Biochemical Characterization of the Methanothermobacter thermautotrophicus Minichromosome Maintenance (MCM) Helicase N-terminal Domains*

Minichromosome maintenance helicases are ring-shaped complexes that play an essential role in archaeal and eukaryal DNA replication by separating the two strands of chromosomal DNA to provide the single-stranded substrate for the replicative polymerases. For the archaeal protein it was shown that the N-terminal portion of the protein, which is composed of domains A, B, and C, is involved in multimer formation and single-stranded DNA binding and may also play a role in regulating the helicase activity. Here, a detailed biochemical characterization of the N-terminal region of the Methanothermobacter thermautotrophicus minichromosome maintenance helicase is described. Using biochemical and biophysical analyses it is shown that domain C of the N-terminal portion, located adjacent to the helicase catalytic domains, is required for protein multimerization and that domain B is the main contact region with single-stranded DNA. It is also shown that although oligomerization is not essential for single-stranded DNA binding and ATPase activity, the presence of domain C is essential for helicase activity.

The structure of the archaeal MCM complex is unclear. The MCM homologues of S. solfataricus (16) and A. fulgidus (18) form hexamers in solution, but mtMCM appears to form dodecamers (10 -13). The crystal structure of the N-terminal portion of the mtMCM protein suggested a dodecameric structure (19). However, electron microscope reconstructions of fulllength mtMCM revealed hexameric (20) or heptameric (21) structures, but not larger multimers.
The mtMCM, and probably other archaeal MCM enzymes, consists of two main portions. The N-terminal region participates in protein multimerization and ssDNA binding and may also have a regulatory function. The C-terminal portion of the protein contains the helicase catalytic domain(s) (11,19). A high resolution three-dimensional structure of the N-terminal portion revealed a dumbbell-shaped double hexamer. Each monomer folds into three distinct domains (Fig. 1A). Domain A, at the N terminus, is mostly ␣-helical. Domain B has three ␤-strands and contains a zinc finger motif. Domain C, positioned between domains A and B (Fig. 1A), contains five ␤-strands that form an oligonucleotide/oligosaccharide binding (OB) fold. This domain connects the N-terminal portion of the enzyme to the catalytic region.
The domain(s) within the N-terminal region responsible for multimerization has not yet been identified. The three-dimensional structure suggested that the zinc finger motif plays a role in double hexamer formation (19). However, biochemical analysis of a zinc finger mutant showed that the mutant protein is impaired in ssDNA binding but not double hexamer formation (13). In addition, studies with the A. fulgidus MCM suggested that domain B, which contains the zinc finger, is not needed for multimerization (18) but probably for ssDNA binding.
A study was therefore initiated to determine the functional roles of each of the three N-terminal domains in the mtMCM protein. In this study it is shown that domain A does not play an essential role in MCM function, because helicase, ATPase, and ssDNA binding activities could be observed in mutant proteins in which the domain had been removed. However, the domain is needed for dsDNA translocation and may play a regulatory role. Domain B is the major contact with ssDNA, and domain C is necessary and sufficient for MCM multimerization and is essential for helicase activity.

Materials
ATP and [␥-32 P]ATP were obtained from Amersham Biosciences, ss⌽X174 DNA was obtained from New England Biolabs, and oligonucleotides were synthesized by the Center for Advanced Research in Biotechnology DNA synthesis facility. The various MCM proteins were purified as previously described (10).

