DNA replication is a fundamental cellular process that is functionally conserved across all three domains of life (bacteria, archaea, and eukaryote). The precise regulation of DNA replication ensures the accurate duplication of genomic information, and replication initiation is the first and most important stage of this regulation. The first model of DNA replication initiation was proposed for Escherichia coli in 1963, postulating that a trans-acting factor binds to a cis-acting site which triggers replication initiation (Jacob et al., 1963). In the subsequent 50 years, this “replicon model” has been demonstrated to be essentially true in all organisms, and the cis-acting site is now known as the replication origin. Bacterial chromosomes are typically replicated from a single origin, whereas the replication of eukaryotic chromosomes initiates from a number of discrete origins (Leonard and Mechali, 2013). DNA replication origins have been well-defined in bacteria and unicellular eukaryotes, and relative topics are covered in a number of excellent reviews (Messer, 2002; Mott and Berger, 2007; Zakrzewska-Czerwinska et al., 2007; Mechali, 2010; Aparicio, 2013). In contrast, focus on DNA replication origins in archaea, the third domain of life, commenced only approximately a decade ago. DNA replication origins have been mapped primarily for a few representatives of archaeal species distributed in the three main phyla, Euryarchaeota, Crenarchaeota, and Thaumarchaeota (Myllykallio et al., 2000; Lundgren et al., 2004; Robinson et al., 2004; Grainge et al., 2006; Norais et al., 2007; Majernik and Chong, 2008; Coker et al., 2009; Pelve et al., 2012, 2013; Wu et al., 2012, 2014). In addition, more detailed characterization has been revealed in several model systems, such as Pyrococcus species (Myllykallio et al., 2000; Matsunaga et al., 2001, 2003), Sulfolobus species (Lundgren et al., 2004; Robinson et al., 2004; Duggin et al., 2008; Samson et al., 2013), Haloferax volcanii (Norais et al., 2007; Hawkins et al., 2013) and Haloarcula hispanica (Wu et al., 2012, 2014). It is now known that archaea use a single or multiple origin(s) to replicate their circular chromosomes (Kelman and Kelman, 2004; Robinson and Bell, 2005; Hyrien et al., 2013). The multiple origins together with their adjacent initiator genes in certain archaeal chromosomes may have arisen from the capture of extrachromosomal elements and appear to be mosaics of distinct replicator–initiator pairings (Robinson and Bell, 2007; Wu et al., 2012).

This replicator–initiator system consists of an origin region and an initiator gene (the cdc6 gene in most cases and whiP in the oriC3 of Sulfolobus species). The origin region usually has a high content of adenine and thymine residues (AT-rich) flanked by several conserved repeated motifs known as origin recognition boxes (ORBs). The initiator protein Cdc6 (also denoted Orc or Orc1/Cdc6 in other papers) shows homology to both Orc1 and Cdc6 of eukaryotes and therefore is considered to be involved in both specific recognition of the origin region and loading of the minichromosome maintenance helicase (MCM; Robinson and Bell, 2005). Despite the conservation of the replicator-initiator structure, archaeal replication origins exhibit considerable diversity in terms of both the ORB elements within different origins and their adjacent initiator genes. The specificity of the initiator genes and their cognate origins was recently established by means of in silico and genetic analyses in Sulfolobus species (Samson et al., 2013) and Haloarcula hispanica (Wu et al., 2012, 2014). The cis organization of the replication origin and the initiator gene (replicator–initiator) is reminiscent of the oriC-dnaA system in bacteria. Recently, we revealed that bacterial-like control mechanisms may be used by different replication origins in haloarchaea, and a model has been proposed for the negative regulation of oriC2 by a downstream cluster of ORB elements in Haloarcula hispanica (Wu et al., 2014).

The goal of this review is to present an overview of progress made over the past decade in our understanding of DNA replication origins of archaeal genomes, including the identification (mapping), characterization and evolution of multiple replication origins on the chromosomes. We focus on the utilization and control of multiple replication origins in haloarchaea, as well as comparisons of replication origins from different archaeal species to draw the generality and evolution of multiple replication origins in archaea.

Similar to bacteria, archaea have simple circular chromosomes (and also several extrachromosomal elements in some archaea); however, many archaea characterized to date harbor multiple replication origins. The approaches developed in bacteria or eukaryotes have been employed to investigate replication origins in archaea, such as GC-skew analysis (Myllykallio et al., 2000), the Z-curve method (Zhang and Zhang, 2003), autonomously replicating sequence (ARS) assay (Berquist and DasSarma, 2003; Norais et al., 2007; Wu et al., 2012), replication initiation point mapping (RIP mapping; Matsunaga et al., 2003), two-dimensional gel analysis (Matsunaga et al., 2001; Robinson et al., 2004), and marker frequency analysis (MFA; Lundgren et al., 2004; Coker et al., 2009; Pelve et al., 2012; Hawkins et al., 2013; Wu et al., 2014). DNA replication origins have been mapped in about a dozen archaeal species.

