The Mcm complex: unwinding the mechanism of a replicative helicase

Abstract

The Mcm2-7 complex serves as the eukaryotic replicative helicase, the molecular motor that both unwinds duplex DNA and powers fork progression during DNA replication. Consistent with its central role in this process, much prior work has illustrated that Mcm2-7 loading and activation are landmark events in the regulation of DNA replication. Unlike any other hexameric helicase, Mcm2-7 is composed of six unique and essential subunits. Although the unusual oligomeric nature of this complex has long hampered biochemical investigations, recent advances with both the eukaryotic as well as the simpler archaeal Mcm complexes provide mechanistic insight into their function. In contrast to better-studied homohexameric helicases, evidence suggests that the six Mcm2-7 complex ATPase active sites are functionally distinct and are likely specialized to accommodate the regulatory constraints of the eukaryotic process.

FIG. 1.

Evolutionary relationship among the Mcm families. The indicated Mcm sequences were aligned using ClustalX (141), and the phylogenic relationships among them were drawn as an unrooted tree using the Drawtree interface of PHYLIP 3.67 (http://mobyle.pasteur.fr/cgi-bin/MobylePortal/portal.py?form=drawtree). It should be noted that although the archaeal Mcm proteins are frequently shown to be most closely related to Mcm4 (42, 129), our more extensive analysis using a larger number of Mcm sequences failed to duplicate these results. Mcm genes from the following organisms were used: Homo sapiens (Hs), Danio rerio (Dr), Drosophila melanogaster (Dm), Caenorhabditis elegans (Ce), Schizosaccharomyces pombe (Sp), Saccharomyces cerevisiae (Sc), Arabidopsis thaliana (At), Oryza sativa (Os), Dictyostelium discoideum (Dd), Tetrahymena thermophila (Tt), Giardia lamblia (Gl), Sulfolobus solfataricus (Sso), Methanobacterium thermoautotrophicum (Mth), “Nanoarchaeum equitans” (Neq), and “Korarchaeum cryptofilum” (Kc). Note that although bacteria lack the Mcm proteins, several phages possess an MCM-like gene (e.g., a Bacillus cereus prophage [173], the “Haloarcula sinaiiensis” archaephage HSTV-1 [R. Hendrix, personal communication], and the archaeal virus BJ1 [E. Pagaling et al., unpublished data; see http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?val=CAL92457]); the BJ1 Mcm homologue has also been included in Fig. 1. GenBank accession numbers are as follows: AtMcm2, NP_175112.2; AtMcm3, NP_199440.1; AtMcm4, NP_179236.3; AtMcm5, NP_178812.1; AtMcm6, NP_680393.1; AtMcm7, NP_192115.1; AtMcm8, NP_187577.1; AtMcm9, NP_179021.2; BJ1 Mcm, YP_919062; CeMcm2, NP_001022416.1; CeMcm3, NC_003283.9; CeMcm4, NP_490962.1; CeMcm5, NP_497858.1; CeMcm6, NP_001023011.1; CeMcm7, NP_504199.1; DdMcm2, XP_637579.1; DdMcm3, NC_007089.3; DdMcm4, NC_007088.4; DdMcm5, NC_007092.2; DdMcm6, NC_007088.4; DdMcm7, NC_007090.2; DdMcm8, XP_639313.1; DdMcm9, XP_637904.1; DmMcm2, NP_477121.1; DmMcm3, NP_511048.2; DmMcm4, NP_477185.1; DmMcm5, NP_524308.2; DmMcm6, NP_511065.1; DmMcm7, NP_523984.1; DrMcm2, NP_775364.1; DrMcm3, NP_997732.1; DrMcm4, NP_944595.1; DrMcm5, NP_848523.2; DrMcm6, NP_001076318.1; DrMcm7, NP_997734.1; GlMcm2, EAA40971.1; GlMcm3, EAA36979.1; GlMcm4, EAA40854.1; GlMcm5, EAA38067.1; GlMcm6, EAA40537.1; GlMcm7, EAA40792.1; HsMcm2, NP_004517.2; HsMcm3, NP_002379.2; HsMcm4, NP_005905.2; HsMcm5, NP_006730.2; HsMcm6, NP_005906.2; HsMcm7, NP_005907.3; HsMcm8, NP_115874.3; HsMcm9, NP_694987.1; Mcm_Kc, ACB08098.1; MthMcm, AAB86236.1; Nanoarchaeum equitans Mcm, AAR39132.1; OsMcm2, NP_001067910.1; OsMcm3, NP_001055835.1; OsMcm4, AP004232.4; OsMcm5, NP_001048396.1; OsMcm6, NP_001054989.1; OsMcm7, NP_001067020.1; ScMcm2, NP_009530.1; ScMcm3, NP_010882.1; ScMcm4, NP_015344.1; ScMcm5, NP_013376.1; ScMcm6, NP_011314.2; ScMcm7, NP_009761.1; SpMcm2, NP_595477.1; SpMcm3, NP_587795.1; SpMcm4, NP_588004.2; SpMcm5, XP_001713071; SpMcm6, NP_596614.1; SpMcm7, NP_596545.1; SsoMcm, AAK41071.1; TtMcm2, XP_001009217.2; TtMcm3, XP_001012838.1; TtMcm4, XP_001018113.1; TtMcm5, XP_001007780.1; TtMcm6, XP_001013258.1; TtMcm7, XP_001008253.2; TtMcm8, XP_001012433.2; TtMcm9, XP_001032062.2.

