In order for any organism to replicate its DNA, a helicase must unwind the duplex DNA in front of the replication fork. In archaea, the replicative helicase is the MCM (minichromosome maintenance) helicase. Although much is known about the biochemical properties of the MCM helicase, the mechanism of assembly at the origin of replication is unknown. In the present paper, several possible mechanisms for the loading process are described.
The MCM (minichromosome maintenance) complex is the replicative helicase in archaea. Several archaeal MCM helicases have been studied, and their biochemical properties have been described. The helicases from most archaea form hexamers in solution, have an ATP-dependent 3′→5′ helicase activity, bind and translocate along ssDNA (single-stranded DNA) and dsDNA (double-stranded DNA), can unwind DNA–RNA hybrids while moving along the DNA strand, and can displace proteins from DNA (reviewed in [1–4]).
High-resolution structural information is not yet available for a full-length MCM helicase, but has been determined for the N-terminal non-catalytic domains of two MCM proteins, from Methanothermobacter thermautotrophicus  and Sulfolobus solfataricus . These have a three-domain structure. Biochemical studies suggest that domain A plays a role in regulation [7–9], and domain B participates in DNA binding via a zinc-finger motif . Domain C, which connects the N-terminal portion of the protein to the C-terminal catalytic region, is involved in protein multimerization [6,7], DNA binding via a β-hairpin motif [5,11,12] and the communication of signals between the N-terminal and C-terminal portions of the MCM protein . Only low-resolution electron micrograph reconstruction studies have been performed on the full-length MCM protein from the archaeon M. thermautotrophicus. These studies suggest that the protein can adopt multiple forms; hexamers, heptamers, dodecamers, filaments and open circles were identified (see  and references therein). Conformational changes upon DNA and nucleotide binding have also been observed .
Replicative helicases in bacteria and viruses require a loader protein or complex to assemble the helicase on to the origin of replication. In Escherichia coli, the DnaC protein functions as the helicase loader for the DnaB helicase. It is currently unknown whether MCM helicases have loader proteins. Since the archaeal replication machinery possesses some of the characteristics of bacteria, it is possible that an archaeal loader exists. One candidate protein for the archaeal helicase loader is the Cdc6 (cell-division cycle 6) protein. This is based on primary amino acid sequence similarity to the eukaryotic initiator protein Cdc6. In addition, the enzymes were shown to interact with MCM , regulate MCM helicase activity [17–19] and affect the oligomeric state of the helicase .
Formation of the replication bubble
Replication begins at origins of replication, which are A/T-rich and contain one or more A/T-rich stretches, known as DUEs (duplex unwinding elements), which are essential for origin function. OBPs (origin-binding proteins) bind ORBs (origin-recognition boxes) and cause localized unwinding. All archaeal origins identified contain multiple short inverted repeats, and some also contain two long inverted repeats bordering the origin region. The short inverted repeats in different archaeal species share sequence similarity (see  and references therein) and serve as ORBs to which the OBPs bind. The archaeal Cdc6 proteins were shown to function in origin recognition (in addition to their putative role as helicase loaders). It was shown that when the archaeal Cdc6 protein (OBP) binds to the inverted repeat at the origin it causes a distortion of the duplex DNA [22,23].
There are at least two mechanisms by which the replication bubble can form (Figure 1). Multiple OBPs binding to multiple ORBs might result in sufficient deformation to result in bubble formation (Figure 1A). Support for this hypothesis comes from E. coli, where the DnaA protein (the OBP) binds to the DnaA boxes (ORB) and results in replication bubble formation. Alternatively, under supercoiling conditions, such as those found in vivo, the inverted repeat sequences may form cruciform structures with stems formed from the inverted repeats and single-stranded loops composed of intervening sequences . The binding of the OBP could stabilize the stems (Figure 1B).
Possible mechanisms of replication bubble formation
Possible mechanisms of archaeal MCM helicase loading at the origin
Helicase assembly in archaea may or may not use a loader protein; investigation into this is ongoing. Several possibilities for helicase loading are presented below.
