro. Created with BioRender.com.

Genome compaction and folding are two other features that distinguish Thermococcales from Sulfolobales. An important fact with regard to DNA topology is that Thermococcales use histones to compact their DNA. The two histones encoded by T. kodakarensis, HTkA and HTkB, which share a conserved histone fold domain with eukaryotic histones, have been examined in some depth, both in vitro and in vivo. HTkA and HTkB were shown to assemble into long oligomers, named hypernucleosomes, in vivo with histone dimers being the smallest particles, wrapping 30 bp of DNA (Maruyama et al. 2013; Mattiroli et al. ²⁰¹⁷; Henneman et al. ²⁰¹⁸; Bowerman, Wereszczynski, and Luger ²⁰²¹; Sanders et al. ²⁰²¹). Using quantitative mass spectrometry, a recent study established that histones were highly abundant in T. kodakarensis (1.76% of total protein), suggesting that the genome of this archaeon is extensively coated with histones (Hocher et al. 2022).

How chromatinization impacts the local topology of chromosomal DNA in vivo in T. kodakarensis (and in general in histone‐encoding Archaea) is not known. In vitro experiments using plasmid DNA show that the orientation of DNA wrapping around the histones greatly depends on the physico‐chemical parameters and protein‐DNA mass ratio, and the formation of both positive and negative supercoils around the histone particles has been reported (Musgrave, Sandman, and Reeve ¹⁹⁹¹; Ronimus and Musgrave ¹⁹⁹⁶). At physiological salt and temperature conditions, histones of T. zilligii, a species closely related to T. kodakarensis, were shown to induce negative supercoiling in plasmid DNA, suggesting that this may be the predominant physiological mode of DNA supercoiling in Thermococcus nucleosomes (Musgrave, Forterre, and Slesarev ²⁰⁰⁰). Notably, similar to eukaryotes, histones are excluded from gene regulatory DNA in T. kodakarensis (Maruyama et al. 2013) suggesting that these regions may be particularly sensitive to heat denaturation and thus a potentially preferential target for topoisomerases in this organism.

Regarding genome folding, only one Hi‐C map has been reported so far for Themococcales: that of T. kodakarensis, and it showed that the genome is structured, at its highest level, into bacterial‐like CIDs, but it remains to be determined how those domains are defined and structured within (Cockram et al. 2021).

The “topological kit” that Thermococcales are equipped with consists of two Type IA topoisomerases (RG and Topo III) and one Type IIB topoisomerase (Topo VI) to which histones need to be added as important players in modulating local DNA topology. Such combination of actors is also found in Methanobacteria and Methanococci, but not in other euryarchaeal lineages (Garnier et al. 2021) that have their own specificities, as exemplified by Haloarchaea, which we will describe in the next section.

RG was shown to be non‐essential at the optimal growth temperature of T. kodakarensis, but how the absence of this enzyme affected DNA topology was not studied. However, at supraoptimal temperatures, the growth rate of the KO strain becomes strongly affected, with 93°C being the threshold temperature where growth stops completely (Atomi, Matsumi, and Imanaka ²⁰⁰⁴). To our knowledge, a genetic investigation of the other two topoisomerases was not reported yet, but it is likely that Topo VI is essential given that it is the only Type II topoisomerase in this organism. As for Topo III, random saturation transposon mutagenesis suggests that this enzyme is essential in another euryarchaeon Methanococcus maripaludis (Sarmiento, Mrázek, and Whitman ²⁰¹³), but this remains to be rigorously tested.

In the absence of established methods to study supercoiling on chromosomes of archaea, most data regarding the variation of supercoiling in Thermococcales come from early experiments where plasmid DNA topology was analyzed under various growth conditions. However, in contrast to similar Sulfolobus studies, no attempt has been made yet to establish a link between observed variations in DNA topology and topoisomerase levels and activities.

In two studied species, one Thermococcus and one Pyrococcus, plasmid topology remained stable throughout different growth phases. Intriguingly, however, when Thermococcus sp. GE31 was grown at the upper limit temperature, the plasmid had lost much of its positive supercoiling, while the opposite—an increase in positive supercoiling—was observed for Sulfolobus plasmids (López‐García and Forterre ¹⁹⁹⁷), suggesting that mechanistic differences exist in how Sulfolobales and Thermococcales adapt to prolonged temperature stress. When short‐term temperature stresses were applied by shifting growth temperature higher or lower than optimal, the Thermococcales followed the same trend as Sulfolobales, whereby the plasmid linking number rapidly decreased following cold shock and rapidly increased following a heat shock (López‐García and Forterre ¹⁹⁹⁷) (Figure 5D). Interestingly, slow cooling of P. abyssi cultures, as opposed to rapidly cooled cells, resulted in more negatively supercoiled plasmid DNA (Marguet, Zivanovic, and Forterre ¹⁹⁹⁶), and the authors speculated that such a situation may be physiologically relevant, with hyperthermophiles frequently subjected to transient cooling in their natural environment. It remains to be established if the observed changes in DNA topology are part of a genuine mechanism to help Thermococcales survive in mesophilic or psychrophilic conditions or simply artefacts due to aberrant topoisomerase activities at non‐optimal temperatures.