FIG. 2.

Organization of the Mcm structural motifs. (A) Cartoon showing the domain structure and linear organization of SsoMcm. Purple denotes structural elements, green denotes β-hairpins, blue denotes cis-acing ATPase elements, red denotes trans-acting ATPase elements, and yellow denotes the presensor 2 insertion. N-T hp, N-terminal β-hairpin; Ext hp, external β-hairpin; WA, Walker A motif; H2I, helix 2 insert β-hairpin; WB, Walker B motif; PS1 hp, presensor 1 β-hairpin; S1, sensor 1; RF, arginine finger motif; Pre-S2, presensor 2 insertion; S2, sensor 2. (B) AAA⁺ active sites are formed at the interface between adjacent subunits. The Walker A, sensor 1, and Walker B motifs act in cis while the arginine finger and sensor 2 motifs act in trans to hydrolyze ATP. The nucleophilic water molecule is oriented by sensor 1 and the Walker B motif. Note that the trans arrangement of the sensor 2 motif appears to be specific for the Mcm subclade of AAA⁺ proteins. (C) Shared motifs among the Mcm proteins. Abbreviations and color coding are the same as described above (A). The S. solfataricus (SsoMcm), G. lamblia Mcm2-7, S. cerevisiae Mcm2-7 (ScMcm2-7), and human Mcm2-7 protein sequences used to generate Fig. 1 were aligned with CLUSTALW (141). The G. lamblia and human sequences, the remaining gaps shared by the seven displayed sequences, and the nonconserved N- and C-terminal regions were then removed due to spatial constraints. Residues within the zinc finger predicted to be important for coordinating Zn²⁺ are highlighted in yellow. Residues conserved among the 19 sequences in the original alignment are shaded in gray.

FIG. 3.

Alignment of Mcm presensor 2 insertions. The presensor 2 inserts (defined in the legend of Fig. 2C) were excised from the Clustal alignment of all 60 eukaryotic Mcm proteins used to generate Fig. 1. The G. lamblia (Gl) proteins align poorly with the related Mcm proteins in this region, and thus, the corresponding Mcm2 and Mcm3 sequences have been truncated for presentation purposes (excised regions marked with the number of amino acids removed in parentheses). Vertical green bars denote conserved amino acid identities (conserved, 9 of 10 sequences within any Mcm group), and yellow corresponds to conserved amino acid similarities (again, at least 9 out of 10 sequences) according to the following groups: aromatic (F, T, and W), hydrophobic (L, I, and V), hydrophilic (S, T, and C), basic (K, R, and H), and acidic (E and D). Within the spacer region, red corresponds to acidic residues, while blue corresponds to basic residues. Asterisks indicate residues with either conserved identities or similarities shared among all six Mcm groups.