The initiator protein DnaA is highly conserved in bacteria, and bacterial replication origins generally possess conserved sequence elements, DnaA boxes. In contrast, the three replication origins in Sulfolobus species differ from each other. Each of the three origins is specifically recognized by its proximally encoded initiator protein, two distinct Cdc6 proteins and WhiP (Dueber et al., 2011; Samson et al., 2013). In addition, the recognition mechanisms appear to be different, as classic ORB and its shorter version (miniORB) are, respectively, observed in the oriC1 and oriC2 regions, while neither is observed in the oriC3 region (Robinson et al., 2004; Samson et al., 2013).

Haloarchaeal genomes generally contain multiple cdc6 genes and replication origins. Recently, we conducted a comparison of the origin-associated Cdc6 homologs and the corresponding predicted ORB elements. Our results suggested that the replication origins from haloarchaeon are notably diverse in terms of ORB elements and their adjacent cdc6 genes, which could be sorted into distinct families. Based on this phylogenetic analysis, linkage-specificity of Cdc6 homologs and the corresponding ORB elements was proposed, suggestive of their specific interaction (Wu et al., 2012). Very recently, we employed comprehensive genetic studies to investigate the specificity of multiple replication origins and cdc6 genes in Haloarcula hispanica, and our results indicated that each Cdc6 protein specifically recognizes its proximal origin (Wu et al., 2014). Thus, multiple replication origins along with their adjacent cdc6 genes appear to be distinct ori-cdc6 systems. These distinct ori-cdc6 systems in haloarchaeon may have many evolutionary advantages: first, it ensures the compatibility of multiple replication origins, which accounts for the observations that multiple Cdc6 proteins from a haloarchaeal genome are distributed into different families (Wu et al., 2012) and that the oriC2-containing plasmid is incompatible with Haloarcula hispanica (Wu et al., 2014); second, distinct ori-cdc6 pairings help minimize competition among multiple origins for initiators and maintain independent control of replication initiation at different origins. Importantly, as haloarchaeal genomes generally contain multiple replicons, distinct ori-cdc6 origins may be favorable for replicon-specific replication control, similar to the different modes of replication origin adopted by the two chromosomes of Vibrio cholerae (Egan and Waldor, 2003).

To understand the molecular mechanisms involved in the specific recognition of origins by initiators, the structures of two origin-bound Cdc6 proteins from Aeropyrum pernix (Gaudier et al., 2007) and S. solfataricus (Dueber et al., 2007) were crystallized. Both of the two Cdc6 proteins contain an N-terminal AAA⁺ domain and a C-terminal WH domain. Intriguingly, both of the studies demonstrated that, in addition to the canonical DNA binding WH domain, the AAA⁺ domains of these two initiators are responsible for recognizing origins (Dueber et al., 2007; Gaudier et al., 2007). Subsequently, biochemical data also demonstrated that both the WH domain and AAA⁺ domain contribute to the origin-binding specificity of the Cdc6 protein (Dueber et al., 2011).

Multiple mechanisms that regulate replication initiation have been well-characterized in both bacteria and unicellular eukaryotes, and are summarized in a number of excellent reviews (Mott and Berger, 2007; Mechali, 2010; Rajewska et al., 2012; Aparicio, 2013). In contrast, the mechanisms by which archaea regulate replication initiation at multiple origins, either on the same chromosome or from different genetic elements, are far less understood. All of the archaeal replication origins characterized to date are dependent on their adjacent initiator gene (the cdc6 gene in most cases; Samson et al., 2013; Wu et al., 2014), and these distinct ori-cdc6 pairings may contribute to their independent control. In addition, the cis location of the cdc6 gene and the origin is proved to not be required for ARS activity in both Haloferax volcanii and Haloarcula hispanica (Norais et al., 2007; Wu et al., 2014). Therefore, we have proposed that direct linkage of the initiator gene to the origin may facilitate its transcription after replication initiation to sequentially control its cognate origin.

Using the Haloarcula hispanica model system, we suggested that some bacterial-like mechanisms may be employed at different replication origins in haloarchaea (Wu et al., 2014). A G-rich inverted-repeat directly inside each ORB element of Haloarcula hispanica oriC1 was shown to be a replication enhancer that stimulated origin activation at oriC1. Because of the repeat’s close location to ORB elements, we proposed that the G-rich inverted-repeat enhances the binding of initiator or regulatory factors at oriC1, similar to many repeated sequences in bacteria that are binding sites for initiation proteins or regulatory factors, playing a crucial role in the control of replication initiation (Rajewska et al., 2012). In addition, a model has been proposed, and partly tested, for the negative regulation of oriC2 by a downstream cluster of Cdc6 binding elements (ORBs), likely via Cdc6E titration, similar to the negative control of replication initiation via a datA locus exhibiting DnaA-titrating activity in E. coli (Kitagawa et al., 1998). More interestingly, many additional predicted replication origins have the oriC2-like structure, suggesting that this strategy of negative replication origin control is used generally by haloarchaea.