Methods
Generation of MCM Mutants-MCM mutants were generated using a PCR-based approach from plasmid containing the gene encoding the wild-type mtMCM enzyme (12). The three-dimensional structure of the N-terminal portion of the molecule (19) served as the guide for construction of the various mutant proteins. The oligonucleotides and templates used are shown in Supplementary Fig. 1. To generate a mutant protein in which the N-or C-terminal portion was deleted (N-ter, Hel, ⌬A, ⌬AB, AB, AC, BC, A, B, and C), a simple PCR reaction was used. For generating the mutant proteins involving deletions of domains within the polypeptide chain (⌬B, ⌬C, ⌬BC, ⌬AC), a two-step PCR strategy was adopted (schematically described in Supplementary Fig. 1B). The first PCR reaction amplified the smaller fragments of the mutants using their respective forward and reverse primers. The reverse primers designed for these PCRs had a 16 -22-bp sequence complementary to the forward primer of the succeeding domain/fragment at the 5Ј end. The products of these PCR reactions were then mixed at equimolar concentration, and a second PCR reaction was performed using the forward primer of the upstream fragment (which has a SalI restriction site) and the reverse primer of the downstream fragment (which has an AatII restriction site). This resulted in a product with one or two deleted domains within the wild-type gene that was cloned into pDBLeu and pPC86 vectors (Invitrogen) between the SalI and AatII sites, yielding fusion proteins to the GAL4 DNA binding or activation domains, respectively. These constructs were used in a two-hybrid analysis.
To generate the Escherichia coli expression vectors containing the various mtMCM mutants, the pDBLeu vectors with the different truncated proteins were used as templates for PCR with primers containing NdeI (in the forward primer) and XhoI (in the reverse primer) restriction sites ( Supplementary Fig. 1). The reverse primer also adds six His residues upstream of the stop codon. The PCR products were cloned into the pET-21a vector (Novagene). All proteins used in the study contain the catalytic domain (except for the N-terminal portion and domain C), and all have a His 6 tag at the C terminus.
Two-hybrid Analysis-For the two-hybrid analysis, pDBLeu and pPC86 vectors containing the various MCM mutant genes were generated (see above). Plasmids encoding the activation and DNA binding domain fusion proteins were co-transformed into yeast MaV203 cells (Invitrogen) according to the manufacturer's protocol. Cells were plated on complete supplement mixture (CSM) plates without Leu and Trp and grown for 2-3 days at 30°C. Colonies were streaked on CSM plates without Leu, Trp, and His and containing 10 mM 3-amino-1,2,4-triazole to suppress glycerol phosphate dehydratase, an enzyme involved in histidine biosynthesis. Colonies were also streaked on CSM plates without Leu, Trp, and Ura. Plates were incubated at 30°C for 2-3 days. Growth on these plates indicates that the proteins fused to the activation and DNA binding domains interact.
Gel Filtration Analysis-One hundred fifty micrograms of the various proteins were applied to a superose-6 (HR10/30; Amersham Biosciences) or superdex-200 (HR10/30; Amersham Biosciences) gel filtration column equilibrated with buffer containing 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 0.5 mM EDTA, and 10% glycerol. Columns were run at 22°C. Fractions (250 l) were collected and analyzed for the presence of the mtMCM proteins by separation on a 10% SDS-PAGE (except for domain C, which was separated on a 15% SDS-PAGE) and staining with Coomassie Brilliant Blue (R-250).
Glycerol Gradient Sedimentation-Glycerol gradient centrifugation was performed by applying 100 g of domain C protein to a 5-ml 20 -40% glycerol gradient in buffer containing 20 mM Tris (pH 7.5), 100 mM NaCl, and 0.5 mM EDTA. After centrifugation at 45,000 rpm (190,000 ϫ g) for 18 h in a Beckman SW 50.1 rotor at 4°C, fractions (200 l) were collected from the bottom of the tube. The distribution of the protein was determined by fractionation on a 12% SDS-PAGE and staining with Coomassie Brilliant Blue (R-250).
ATPase Assay-ATPase activity was measured in reaction mixtures (15 l) containing 25 mM Hepes-NaOH (pH 7.5), 5 mM MgCl 2 , 1 mM dithiothreitol, 100 g/ml bovine serum albumin, 1.5 nmol of ATP containing 2.5 Ci of [␥-32 P]ATP (3000 Ci/mmol; Amersham Biosciences), and 0.5 or 1.5 pmol of proteins (as monomers) in the presence or absence of 50 ng of ss⌽X174 DNA. After incubation at 60°C for 60 min, an aliquot (1 l) was spotted onto a polyethyleneimine cellulose thin layer plate, and ATP and P i were separated by chromatography in 1 M formic acid ϩ 0.5 M LiCl. The extent of ATP hydrolysis was quantitated by PhosphorImager (Amersham Biosciences) analysis.