FIG. 4.

Mcm involvement in eukaryotic DNA replication initiation and elongation. (A) During G₁ phase, Cdc6 and Cdt1 recruit and load Mcm2-7 to origins of replication (marked by the binding of Orc1-6) to form a stable and inactive complex called the pre-RC. (B) In late G₁/early S phase, the pre-RC is somehow activated for DNA unwinding by the CDKs and DDK. This facilitates the loading of additional replication factors (e.g., Cdc45, Mcm10, GINS, polymerase α/primase, and DNA polymerases δ and ɛ) and unwinding of the DNA at the origin (not shown). (C) During S phase, Cdc6 and Cdt1 are degraded or inactivated (represented by dashed outlines) to block additional pre-RC formation, and bidirectional DNA replication ensues. For diagrammatic purposes, only one replication fork is shown, and many replication factors (e.g., the processivity clamp [PCNA], its loader [replication factor C], and factors required for Okazaki fragment processing, etc.) are omitted. When the replication fork encounters lesions in the DNA (red asterisk), the S-phase checkpoint response (via the Mrc1/Tof1/Csm3 [M/T/C] complex) slows or stops fork progression and stabilizes the association of Mcm2-7 with the replication fork during DNA repair.

FIG. 5.

Mechanistic models of hexameric helicase function. (A) Coordination of ATP hydrolysis. In each model, circles represent ATPase active sites, and the alternating position of the dark green circles represents the location of ATP hydrolysis as a function of time within the respective hexamers. (B) Models for DNA unwinding. (Adapted from reference with permission from Elsevier.) In the steric-exclusion model, the helicase encircles and translocates along one strand of ssDNA and unwinds the duplex DNA by the exclusion of the other strand. In the rotary-pump model, the helicases load at origins and translocate away from them, where they are eventually anchored (dark vertical lines). They then rotate the intervening DNA in opposite directions, causing the unwinding of the origin. In the dsDNA pump model, two helicases form a head-to-head complex and pump dsDNA toward the origin, where it is extruded as single strands. In the ploughshare model, the helicase encircles dsDNA and, after local melting of the DNA duplex at the origin, “drags” a rigid protein or protein domain (triangle) that acts as a wedge to separate the DNA strands. Arrows indicate the direction of DNA and/or helicase movement.

FIG. 6.

Involvement of archaeal Mcm structural motifs in ATP hydrolysis and DNA unwinding. (Adapted from reference with permission from Elsevier.) Data from mutant doping studies with Walker A (K346A) (A and D), sensor 2 (R560A) (B and D), and arginine finger-alanine mutants (R473A) (C and D) are shown. The indicated mutants were mixed with wild-type SsoMcm and then assayed for ATPase (A to C) in the presence and absence of DNA or helicase (D) activities; assays were performed between 3 and 10 times, and error bars represent the standard deviations of the results. Red or blue lines represent mathematical simulations (for details, see reference 180). In A and B, the indicated simulation assumes that ATPase activity is linearly proportional to the number of wild-type Mcm active sites present. In C, the simulation is according to the “wild-type-mutant pairs simulation” model with a value of s equal to 3. This simulation assumes that although the active sites containing an arginine finger mutation are catalytically inactive, they still bind ATP and stimulate the activity of the adjacent wild-type ATPase active site. In D, the red simulation is the “pairs” model with a P value of 2 and follows the assumption that the activity at any wild-type Mcm ATPase active site depends on the presence of an adjacent wild-type active site. The exponential decrease that would be observed if a single mutant subunit within the hexamer blocks helicase activity is shown in blue.

FIG. 7.