Unlike bacteria, where the global supercoiling level is tightly regulated by opposing activities of DNA gyrase and Topo I (Drlica 1992), T. kodakarensis seems to be highly resistant when exposed to chronic and pervasive topological stress (Villain et al. 2021). This was achieved by expressing a thermostable gyrase in this organism which, as judged from plasmid topology, drastically changed the topology of DNA by converting the naturally positively supercoiled DNA of T. kodakarensis to moderately negatively supercoiled DNA (σ ~ −0.03 as compared to σ ~ −0.06 in E. coli). Intriguingly, among endogenous topoisomerases, only the expression of reverse gyrase was upregulated, albeit lowly (0.25‐fold increase), while the expression of the other two topoisomerases remained unchanged suggesting that gyrase activity was not compensated, as one could expect, by the increased abundance of any of the three endogenous topoisomerases. Despite this unnatural topological state of its DNA, the organism grew at a normal rate under optimal growth conditions, but the cells were in average smaller (~30% volume loss). Interestingly, these cells strongly overproduced the archaellum (archaeal flagellum) components suggesting that they responded to stress by becoming hyper‐flagellated (Villain et al. 2021). As the gyrase‐induced artificial increase in negative supercoiling could mimic a cold shock, one could speculate that overexpression of archaella is a mechanism allowing Thermoccoccus to swim away from cold water towards higher temperatures, thus hinting that, as in bacteria (Dorman and Dorman 2016), changes in DNA topology could serve as a sensor such that the appropriate global transcriptional response is induced to help the cells adapt to environmental cues.

As for the housekeeping activities of topoisomerases from Thermococcales in handling supercoil build‐up during transcription (Figure 5B), DNA replication (Figure 5C) as well as chromosome decatenation (Figure 5E), plausible hypotheses can be proposed to serve as a basis for a future experimental work.

6. Regulation of DNA Topology in H. volcanii

H. volcanii belongs to the phylum Halobacteriota (formerly Euryarchaeota) (Figure 2), and it was initially isolated in 1975 from the Dead Sea (Mullakhanbhai and Larsen 1975). It is an obligate halophile and mesophile that grows aerobically at 45°C and 1.7–2.5 M NaCl and uses a “salt‐in” mechanism to keep the cytoplasmic salt concentration at the same molarity as the external environment (Gunde‐Cimerman, Plemenitaš, and Oren ²⁰¹⁸).

H. volcanii is highly polyploid, with up to 40 copies of its chromosomes (2.8 Mb) per cell (Maurer et al. 2017), and it also harbors three circular mini‐chromosomes: pHV4 (636 kb), pHV3 (438 kb) and pHV1 (85 kb) (Breuert et al. 2006; Hartman et al. ²⁰¹⁰) (Figure 6A). The main chromosome features three active replication origins, and, similar to Thermococcus, all can be deleted without adverse effect on the growth of the organism. Instead, and rather unexpectedly, H. volcanii grows faster (!) in their absence, albeit under controlled laboratory conditions (Hawkins et al. 2013).

FIGURE 6.

Genome characteristics and DNA topology regulation in Haloferax volcanii. (A) Chromosome structure of H. volcanii. The organism is polyploid with 2–40 copies of the genome per cell. The chromosome is folded into CIDs of variable size that contain loop structures (not shown) stabilized by a canonical SMC protein. H. volcanii encodes one histone protein (HstA) that does not play a canonical role in chromosome packaging, one confirmed NAP (MC1) and two putative ones (HVO_1577 and HVO_2029). For simplicity, the binding mode and the relative abundance of these NAPs are not accounted for in the scheme. Chromosomes are replicated bidirectionally from three origins of replication; interestingly though, the origins can be deleted without a negative impact upon growth. Recombination‐dependent replication is proposed as an alternative to canonical replication mechanism initiated from replication origins. (B) Proposed model for how transcription‐induced supercoiling is handled by DNA topoisomerases of Haloferax. The positively supercoiled DNA accumulating ahead of the transcribing RNA polymerase is the likely substrate for both DNA gyrase and Topo VI. Topo III and perhaps also Topo VI could be involved in relaxing negative DNA supercoils occurring behind the polymerase. The sense of progression of the RNA polymerase is indicated by the arrow and the growing messenger RNA is depicted in violet. (C) Proposed model for solving DNA replication‐induced topological constraints. For clarity, only the replicative helicase (MCM) and the DNA polymerases are depicted. The progression of the replication fork induces positive supercoiling ahead of the polymerase, which could be relaxed by the action of both DNA gyrase and Topo VI. The precatenanes forming behind the polymerase could be disentangled by the Topo III. The direction of replication is indicated by the arrows. (D) Observed topological changes in plasmid DNA induced by temperature stress. Growth of H. volcanii at a lower‐than‐optimal temperature (30°C) results in an increase in negative supercoiling, while growth at a supraoptimal temperature (50°C) leads to a lower amount of negative supercoiling as compared to plasmid DNA isolated from cells grown at 37°C (Mojica et al. 1994). The increase in negative supercoiling could be due to the negative supercoiling activity of DNA gyrase, while Topo VI and/or Topo III could relax negative supercoiling. (E) Proposed model for chromosome decatenation. Both DNA gyrase and Topo VI are plausible candidates for separating two fully replicated and catenated chromosomes by introducing a transient double‐stranded brake in one of the molecules and passing the second molecule through the break (the right part of the panel). In case the replication ends with the formation of hemicatenanes, Topo III would be a likely candidate for solving these structures (the left part of the panel). Created with BioRender.com.