DNA helicase activity was measured in reaction mixtures (15 l) containing 20 mM Tris-HCl (pH 8.5), 10 mM MgCl 2 , 2 mM dithiothreitol, 100 g/ml bovine serum albumin, 5 mM ATP, 10 fmol of 32 P-labeled DNA substrate (3,000 cpm/fmol), and the various MCM mutant proteins as indicated in Fig. 6. After incubation at 60°C for 1 h, reactions were stopped by adding 5 l of 5ϫ loading buffer (100 mM EDTA, 1% SDS, 0.1% xylene cyanol, 0.1% bromphenol blue, and 50% glycerol). Aliquots were loaded onto an 8% native polyacrylamide gel in 0.5ϫ TBE and electrophoresed for 1.5 h at 200 V. The helicase activity was visualized and quantitated by phosphorimaging. All helicase experiments were repeated three times with almost identical results. Representative gels are shown in Fig. 6.
Streptavidin Displacement Assay-The streptavidin displacement assay was performed as previously described (15) using a 32 P-labeled 50-mer biotinylated oligonucleotide, 5Ј-GGGACGCGTCGGCCTGGCACG(Biotin-dT)CGGCCGCTGCGGCCAGGCACCCGATGGC-3Ј. Ten fmol of the oligonucleotide were incubated with 50 nM streptavidin in a helicase reaction mixture (15 l) at 30°C for 10 min, followed by the addition of 50 M free biotin (to trap and sequester streptavidin released during the helicase reaction) and varying amounts of protein. After incubation for 1 h at 60°C, 5 l of 5ϫ loading buffer was added to each sample, followed by electrophoretic analysis on a native 8% polyacrylamide gel in 0.5ϫ TBE. The gels were analyzed using phosphorimaging.
Multiple Alignment and Surface Conservation-The multiple alignment of the N-terminal portion of the archaeal MCM proteins was constructed by searching the sequence of MCM protein chain A (Protein Data Bank accession number 1ltl) against the National Center for Biotechnology Information (NCBI) non-redundant protein data base using the NCBI Blast PSI-BLAST program. After five rounds of PSI-BLAST, sequence relatives with expectation scores Ͻ0.005 from the archaea domain were pooled and aligned using the ClustalW program.
The surface conservation shown in Fig. 7 was made by scoring the relative conservation of a given column compared with all other columns in the multiple alignments. The raw score for each column was calculated using standard Shannon's information theory entropy formula (23) and converted into normalized Z-score distribution by calculating average entropy and S.D. for all columns in the alignment. The molecular surfaces and surface potential were built using the Grasp program.

Domain C of the mtMCM N-terminal Region Is Required for
Multimer Formation-As a first step in determining the region(s) of the mtMCM needed for multimerization, a two-hybrid analysis was performed. Previous studies have shown that the N-terminal portion of the protein participates in hexamer/ double hexamer formation (11,19), and a high resolution threedimensional structure of the mtMCM N-terminal portion revealed a three-domain (A, B, and C) structure (Fig. 1A) (19). Using the three-dimensional structure as a guide, a number of constructs were made to express various deletion mutants of the N-terminal domains. These include the expression of the individual domains and the deletion of single or multiple domains. The genes encoding the deletion proteins were gener-ated using a PCR-based approach (see "Experimental Procedures") and cloned into the pDBLeu and pPC86 vectors (Invitrogen), resulting in fusion proteins with the GAL4 DNA binding and activation domains, respectively. The various constructs were analyzed for their ability to interact with the full-length protein, the N-terminal portion of the molecule, and for self-interaction (Fig. 1B). As shown in Fig. 1B, each protein containing domain C showed interaction with itself, the fulllength, and the N-terminal portion of the mtMCM protein.
Furthermore, domain C is the only intact domain that demonstrated self-interaction.