Archaeal Mcm structural architecture and proposed unwinding modes. (A) Location of Mcm structural motifs within the monomer crystal structure of SsoMcm. The images were made using PYMOL (http://pymol.sourceforge.net/) and Protein Data Bank accession number 3F9V, which includes amino acids 7 to 601 of SsoMcm (30). The coloring and abbreviations are the same as those defined in the legend of Fig. 2A. Each of the four different views are 90° rotations of the monomer crystal structure. The trans and cis faces show the dimer interface between subunits that contains the trans- or cis-acting ATPase motifs. The external face views the monomer from the outside of the hexamer looking toward the central channel, while the channel face views the monomer from inside the central channel looking out. (B) Top and side views of the hexamer organization of SsoMcm. (Reprinted from reference with permission of the publisher. Copyright 2008 National Academy of Sciences, U.S.A.) (C) Schematic representation of an SsoMcm hexameric helicase. The four β-hairpins (N-terminal β-hairpin [NT], helix 2 insert β-hairpin [H2I], presensor 1 β-hairpin [PS1], and external β-hairpin [EXT]) are represented by short solid bars; the central channel and side channels are in darker shading. (D) dsDNA pump mode showing ssDNA extrusion from the side channel. (E) Steric exclusion model of a single SsoMcm helicase. DNA is shown as gray lines. Arrows indicate the direction of helicase movement. (Panels B to E reprinted and legends adapted from reference with permission of the publisher. Copyright 2008 National Academy of Sciences, U.S.A.)

FIG. 8.

Mcm2-7 helicase activity is glutamate dependent. Lanes contained either no Mcm (1 and 2), 400 ng Mcm467 (lanes 3 to 6), or 400 ng Mcm2-7 (lanes 7 to 10) and were supplemented with potassium glutamate as indicated. (Figure reprinted and legend adapted from reference with permission from Elsevier.)

FIG. 9.

Speculative model for coupling Mcm2/5 gate activity with DNA unwinding. (A) Wild-type Mcm2-7. The Mcm2/5 gate alternates between open and closed conformations. In the open conformation, the helicase activity is turned off by conformational changes propagated through the complex; in the closed conformation, a different set of conformational changes is propagated through the complex that activates helicase activity. (B) Mutations that destroy normal gate activity (i.e., Mcm5KA and Mcm2RA) prevent propagation of the activating conformation. (C and D) Mcm2KA (C) or Mcm3KA (D) mutations cause a partial reduction in helicase activity by a loss of ATP binding at the affected active site (indicated) that partially blocks the activating conformation from reaching the Mcm7/4 site. (E) The double-mutant complex has little or no ATPase activity since no activation signal reaches the Mcm7/4 site. Green circles represent wild-type subunits, red hexagons represent arginine finger R→A mutations, and red squares represent Walker A K→A mutations.

FIG. 10.

Speculative model for Mcm2/5 gate involvement during DNA replication. (A) In vitro, the Mcm2-7 complex exists in equilibrium between open and closed states, but its topology in vivo is unknown. (B) The open state may facilitate DNA loading. Cdc6 and Cdt1 recruit the Mcm2-7 complex to origins of replication (marked by Orc1-6 [not shown]). An open Mcm conformation may be required to load DNA into the central channel. In late G₁/early S phase, the Mcm2-7 complex is activated for DNA unwinding. The closure of the Mcm ring and helicase activation may occur by several means. (C) Cdc7/Dbf4 phosphorylation may close the Mcm2/5 gate and activate DNA unwinding. Small red circles indicate phosphorylation. (D) Alternatively, the loading of Mcm10, the GINS complex, and Cdc45 may either (i) activate the helicase, (ii) act as processivity factors by stabilizing the closure of the Mcm2/5 gate to prevent dissociation from the DNA (depicted), or (iii) regulate the activity of individual Mcm active sites in some unknown fashion. Purple hexagons, Mcm10; blue parallelograms, GINS heterotetramer; red ovals, Cdc45.

FIG. 11.

Models of Mcm2-7 evolution. (A) Divergent evolution of Mcm2-7. (B) Convergent evolution of Mcm2-7. See the text for details.