Duplication of Eukaryotic Chromosomes
Chromosomes are the permanent repositories of the information that directs cell function. Remarkably, during an average human lifetime each person will accurately synthesize millions of miles of DNA. Inaccurate, incomplete or excessive DNA replication can lead to cell death, developmental abnormality or cancerous transformation. Consistent with this critical role, chromosome duplication is carefully coordinated with cell division. Each time a cell divides it must accurately replicate the DNA at the foundation of each chromosome and reassemble the proteins that interpret this essential cellular blueprint.
Our studies focus on the events that occur at the starting points of chromosome duplication called origins of DNA replication. These DNA sequences are found at multiple sites on each eukaryotic chromosome and direct the assembly of a pair of complex DNA synthesis machines (replisomes) that will replicate the adjacent DNA. Each replisome consists of multiple enzymes whose function must be coordinated, including a DNA helicase, three DNA polymerases and an RNA polymerase. The events that occur at origins can be broken down into four steps: origin recognition, helicase loading, helicase activation and replisome assembly. To maintain proper number of each chromosome, helicase loading is tightly restricted to the G1 phase of the cell cycle whereas helicase activation and replisome assembly occur during S phase. By separating these events in the cell cycle, each origin of replication has only one opportunity to initiate replication per cell division, thereby preventing rereplication of any part of the genome (Fig. 1).
The six-protein origin recognition complex (ORC) mediates the initial selection of origin DNA sequences. ORC was initially discovered in the budding yeast, S. cerevisiae, based on its ability to bind to a conserved sequence present in this organism’s short (100-120 bp), well characterized origins of DNA replication. Interestingly, this sequence is present in many positions across the genome, yet only a subset of these sites are bound by ORC. In collaboration with Dr. David MacAlpine’s lab, we used genome-wide mapping of nucleosomes to demonstrate that nucleosome-depleted regions are a critical determinant of the sites bound by ORC (Fig. 2). Once bound to origin DNA, ORC is required to establish the precise position of nucleosomes on both sides of the origin. Importantly, previous experiments in the lab indicate that this positioning is critical for subsequent helicase loading. We are currently focused on understanding how these nucleosome-depleted regions are encoded at origins of replication and how the positioned nucleosomes facilitate helicase loading.
Unlike the situation in S. cerevisiae cells, the sequences that act at origins of replication are not well-defined in metazoan cells. This is consistent with the apparent lack of sequence-specific DNA binding activity for ORC derived from these organisms. To understand how origin selection occurs in metazoan cells, we have used genomic approaches to map ORC binding sites and replication origins. We initially applied these approaches to the fruit fly, D. melanogaster, and we are now extending these approaches to human cells with the goal of combining this information with biochemical studies of metazoan ORC to identify the determinants that mediate ORC localization and origin selection.
As cells enter the G1 phase of the cell cycle, chromatin-associated ORC recruits two other helicase loading factors Cdc6 and Cdt1 to the origin DNA (Fig. 3). Due to its association with Cdt1, the eukaryotic replicative DNA helicase, the Mcm2-7 complex, is also recruited to the origin DNA. In an ATP-dependent series of events, ORC, Cdc6 and Cdt1 load the DNA helicase onto the adjacent DNA. After origin loading, the two ring-shaped, hexameric Mcm2-7 complexes encircle double-stranded DNA (dsDNA) as a head-to-head double hexamer. This loaded form of the Mcm2-7 helicase marks all potential origins of replication but lacks helicase activity. This process is also called pre-replicative complex (pre-RC) formation or origin licensing.
We have investigated the process of helicase loading in S. cerevisiae cells extensively. We have used in vitro helicase loading assays to demonstrate that ATP hydrolysis by Cdc6 and ORC is required at distinct times during helicase loading. More recently we sought to understand how the asymmetric ORC and origin DNA loads a symmetric head-to-head pair of Mcm2-7 complexes. We found that the smallest ORC subunit (Orc6) has two binding sites for Cdt1. The intermediate formed upon initial origin recruitment of Mcm2-7 and the helicase loading proteins includes only one ORC molecule but multiple (most likely two) Cdt1 proteins and their associated Mcm2-7 (Fig. 3B). We identified mutants in Cdt1 that interfere with the formation of this complex and showed that they are lethal and prevent helicase loading. The existence of a multi-Cdt1/Mcm2-7 intermediate prior to helicase loading provides an easy solution for the establishment of the symmetry necessary for the formation of a bidirectional pair of replication forks. Our latest studies incorporate single-molecule approaches to address the changing protein-protein interactions during helicase loading and the conformation of the Mcm2-7 double hexamer to understand how helicase loading results in a complex that is both inactive and poised for activation in S phase.