To confirm the observation made with the two-hybrid analysis, all deletion proteins containing the catalytic domain, the N-terminal region, and domain C alone were expressed and purified from E. coli and analyzed on a superose-6 gel filtration column. As shown in Fig. 2, truncated proteins containing  F, H, and I).
Several of the mutant proteins that contain domain C, however, are less stable than the full-length protein as is evident from the presence of two peaks (dodecamers and hexamers) or by the "trailing" of the proteins, suggesting that the complex dissociated during fractionation on the sizing column (e.g. panel G). These observations suggest that either additional domains, besides domain C, are involved in double hexamer formation or the deletion of the other domains may affect the overall stability of the mtMCM complex.
The data presented in Figs. 1 and 2 suggest that domain C is responsible for multimerization of the mtMCM. The superose-6 column, however, is not the most suitable one for the analysis of domain C alone because the domain is only 17.9 kDa in size. Therefore, a superdex-200 column was used to analyze this domain (Fig. 3A). This analysis revealed the presence of a major peak at fractions 58 -60. To get a more accurate size determination, a glycerol gradient sedimentation was also performed (Fig. 3B). In the glycerol gradient a major peak was observed at fractions 11 and 12. Combining the S value and Stokes radius in the mass equation (24) yields a native mass of 103 kDa for the major peak on gel filtration (Fig. 3A). This is in good agreement with the expected size of a hexameric structure (107.8 kDa). However, a minor peak at fractions 46 -48 was also observed following the gel filtration (Fig. 3A). This elution position is consistent with a dodecameric structure (216 kDa). In addition, the glycerol gradient sedimentation shows a trailing of the proteins from the dodecameric to hexameric structures. These results may suggest that although domain C alone can form double hexamers, they are not stable and readily dissociate to hexamers during the gel filtration and glycerol gradient analyses.
Domain B of the N-terminal Portion of mtMCM Is Needed for Efficient ssDNA Binding-During DNA unwinding, the helicase translocates along one strand of the duplex and displaces the complementary strand. Hence, all DNA helicases, including mtMCM, have been shown to interact with ssDNA (25). Previous studies demonstrated that the zinc finger motif is needed for ssDNA binding by mtMCM (13) as well as by the eukaryotic helicase (26). The results of these studies, however, suggested that other regions of the proteins are also needed for efficient DNA binding. Thus, the ability of the various mutant proteins to interact with ssDNA was determined using a filter binding assay (Fig. 4A) and gel mobility shift assay (Fig. 4B).
As shown in Fig. 4, all truncated proteins show substantially reduced levels of ssDNA binding in comparison to the fulllength enzyme. Deletion of domain B has the most severe effect on ssDNA binding, because no ssDNA binding could be detected in a truncated protein in which only domain B was deleted (Fig. 4A, ⌬B). This observation suggests that domain B is the major (but probably not the only) region of the mtMCM that contacts ssDNA. The zinc finger motif is located in domain B, and thus these results are consistent with past reports illustrating the need for an intact zinc finger for efficient DNA binding (13,26). Interestingly, deletion of domain B in conjunction with another domain (either A or C) resulted in a protein with better ssDNA binding in comparison to a protein in which only domain B was removed. It is possible that removing such a large part of the protein exposed charged regions that may show affinity to the negatively charged DNA. Nevertheless, all these mutant proteins show a substantial reduction in ssDNA binding in comparison to the wild-type enzyme.
On the other hand, a mutant protein lacking domain C that could not form hexamers (Fig. 2F) retained about 40% of ssDNA binding in comparison to the wild-type enzyme (Fig.  4A). These results are consistent with previously reported observations (10, 11) that show that monomeric mtMCM protein retains detectable ssDNA binding. This indicates that hexamer/dodecamer formation is not needed for ssDNA binding.