Despite the bacterial-like structure of archaeal replication origins, archaea use eukaryotic-type replication machinery (Robinson and Bell, 2005), indicating that archaea may adopt eukaryotic-like mechanisms to control replication proteins and thus replication initiation. Interestingly, genome-wide transcription mapping indicated that serine–threonine protein kinases show cyclic induction in Sulfolobus species, indicating that regulatory factors similar to eukaryotic cyclin-dependent kinase (CDK) complexes may be present in archaea (Lundgren and Bernander, 2007). Recently, an ATP-ADP binary switch model for Cdc6-mediated replication control was proposed in S. islandicus, postulating that binding of ATP remodels Cdc6 conformation for efficient MCM recruitment, and subsequent ATP hydrolysis renders Cdc6 incapable of further recruiting MCM (Samson et al., 2013). In addition, as almost all replication origins are dependent on Cdc6 proteins, conformational changes of Cdc6 proteins may play important roles in coordinating replication initiation at different origins within a cell.

Although considerable diversity of replication origins has been observed in haloarchaea, comparison analysis revealed a conserved replication origin, oriC1, which is positioned in the main chromosome of all analyzed haloarchaeal genomes (Coker et al., 2009; Wu et al., 2012). Both the ORBs within oriC1 and the oriC1-associated Cdc6 homologs are highly conserved. In addition, gene order analysis found that genes around oriC1 are highly syntenic among haloarchaea (Figure 2; Capes et al., 2011). Notably, other studies (Robinson et al., 2004; Coker et al., 2009) and our results indicated that the oriC1 replication origin is broadly conserved in archaea, in terms of both function and structure, which strongly suggested that the ancestral chromosome was dependent on oriC1. Variations were observed in oriC1 homologs from different archaeal phyla, which may contribute to the adaptability of archaea to different extreme environments. For example, an extended halophile-specific “G-string” element has been identified at the end of each ORB in haloarchaea, and these “G-string” elements have been proven to be essential for autonomous replication based on the oriC1 in Haloarcula hispanica (Wu et al., 2014).

Multiple replication origins along with their adjacent cdc6 genes appear to be mosaics of distinct replicator–initiator systems. A comparison between Aeropyrum and Sulfolobus origins suggested that the capture of extrachromosomal elements accounts for replicon evolution (Robinson and Bell, 2007). In particular, it has been proposed that the three replication origins of the Sulfolobus species arose by the integration of extrachromosomal elements into a single-origin ancestral chromosome (oriC1-cdc6-1), and the acquisition of oriC3-whiP occurred prior to the integration of oriC2-cdc6-3 (Samson et al., 2013). Similarly, genomic context analyses of ori-cdc6 systems in haloarchaea revealed that 40% of predicted replication origins were observed with transposases or integrases nearby, indicative of the translocation of a subset of replication origins among haloarchaea. In addition, comparative analyses of the selected replication origins suggested that different evolutionary mechanisms, including ancestral conservation and coupled acquisition and deletion events, may account for the current mosaics of multiple replication origins in the haloarchaeal genomes. Importantly, a comparative genomic analysis of two Haloarcula species, Haloarcula hispanica and Haloarcula marismortui, revealed that the species-specific origins are located in extremely variable regions, suggesting that these novel origins were recently acquired, via either integration into the chromosome or rearrangement of extrachromosomal elements (Wu et al., 2012). Further work may focus on comparisons of replication origins from closely related species to reveal the dynamics of origin evolution and whether origin evolution alters the mode of genomic replication.

To date, the number of archaea with mapped replication origins is still limited, which to some extent has affected us to get a panoramic view of the generality and evolution of replication origins in archaea. In addition to the mapping of replication origins, the development of prediction algorithms for replication origins in archaeal genomes and the construction of databases with these predicted origins (Gao et al., 2013) will be useful for comparing replication origins from a broader range of archaeal species. Fortunately, the rapid increase in the number of complete archaeal genomic sequences that are publically available will promote our studies of archaeal replication origins.

In addition, the control and coordination of replication initiation at multiple origins in archaea is far less understood. The multireplicon structure of haloarchaeal genomes allows for precise control and coordination of replication initiation at multiple origins. As the chromosome and extrachromosomal elements within a haloarchaeon are generally different sizes and have different copy numbers (Breuert et al., 2006; Liu et al., 2013), it will be interesting to reveal whether they initiate synchronously and how they maintain different copy numbers, as well as what roles multiple replication origins play in governing polyploidy in haloarchaea. In addition, the coordination of multiple origins may play important roles in maintaining the multireplicon structure of haloarchaeal genomes. As most replication origins are dependent on Cdc6 proteins in haloarchaea (excluding the origins of small plasmids), we propose that the coordination of replication initiation at different origins may be obtained by conformational changes of Cdc6 proteins via an ATP-ADP binary switch, which has recently been proposed for chromosome replication in S. islandicus (Samson et al., 2013). Thus, more exhaustive work should be taken into account to uncover the control and coordination of the replication initiation from multiple origins, either on the same chromosome or from different genetic elements, in haloarchaeal multireplicon genomes.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.