Helicase loader-dependent assembly
If archaea employ a loader protein, as do bacteria, helicase assembly at the origin could occur as it does in bacteria. Much is known about the mechanism of helicase loading at the E. coli origin of replication (oriC). The DnaC protein functions as a helicase loader, assembling the DnaB helicase at the origin in an ATP-dependent reaction. The protein forms a complex with the hexameric ring of DnaB, resulting in a DnaB6–DnaC6 complex. The complex then interacts with origin-bound DnaA. When the DnaA protein binds to the origin it causes a distortion in the DNA and forms the replication bubble. After binding the DnaA–DNA complex, the DnaC protein opens the DnaB hexameric ring and assembles it on to the replication bubble. Thus the DnaC protein functions as a ‘ring breaker’ (Figure 2A) .
Possible mechanisms of loader-dependent helicase assembly
Biochemical studies with a number of archaeal Cdc6 proteins show that they have biochemical properties that are similar to those of the E. coli DnaC loader. Both proteins interact with the helicase and inhibit helicase activity in vitro and both bind and hydrolyse ATP. These similarities led to the hypothesis that the archaeal Cdc6 may be the functional homologue of the E. coli DnaC protein and function as a ring breaker (Figure 2A) .
The archaeal replication machinery may be a modified version of the E. coli ring breaker model that involves dissociation of the helicase (Figure 2B). Studies with the Cdc6 and MCM proteins from the archaeon M. thermautotrophicus showed that interaction between the Cdc6 and MCM proteins dissociates the helicase, and no large complexes, which may include hexamers of MCM that associate with Cdc6 protein, could be observed . These observations are different from those made with the E. coli proteins in which a complex between hexamers of the DnaC and DnaB proteins can easily be detected. Thus it was suggested that the Cdc6 protein may still have a MCM loading function, but, instead of opening the ring at one interface, within the hexameric structure, it dissociates the MCM ring before assembly. This would make Cdc6 function as a ‘ring dissociater’  (Figure 2B).
A potential mechanism could be to assemble monomeric subunits on to the origin to form the active helicase, referred to as a ‘ring maker’ . However, as most archaeal MCM proteins are stable hexamers or other multimers, this mechanism is unlikely.
Helicase loader-independent assembly
If the initial replication bubble (Figure 1) is sufficiently large, the helicase may be able to assemble itself on it without the need for a loader (Figure 3). There are several observations that may support this hypothesis. The archaeal MCM complex was shown to assemble around circular closed plasmid ssDNA (see  and references therein). The E. coli DnaB helicase, on the other hand, cannot assemble around such a substrate (and therefore requires a helicase loader). Also, the size of the single-stranded loop required for MCM assembly is approx. 50 bases [26a]. A loop of this size may be available between the long inverted repeats in several archaeal origins (for example, in Pyrococcus abyssi, the distance is 240 bases , and in S. solfataricus, the distance is approx. 65 bases ). Therefore bubble formation of sufficient size could occur via the mechanisms shown in Figure 1.
Helicase loader-independent assembly
The MCM helicase may open to load on to DNA
The hexameric ring can assemble itself through complex ‘breathing’, resulting in temporary opening of the ring allowing assembly on to the DNA (Figure 3A). The interaction with DNA may stabilize the ring and prevent dissociation.
The MCM helicase may thread itself on to the initial replication bubble
Even if the initial bubble is too small for hexamer assembly, the MCM might still thread itself on to it. The electron micrograph reconstruction studies with the M. thermautotrophicus MCM complex revealed that the protein can form open circles and filaments [14,29]. Both of these helix-like structures have an opening in the ring. It was therefore suggested that this could facilitate self-loading of the helicase . This loading could take place even on a small replication bubble created by the OBPs if the helical MCM protein can wrap itself around the DNA strand. The open ring could act as a ‘corkscrew’ which would melt the DNA and form the bubble as it wound on to the DNA. Structural studies show that, when Cdc6 proteins bind the origin, they form a distortion in the duplex [22,23]. Although there are multiple ORBs in most archaeal origins, the distortion formed by them may not result in a bubble sufficiently large for helicase assembly; only a small ssDNA region may be exposed. If the MCM complex could melt the DNA using a corkscrew motion while it threads on, the small bubble formed may be sufficient for helicase self-assembly (Figure 3B).