The ability of the truncated proteins to interact with ssDNA was also determined using a gel mobility shift assay (Fig. 4B). The results are similar to those observed with the filter binding assay. Domain B is most critical for ssDNA binding because proteins lacking the domain (⌬B, ⌬BC, and ⌬AB) bind poorly to ssDNA (Fig. 4B, lanes 8, 9, 12, 13, 16, and 17) in comparison to the full-length or truncated proteins containing domain B. Proteins deleted for domain C (⌬C, ⌬AC) retained ssDNA binding ( lanes 10, 11, 14 and 15). Interestingly, these proteins resulted in slower migrating bands in comparison to the proteins that include domain C (full-length, N-terminal portion and ⌬A, lanes 2-7). This is likely because of the different structure of the proteins. The proteins without domain C (⌬C, ⌬AC) are monomeric and thus are not expected to result in similar shifts to those created by the hexameric/dodecameric proteins. In agreement with previous observations (13), the full-length protein formed a faster and a slower migrating band (lanes 2 and 3). Both Simian Virus 40 Large-T antigen and Schizosaccharomyces pombe MCM helicases were also shown to form a faster migrating band that was shown to be a result of a hexamer binding to the DNA and a slower migrating band that was a result of a double hexamer binding (27,28). The same explanation has been proposed for the mtMCM protein in which the faster migrating band contains only one hexamer, whereas the slower one contains a double hexamer (13).
mtMCM ATPase activity was shown to be stimulated in the presence of ssDNA (10 -13). Thus, another indirect approach to demonstrate the domains required for ssDNA binding involves performing ATPase assays in the presence and absence of DNA. Therefore, the ATPase activity of the various mtMCM mutant proteins and stimulation of that activity by ssDNA was examined. As shown in Fig. 5, and similar to previously reported observations (10 -13), the ATPase activity of the fulllength enzyme was stimulated ϳ4-fold in the presence of ssDNA. Only the truncated protein lacking domain A (⌬A) shows a similar level of stimulation of activity by ssDNA, although the ATPase activity of the mutant protein is lower than the wild-type enzyme. None of the other truncated proteins showed stimulation of ATPase activity by DNA. Also, whereas the protein lacking domain A contained about 30% of the ATPase activity compared with the full-length enzyme, the ATPase activities of all the other truncated proteins were substantially reduced.
mtMCM Hexamerization Is Required for Helicase Activity-Next, the requirement for the three N-terminal domains (A, B, and C) for helicase activity was examined. Based on the data presented above (Figs. 4 and 5), it was anticipated that most of the mutants would have very low, if any, helicase activity. Therefore, the helicase and dsDNA translocation experiments were performed using a higher protein concentration (10 -50fold) than that required to detect helicase activity with the full-length enzyme (14,15).
As shown in Fig. 6A, only the intact protein has an appreciable helicase activity, because 1 and 3 pmol of enzyme (as monomer) displaced 25 and 66% of the substrate, respectively (lanes 3 and 4). The only truncated proteins with detectable helicase activity are those missing domain A (⌬A, lanes 7 and 8) or B (⌬B, lanes 9 and 10). Because these proteins are hexameric/dodecameric (Fig. 2, D and E), the results demonstrate the need for ring formation by the proteins for helicase activity. The helicase activity of these mutant proteins, however, is substantially less than the full-length enzyme. The deletion of domain C or deletion of any two domains completely abolished helicase activity. As expected, the N-terminal domain alone showed no helicase activity (Fig. 6A, lanes 5 and 6). The experiments shown in Fig. 6A were performed with flat substrate. Similar results were observed using a forked DNA substrate (data not shown).
The reduced helicase activity of the truncated proteins lacking domain B may be attributed to a reduced affinity for ssDNA. It was previously shown that a zinc finger mutant is impaired in ssDNA binding (13), and domain B deletion results in a similar reduction in ssDNA binding (Fig. 4). Indeed, the helicase activity of the protein lacking domain B is similar to that observed with a zinc finger mutant (Fig. 6A, compare lanes  9 and 10 to lanes 19 and 20), supporting the idea that the reduction in activity is, in part, because of impaired ability to interact with DNA.