In S. cerevisiae cells, cyclin-dependent kinases (CDKs) inhibit helicase loading outside of the G1 phase of the cell cycle. There are at least three helicase loading proteins that are inhibited by CDK phosphorylation: Cdc6, Mcm2-7 and ORC. Unlike the CDK-dependent degradation of Cdc6 or nuclear export of Mcm2-7, ORC remains present and on the DNA throughout S, G2 and M phases. Thus, the mechanism of ORC CDK inhibition must be intrinsic to ORC function during helicase loading. We showed that Orc6 was the primary target of CDK inhibition and that this modification inhibited one of the two Cdt1 binding sites on Orc6. Importantly, this phosphorylation prevents formation of the multi-Cdt1 helicase loading intermediate we identified above, indicating that this intermediate is not only important for helicase loading but it is the target of a key regulatory mechanism.
Helicase Activation and Replisome Assembly
Mcm2-7 complexes that are loaded onto the origin DNA during G1 are only activated upon entry into the S phase of the cell cycle (Fig. 4). Helicase activation requires the action of the Dbf4-dependent Cdc7 kinase (DDK) and S-phase cyclin-dependent kinase (S-CDK). The action of these kinases leads to the recruitment of two critical helicase activating factors, Cdc45 and GINS and the activation of the helicase. After helicase activation, the cell must assemble the rest of the DNA replisome including the three replicative DNA polymerases that are dedicated to leading (Pol e) or lagging (Pol a/primase, Pol d) strand synthesis.
We found that the loaded form of Mcm2-7 is preferentially targeted by DDK and this preference is directed by priming phosphorylation of Mcm2-7. We have mapped both DDK and priming Mcm2-7 phosphorylation sites. Genetic and biochemical analysis of these sites found that the priming phosphorylation is mediated by multiple kinases including the ATR-related Mec1 kinase.
We have recently developed new assays that recapitulate the origin-dependent initiation of replication using S. cerevisiae extracts and proteins. Importantly, these assays show the hallmarks of in vivo replication initiation including dependence on known replication factors, regulatory kinases and a defined origin of replication. We can readily deplete (and purify) individual replication proteins from the assay, allowing us to assess when during the replication process each protein functions. Using these and other approaches we determined that DDK and S-CDK have distinct and sequential functions during helicase activation and that DDK is active prior to S phase at early origins of replication. By analyzing replisome assembly, we found that the leading strand DNA polymerase is recruited prior to its lagging strand counterparts. This order of events ensures that the leading strand DNA polymerase is present before any RNA primer synthesis so that it can elongate the first primers.
The availability of the origin-dependent replication assay derived from the genetically tractable S. cerevisiae opens many new avenues to study chromosome duplication. Although Cdc45 and GINS are direct activators of Mcm2-7, it is increasingly clear that the Mcm2-7 double hexamer encircling dsDNA separates into single hexamers surrounding ssDNA during initiation (Fig. 5). We are currently using the replication assay and new assays derived from it to understand how the numerous proteins involved in helicase activation (Fig. 4) trigger and drive these changes. We have recently incorporated nucleosomal DNA templates into the replication assay and we are using these assays investigate the impact of nucleosomes on replication initiation and elongation. Finally, this assay will also allow us to investigate other replication-fork-dependent events including the intra-S-phase checkpoint and chromosome cohesion.
Figure 1: Temporal segregation of replication initiation events prevents multiple rounds of replication per cell cycle. Loading state: Helicase loading is allowed during G1 but helicase activation and replisome assembly are prevented due to the lack of key activating kinases. Activation state: Previously loaded helicases are activated and this leads to replisome assembly. During S, G2, and M-phase multiple mechanisms prevent new helicase loading. Only when the cell completes chromosome segregation and cell division can a new round of helicase loading be initiated.