H. volcanii encodes one histone ortholog, HstA; however unlike Thermococcales, it is not abundant at either the transcript (Rojec et al. 2019) or protein level (Hocher et al. 2022). Following this observation, Hocher and colleagues suggested that HstA likely has a limited role in DNA compaction, with an as‐of‐yet uncharacterized protein or set of proteins actually being responsible for the widespread protection from micrococcal nuclease digestion that was previously observed in this species (Ammar et al. 2012). In line with this hypothesis, a recent study showed that HstA binds to the genome infrequently and in discrete regions that can be both coding and intergenic, thus exhibiting features that are not compatible with this protein facilitating genome‐wide chromatin formation (Sakrikar et al. 2023).

The three‐dimensional structure of H. volcanii genome was studied in some detail, and it highlighted significant differences with that of the other studied model, Sulfolobus (Cockram et al. 2021). The main chromosome of H. volcanii is organized into CIDs ranging in size from 25 to 570 kb, and their boundaries correlate with strong gene expression. However, the chromosome is not organized into transcriptionally active and inactive compartments as in Sulfolobus. A high number of discreet loop structures (64 in average) was found nested within CIDs, and their presence was associated with transcriptional activity and the canonical structural maintenance of chromosomes (SMC) protein (Cockram et al. 2021).

Different from the two models we described previously, in addition to Topo III and Topo VI, H. volcanii encodes DNA gyrase (Garnier et al. 2021), and this feature was exploited by early studies in the 1990s to set up the first genetic tools for this archaeon that were relient upon the sensitivity of this organism to gyrase inhibiting drugs (Holmes and Dyall‐Smith ¹⁹⁹⁰). This, together with easy cultivation, opened the way toward the development of an extensive repertoire of genetic, molecular biology and biochemical tools, making H. volcanii one of the main archaeal model organisms (Pohlschroder and Schulze 2019).

As described above, DNA gyrase, a Type IIA enzyme, was acquired from a bacterial donor at an early stage in the evolution of the archaeal domain and has since spread to several phylogenetically distinct lineages, including Asgard archaea (Villain et al. 2022). The introduction of an enzyme with a strong impact on DNA topology, and thus all DNA transactions, was presumably not a small matter for the recipient archaeal cell. H. volcanii is therefore a fitting model to study how the DNA gyrase and endogenous topoisomerases co‐evolved in response to this event. A particular question arises as to how DNA gyrase and Topo VI activities have been accommodated to avoid functional overlap, as both of these enzymes relax positive supercoils and can act as decatenases.

In the 1990s, it was shown that DNA gyrase activity is a dominant one in H. volcanii and is responsible for the negatively supercoiled state of native plasmids and megaplasmids in this organism (Sioud et al. 1988; López‐García et al. ¹⁹⁹⁴; Charbonnier and Forterre ¹⁹⁹⁴). When H. volcanii was grown at supraoptimal temperatures, the plasmid linking number increased, similar to what was observed in Sulfolobus species, suggesting that the temperature effect on DNA topology was not compensated by endogenous topoisomerases (Mojica et al. 1994). Interestingly, under optimal growth conditions, the in vivo relaxation of DNA by intercalating drug chloroquine (which in vitro results in increase in positive supercoiling to compensate for lower twist) induced an increase in negative supercoiling of plasmid DNA. This suggests that, in those conditions, DNA gyrase is used to re‐establish native supercoiling level, indicating the existence of a homeostatic control of DNA supercoiling in H. volcanii through the activity of DNA gyrase (Mojica et al. 1994) (Figure 6D).