It was recently demonstrated that the M. thermautotrophicus and eukaryotic MCM complexes are capable of translocating along duplex DNA (15,29). Thus, the mutant proteins that show some level of helicase activity (Fig. 6A, ⌬A and ⌬B) were tested for their ability to move along duplex DNA. As shown in Fig. 6B, neither of the truncated proteins is capable of moving along the duplex (lanes 5-8). At similar protein concentrations, however, the full-length enzyme results in efficient translocation along the duplex (lanes 3 and 4). These observations further suggest that the removal of any domain from the Nterminal portion of mtMCM impairs its ability to translocate along DNA. Additional evidence of the impaired ability of the mutant proteins in efficient DNA translocation comes from the observation that the mutant proteins cannot displace streptavidin from biotinylated oligonucleotides, whereas the fulllength enzyme can (Ref. 15; data not shown). DISCUSSION The biochemical analysis of the truncated mtMCM proteins described in this study, together with past observations, revealed that each of the domains in the N-terminal portion of the molecule has a different role in MCM function. Domain A may play a regulatory role, Domain C is necessary and sufficient for protein multimerization, and Domain B is the main contact with ssDNA.
MCM Hexamer Formation-To date, the structures of a number of archaeal MCM complexes have been determined. Although all proteins are homologous, they appear to aggregate differently. Although the mtMCM appears to form double hexamers (10 -13), the enzymes from S. solfataricus (16), A. fulgidus, and Aeropyrum pernix (18) are hexamers. Electron micrograph reconstruction of mtMCM revealed ring-shaped hexamers (20) or heptamers (21), but no dodecamers could be observed. The observation that the M. thermautotrophicus enzyme forms only single ring structures at low concentrations (such as those used for electron microscopy) may suggest that the double hexamers are formed by nonspecific hydrophobic or ionic interactions involving the two hexamers. These nonspecific interactions may be present only in the mtMCM enzyme, because to date this is the only MCM complex of either archaea or eukarya in which dodecameric structures have been reported. In eukarya, however, the Mcm4,6,7 complex was shown to form double hexamers in the presence of a forked DNA substrate (28).
The observation that domain C is involved in hexamer and double hexamer formation is consistent with a previously published observation showing that a truncated mtMCM protein, in which the first 111 amino acids have been removed, is monomeric (11). Domain C starts at amino acid 92 (19), and thus the N-terminal portion of domain C is deleted in the mutant protein. Taken together, these observations suggest that the N-terminal part of domain C (located N-terminal to domain B in the primary amino acid sequence) may be required for multimerization. The deletion, however, is not likely to affect the overall structure of the molecule because, similar to FIG. 4. mtMCM multimerization is not required for ssDNA binding. A, filter binding assays were performed as described under "Experimental Procedures" using 32 P-labeled N120T oligonucleotide in the presence of 0.1, 0.3, 0.9, and 2.7 pmol of protein (as monomer). The average result of three experiments is shown. B, gel mobility shift assays were performed as described under "Experimental Procedures" using 32 P-labeled N120T oligonucleotide and 1 and 3 pmol of proteins (as monomer). Lane 1, substrate only. the data described here (Fig. 4), the deleted mutant, although monomeric, retains its ability to bind DNA (11).
MCM homologues from other archaeons may also multimerize via domain C. The region within domain C needed for multimerization, however, may be different from that in mtMCM. Removal of the N-terminal part of domain C and the entire domain B from the A. fulgidus and A. pernix MCMs does not affect the hexameric structure of the proteins (18), suggesting that the C-terminal portion of domain C, adjacent to the catalytic domain, is needed for hexamer formation. Thus, mtMCM may be different from these enzymes not only by forming dodecamers but also in the region involved in multimer formation.