The oligomeric state of the MCM helicase may change during loading
Electron micrograph reconstruction studies with the M. thermautotrophicus MCM have shown that the enzyme can form hexamers and heptamers [14,30,31]. It was suggested that DNA binding by the helicase might result in a switch from heptamer to hexamer and facilitate the assembly of the active helicase . It is also possible that, during assembly at the origin, the helicase loader removes one of the protomers from the heptamer and closes the hexamer around the DNA (Figure 3C) .
What is the role of duplex DNA translocation by the MCM helicase?
All MCM proteins studied bind and translocate along duplex DNA. It is not yet clear whether, during replication, the helicase moves on ssDNA or dsDNA, and models for both possibilities have been proposed . It has also not been established whether the helicase assembles at the origin before origin melting by the OBPs, and thus may be loaded on dsDNA, or after origin melting, and thus may be loaded on either ssDNA or dsDNA.
It is possible, however, that duplex translocation plays an indirect role in helicase loading by speeding the process by which the helicase locates the origin. Most of the genes encoding archaeal OBPs are located in close proximity to the origin of replication. It has been suggested that this proximity enables the proteins to bind to the origin as soon as they are synthesized . However, the genes encoding the MCM proteins are located in different locations along the chromosome, and not always near the origin. The helicase therefore has to travel to find the origin, either by itself or in combination with the helicase loader. As diffusion in one dimension is more efficient than diffusion in three dimensions, one possibility is that the helicase binds to the duplex DNA of the chromosome and translocates along the duplex until it reaches the origin. At the origin, the helicase may either encounter the replication bubble and self-load, as described above, or the helicase may encounter the helicase loader associated with the origin.
Several biochemical properties of MCM support this hypothesis. The MCM proteins from several archaea can associate with circular dsDNA and translocate along the duplex [33,34]. It was also shown that the MCM protein can displace histone and other proteins from DNA, suggesting that translocation can occur in vivo .
Do all archaea utilize the same loading mechanism?
As described above, the mechanism of helicase assembly at the origin of replication is not yet known, although several mechanisms have been proposed. It is also not clear whether all archaeal species use the same loading mechanism or whether each organism, or group, developed their own assembly process. Several lines of evidence suggest that more than one loading mechanism may exist. (i) Not all archaeal species contain a clear homologue of Cdc6, and some species contain only one homologue. In species lacking a Cdc6 homologue, the loader may not exist (loader-independent assembly) or may not have been identified. In some species, the Cdc6 protein may function both in origin recognition and as the helicase loader. (ii) Not all archaeal MCM complexes behave the same. Different forms in solution have been observed (for examples, see [33,36]) and different activities have been noted in different species. For example, some enzymes are active on their own, but some require additional factors such as Cdc6  or the GINS complex . (iii) Some archaea contain multiple MCM homologues. If these form hetero-oligomers, the loading mechanism may be different from that of a homomultimer. (iv) The relative abundance of MCM helicase within the cell appears to differ between different organisms. Studies have shown that P. abyssi contains approx. 400 molecules of MCM in rapidly dividing cells , M. thermautotrophicus has approx. 300 molecules (N. Sakakibara and Z. Kelman, unpublished work), but Thermoplasma acidophilum contains approx. 1000 molecules per cell . The abundance of the MCM and other initiation proteins (e.g. Cdc6) may influence helicase assembly. (v) The Cdc6 proteins from different archaea have different effects on helicase activity in vitro (for examples, see [17,18]). These differences could possibly be a reflection of different loading mechanisms.
Support for different loading processes in different archaea comes from studies with bacteria. It was shown that, although the E. coli helicase is loaded at the origin via a ‘ring-breaker’ process, the helicase from Bacillus subtilis is loaded via a ‘ring-maker’ mechanism . It is therefore possible that different archaea may also use different loading mechanisms.
Molecular Biology of Archaea: Biochemical Society Focused Meeting held at University of St Andrews, U.K., 19–21 August 2008. Organized and Edited by Stephen Bell (Oxford, U.K.) and Malcolm White (St Andrews, U.K.).
We apologize to colleagues whose primary work was not cited due to space limitations.
This work was supported by a grant from the National Science Foundation [grant number MCB-0815646] and a Research Scholar Grant from the American Cancer Society [grant number RSG-04-050-01-GMC] awarded to Z.K.