Figure 2: Nucleosomes influence ORC binding in vivo. Only a subset of the potential ORC binding sites (called “ARS concensus sequence” or ACS) in the S. cerevisiae genome are bound by ORC. The ACS sites bound are typically located asymmetrically within a nucleosome-depleted region. After DNA binding ORC positions the nucleosomes on either side of the nucleosome depleted region, providing a site for subsequent loading of the Mcm2-7 DNA helicase.
Figure 3: Model for eukaryotic helicase loading. (A) Also known as origin licensing or pre-replicative complex (pre-RC) formation, helicase loading begins as cells complete cell division and enter G1-phase with origin-bound ORC. (B) Origin-bound ORC recruits Cdc6 and at this point both proteins are bound to ATP. (C) Mcm2-7 and Cdt1 form a complex in advance of origin recruitment. Helicase recruitment occurs when a pair of the Mcm2-7/Cdt1 complexes are recruited by two Cdt1 binding sites on Orc6 (labeled N-terminal and C-terminal) which bind two Orc6 binding sites on Cdt1 (labeled 1 and 2). (D) Helicase loading occurs when Cdc6 hydrolyzes its bound ATP, leading to release of Cdt1 from Mcm2-7 and two Mcm2-7 hexamers encircling the origin DNA in the form of a head-to-head double hexamer. The remaining helicase loading proteins are released at this stage and this event is likely coupled to ORC ATP hydrolysis.
Figure 4: Model for eukaryotic helicase activation and replisome assembly. The initial event in helicase activation is the phosphorylation of loaded Mcm2-7 complexes by DDK. This modification leads to the recruitment of two proteins including the helicase activating proteins, Cdc45. These events are independent of S-CDK activity and form a stable DDK-dependent complex. Induction of S-CDKs leads to the recruitment of a number of additional replication factors, including the second helicase activating protein, the GINS complex, and the leading strand DNA polymerase e. Recruitment of the lagging strand DNA polymerases requires DNA unwinding and the function of the Mcm10 replication factor.
Figure 5: A model for changes to the Mcm2-7 complex during helicase activation. As initially loaded, the Mcm2-7 exists as a double hexamer encircling dsDNA. Cdc45 and GINS binding is proposed to lead to helicase activation and the separation of Mcm2-7 into single hexamers that encircle ssDNA. The proteins and mechanisms behind this transition are currently unclear.
Takara, T. J., and Bell, S. P. (2011). Multiple Cdt1 molecules act at each origin to load replication-competent Mcm2-7 helicases. EMBO J.
Heller, R. C., Kang, S., Lam, W. M., Chen, S., Chan, C. S., and Bell, S. P. (2011). Eukaryotic origin-dependent DNA replication in vitro reveals sequential action of DDK and S-CDK kinases. Cell 146, 80–91.
Chen, S., and Bell, S. P. (2011). CDK prevents Mcm2-7 helicase loading by inhibiting Cdt1 interaction with Orc6. Genes Dev 25, 363–372.
Randell, J. C. W., Fan, A., Chan, C., Francis, L. I., Heller, R. C., Galani, K., and Bell, S. P. (2010). Mec1 is one of multiple kinases that prime the Mcm2-7 helicase for phosphorylation by Cdc7. Mol Cell 40, 353–363.
Tsakraklides, V., and Bell, S. P. (2010). Dynamics of pre-replicative complex assembly. Journal of Biological Chemistry 285, 9437–9443.
Eaton, M. L., Galani, K., Kang, S., Bell, S. P., and MacAlpine, D. M. (2010). Conserved nucleosome positioning defines replication origins. Genes Dev 24, 748–753.
Francis, L. I., Randell, J. C. W., Takara, T. J., Uchima, L., and Bell, S. P. (2009). Incorporation into the prereplicative complex activates the Mcm2-7 helicase for Cdc7-Dbf4 phosphorylation. Genes Dev 23, 643–654.