Research Interests
We are studying how cells accurately replicate their DNA, a process that begins at specific DNA sequences termed replication origins. Various genome-wide approaches have identified from 320 to 420 possible replication origins in budding yeast, the model organism we study. In G1 phase, each origin assembles approximately 40 polypeptides in a temporally defined order, culminating in the initiation of DNA replication at the G1/S phase boundary. The first stage of this process is called pre-replicative complex assembly and requires the origin recognition complex (ORC), Cdc6p, and Cdt1p. ORC directly binds to DNA and then recruits Cdt1p and Cdc6p during early G1 phase. These three proteins cooperate to load the MCM DNA helicase at origins in an ATP-dependent reaction. Cyclin-dependent kinases and the Cdc7p-Dbf4p kinase then catalyze the association of additional proteins with the MCM helicase, ultimately causing the initiation of bi-directional DNA synthesis (Figure 1). In our lab we are studying how Cdc6p-ATP functions to load the MCM helicase within a chromatin context in budding yeast. We also are studying the Cdc7p-Dbf4p kinase in both yeast and human cells.
We previously discovered that deletion of SIR2, encoding a histone deacetylase, rescued the temperature sensitivity of several mutants that were defective in pre-RC assembly, including a cdc6-4 mutant. We screened the replication origins on chromosomes III and VI to identify those origins that were inhibited by SIR2 and identified five SIR2-sensitive origins: ARS305, ARS315, ARS317, ARS603, and ARS606. We determined the detailed structure of two origins on chromosome III and found that these origins contain inhibitory elements distal to the ORC binding site. By utilizing data from another group that mapped stably positioned nucleosomes on chromosome III, we found that these inhibitory elements were positioned within stably bound nucleosomes. Furthermore, the positioned nucleosomes were very close to or overlapping the site of pre-RC assembly. Origins that were not inhibited by SIR2 did not have nearby nucleosomes in this region. Since genetically, SIR2 inhibited origins through the inhibitory element, we suggest that the acetylation state of this nucleosome affects pre-RC assembly. We do not know whether Sir2p is acting directly at these inhibitory sites or indirectly by regulating another gene, but we put forth the following model to explain our findings (Figure 2). Because pre-RC assembly occurs on naked DNA and because nucleosomes within the origin inhibit pre-RC assembly, some origins may exist within a nucleosome environment that is not optimal for pre-RC assembly. In those cases, either a particular modification of a nearby nucleosome or a protein that binds to the nucleosome in a SIR2-dependent manner inhibits pre-RC assembly. Since Sir2p was previously thought to only act at very specific heterochromatic regions in the genome, we are interested to discover exactly how Sir2p acts at these euchromatic sites.
 |
Figure 1. The stages of replication initiation. In the first stage, a multi-subunit complex called the pre-replicative complex (pre-RC) is assembled at replication origins. This complex consists of ORC, Cdt1p, Cdc6p, and the MCM helicase. In the second stage, Cdc7p-Dbf4p and Cdk kinases activate the MCM helicase, resulting in origin unwinding and the association of DNA polymerases at the origin. |
 |
Figure 2. Model for SIR2-mediated inhibition of DNA replication. A) Schematic of the ARS315 element that contains an inhibitory sequence element (IS) downstream of the site for pre-RC assembly. B) The IS element positions a nucleosome close to B2, which is inhibitory for pre-RC assembly. Deletion of SIR2 or mutation of the IS element allows repositioning of the nucleosome to allow pre-RC assembly. |
Cdc7p-Dbf4p is a two-subunit protein kinase required for initiating DNA replication after MCM helicase loading. The Cdc7p kinase subunit binds Dbf4p, which activates its kinase activity. Although Cdc7p-Dbf4p is required to promote DNA replication, it also has an undefined role in the repair of certain DNA lesions. In order to define the amino acids required for its roles in DNA replication and repair, we are analyzing Dbf4p using a mutational approach. We found that the Dbf4p N-terminus is dispensable for DNA replication, but it encodes functions that participate in the repair of DNA lesions and the firing of late-replication origins. An N-terminal BRCT-like motif may mediate these activities. It is also possible that Dbf4p maintains replication fork stability by targeting Cdc7p kinase to stalled replication forks. Interestingly, these are separable activities from Dbf4p’s essential role in promoting initiation of DNA replication. In addition, we have identified two regions within Dbf4p—a C-terminal Zn-finger motif and a separate region—that mediate binding to and activation of the Cdc7p kinase. The C-terminal region is also not essential for Dbf4p activity, but loss of this region dramatically lowers Cdc7p kinase activity.
We also are studying the human Cdc7-Dbf4 protein kinase. We previously raised monoclonal antibodies that recognize both human subunits and used these to screen human cancer cell lines and primary human tumors for HsCdc7-Dbf4 abundance. Although both subunits are expressed at very low levels in normal cycling cells (and are perhaps absent in post-mitotic cells), they are up-regulated in a substantial number of tumor cell lines. HsCdc7 protein is also highly expressed in some primary breast and colon tumors. By screening expression data from a panel of more than 650 primary human tumors, we also found that CDC7 and DBF4 mRNA expression are coordinately up-regulated in many tumors of diverse origin. Similarly, we screened several primary tumors and tumor cell lines for gene copy changes in CDC7 and DBF4. We found that the DBF4 gene copy number is often elevated in those cells expressing higher levels of Cdc7-Dbf4 kinase.
Because HsCdc7 is an essential kinase for DNA replication, its increased expression level in some tumors and tumor cell lines may reflect higher rates of cellular proliferation. However, we find no correlation between doubling time and Cdc7-Dbf4 expression in the NCI60 tumor cell lines. Furthermore, several published studies suggest that increased Cdc7-Dbf4 expression can inhibit the growth of rodent cell lines but has no effect on the growth of human cells. We suggest that since HsCdc7-Dbf4 is likely involved in other aspects of chromosome metabolism (e.g., DNA repair) and functions in the S-phase checkpoint, its increased expression in some tumor cell lines may offer an advantage for handling the chromosome instability that occurs in many human tumors. Our aim is to understand the mechanism(s) that allow increased expression of Cdc7-Dbf4 kinase in tumor cells and to investigate its phenotypic effects.
External Collaborators
- Angelika Amon, Massachusetts Institute of Technology, Cambridge
- Catherine Fox, University of Wisconsin–Madison
- Carol Newlon, University of Medicine and Dentistry of New Jersey, Newark
- Alain Verreault, University of Montreal, Quebec, Canada