MCM Interactions with ssDNA-The biochemical analysis of the truncated MCM proteins described here and elsewhere (13) demonstrates that domain B is the main contact with ssDNA, probably via the zinc finger fold located within the domain. It is possible, however, that domain C also participates in ssDNA binding because it contains an OB fold. Interestingly, the three-dimensional structure of domain C is superimposable on the OB2 domain of BRCA2, and domain B is located in a position similar to the Tower domain of the BRCA2 molecule (30). DNA binding was shown to occur in the cleft between the OB2 and Tower domains. It is possible that in mtMCM protein DNA interactions occur in the cleft between domains B and C in addition to the interactions of ssDNA with the zinc finger located in domain B. FIG. 6. mtMCM multimerization is needed for helicase activity and dsDNA translocation. A, DNA helicase assays of the various mutant proteins were performed as described under "Experimental Procedures" using 10 fmol of substrate and 1 and 3 pmol of proteins (as monomer). The percent displacement of the 32 P-labeled oligonucleotide from the duplex DNA substrate is indicated as %. Lane 1, substrate only; lane 2, boiled substrate. B, dsDNA translocation assays were performed as described under "Experimental Procedures" using 1 and 3 pmol of proteins (as monomer). The 32 Plabeled oligonucleotide is marked in bold. The percent displacement of the labeled oligonucleotide is indicated as %.  (18). These observations led to the proposal that domain A may have a regulatory function(s) in vivo (18).
Furthermore, the surface of domain A is less conserved among MCMs of 27 archaeal species studied in comparison to domains B and C (Fig. 7). This observation suggests the possibility that the domain may be needed for protein-protein interactions as a regulatory mechanism and/or during the initiation process. Because in different organisms the structures of proteins interacting with MCM may be different, the MCM will have to adjust its structure and surface residues to facilitate these interactions. One obvious candidate for such interactions is the binding of the archaeal Cdc6 protein to MCM (14,17). The interactions between the two proteins were shown to regulate the helicase activity of MCM (14,17). In addition, it was demonstrated that the interactions are species-specific (14) and thus may be mediated by domain A.
The hypothesis that domain A plays a regulatory role may also be supported by the observation that the domain is required for dsDNA translocation (Fig. 6B). It was hypothesized that dsDNA translocation by the archaeal, eukaryal, and bacterial helicases may play a regulatory role during the process of initiation of DNA replication (15,29,31,32).
Supporting evidence for the role of domain A in regulating MCM helicase activity comes from studies conducted with an eukaryotic Mcm5 mutant (mcm5-bob1) (33). In this mutant a conserved Pro is substituted with Lys. Structural and biochemical studies of this mutant prompted Fletcher et al. (19) to propose a role for the mutation, and thus domain A, in the regulation of MCM activity.
Does the N-terminal Region of MCM Play Similar Roles in Eukarya?-The eukaryotic MCM is a family of six polypeptides with varying sizes but with shared helicase domains within their central part (2,6). Although the proteins were shown to form hexamers in solution, high resolution structures of the individual proteins or the complex have not yet been obtained. Only low resolution electron micrograph structures of the het-erohexameric Mcm2-7 (34) and of Mcm4,6,7 (35) complexes are available. These structures revealed a globular shape with a central cavity. The regions and/or domains needed for multimerization, however, are not yet known. It was shown, however, that the N-terminal 112 residues of murine Mcm4 are not required for hexamer formation (26). This region, however, is unique to Mcm4 and is not found in other eukaryotic MCM family members or the archaeal homologues ( Supplementary  Fig. 2).
The eukaryotic MCM helicase is a heterohexamer, which is different from the homohexamers of the archaeal enzymes. Therefore, the multimerization domain might be different in eukarya. To determine whether similar regions may be involved in multimerization of the archaeal and eukaryal complexes, a sequence alignment (ClustalW) was performed between the N-terminal domains of mtMCM and the human Mcm2-7 proteins (Supplementary Fig. 2). Similar to the observation in the archaeal proteins (Fig. 7), domain A is the least conserved among the six eukaryotic polypeptides. Domain C, on the other hand, is much more conserved with about 25% identity among the six eukaryotic family members. Furthermore, domain C does not contain any insertions or deletions within the primary amino acid sequence, suggesting that a structural fold may exist that cannot tolerate additional loops and alterations. Hence, these observations may suggest that in eukarya, similar to archaea, the region of MCM located next to the catalytic domain is involved in hexamer formation.