In Eukaryotes, the regulation of topoisomerase functions is accomplished by post‐translational modifications (PTMs) that modulate enzyme activity and likely play key roles in determining sites of enzyme action and enzyme stability (Chikamori et al. 2010; Bedez et al. ²⁰¹⁸). Similarly, proteomics studies showed that E. coli Topo I was acetylated at lysine residues that make contacts with DNA (K. Zhang et al. ²⁰¹³) and that loss of deacetylase CobB results in reduced intracellular TopA catalytic activity and increases negative DNA supercoiling in vivo (Zhou et al. 2017). Whether a PTM‐based mechanism operates in Archaea, specifically in relation with DNA topoisomerases, has, to our knowledge, not yet been studied. However, valuable insights came from a recent study that investigated the dynamics of the acetylome in H. volcanii under oxidative stress conditions (Couto‐Rodríguez et al. ²⁰²³). Strikingly, this study reported that all three endogenous topoisomerases carried acetylated lysins, and notably, increased acetylation was associated with reduced abundance of topoisomerases, hinting that Archaea, as the other two domains of life, relay upon PTMs as regulators of topoisomerase activity (Couto‐Rodríguez et al. ²⁰²³). The inspection of the location of the acetylated lysins in the predicted or solved structures of these topoisomerases suggests that these PTM's might influence DNA binding or activity of topoisomerases in H. volcanii.

Similar to the two models described above, a model, guided mostly by biochemical data, can be devised as to how the topoisomerases of H. volcanii cooperate to release topological constraints that accumulate during transcription (Figure 6B), chromosome replication (Figure 6C) and decatenation of replicated chromosomes (Figure 6E).

7. Perspectives and Open Questions

With their typically prokaryotic features and unique position in the tree of life as direct ancestors of Eukaryotes, the Archaea are ideal models to understand the transition from a small and relatively simple prokaryotic cell to a significantly more complex and larger eukaryotic cell, a phenomenon that remains one of the greatest unresolved mysteries in Biology. Cells have to deal with entangled DNA all the time, and this physical constraint is central to all fundamental DNA processes and defines cellular physiology. As we have seen, the Archaea use a variety of strategies to manage the topology of their DNA using a unique combination of bacterial, eukaryotic‐like and typical archaeal features, the workings of which in vivo are barely beginning to be studied. If we are to close this knowledge gap, in addition to classical genetic and biochemical approaches, the field will need to implement novel NGS‐supported genome‐wide assays that have been successfully used in Bacteria and Eukaryotes.

In particular, the genome wide mapping of negative and positive supercoiling using intercalating drug psoralen and its derivatives (e.g., TMP‐seq (Kouzine, Baranello, and Levens ²⁰¹⁸), Psora‐seq (Visser et al. 2022)) and GapR protein (GapR‐seq (Guo et al. 2021)) in combination with techniques to map topoisomerase binding and cleavage across chromosomes (McKie, Maxwell, and Neuman ²⁰²⁰) are dearly needed to address the physiological role of archaeal topoisomerases and understand how supercoiling affects 3D spatial organization, stability, and expression of archaeal genomes. This may be a challenging endeavor though, as the extremophilic growth conditions of main archaeal model organisms (high salt, high temperature, extremely acidic medium) could interfere with the stability of psoralen derivatives or with the efficiency of UV‐induced photobinding. The same goes for GapR protein, which is so far only found in mesophilic bacteria and would presumably not withstand high temperature or high salt conditions. The very high GC content of the genome of Haloferax could be an obstacle too, as psoralen is known to preferentially intercalate into AT‐rich DNA sequences.

Despite these potential difficulties it is important to develop these tools for archaea as they will enable us to address some of the longstanding and key questions in the field:

1. What are the principles and diversity of DNA supercoiling regulation strategies in Archaea and how are those mechanisms related to eukaryotic counterparts?

2. What is the impact of supercoiling on DNA replication initiation/termination, gene expression, genome folding and stability?

3. How is global genome topology modulated in response to stress and what are the underlying molecular mechanisms?

4. What is the contribution of DNA‐binding proteins (histones, NAPs) to the regulation of DNA topology and how is the crosstalk achieved with topoisomerases?

5. How is the propagation of mobile genetic elements and horizontal gene transfers in general affected by DNA topology?

Answering these questions will significantly advance the understanding of archaeal fundamental biology but will also have more broad impact, notably for the understanding of the evolution of DNA metabolism in general, for a better appreciation of the diversity of DNA topology regulation strategies, or gene expression tuning for applications in synthetic biology. Given the relevance of research on DNA topology regulation in Archaea and the strong potential for original and high‐impact results, we are confident that a bright and exciting future awaits this avenue of research.

Author Contributions

Paul Villain: conceptualization, investigation, writing – review and editing. Tamara Basta: conceptualization, investigation, funding acquisition, writing – original draft, writing – review and editing.

Ethics Statement

The work presented here did not include human or animal subjects nor human or animal material or data. Thus, no formal consent or approval was necessary.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.