Graduation Year

2016

Document Type

Thesis

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Biology (Cell Biology, Microbiology, Molecular Biology)

Major Professor

Mark Alexandrow, Ph.D.

Committee Member

Douglas Cress, Ph.D.

Committee Member

Teresita Muñoz-Antonia, Ph.D.

Committee Member

Meera Nanjundan, Ph.D.

Keywords

Cancer, DNA replication, retinoblastoma, MCM, Cdc6, Cdt1

Abstract

Tumorigenesis is a multifaceted set of events consisting of the deregulation of several cell-autonomous and tissue microenvironmental processes that ultimately leads to the acquisition of malignant disease. Transforming growth factor beta (TGFß) and its family members are regulatory cytokines that function to ensure proper organismal development and the maintenance of homeostasis by controlling cellular differentiation, proliferation, adhesion, and survival, as well as by modulating components of the cellular microenvironment and immune system. The pleiotropic control by TGFß of these cell intrinsic and extrinsic factors is intimately linked to the prevention of tumor formation, the specifics of which are dependent on the various cellular and/or molecular signaling contexts that exist for TGFß. The diverse roles and the various levels of signal control for TGFß lend themselves to certain characteristics that are more advantageous for cancers to usurp in order to promote tumorigenesis, while other anti-tumorigenic roles for TGFß are more beneficial to tumor development if they are circumvented or disabled.

Transforming growth factor ß1 (TGF-ß1) exerts its anti-tumor effects in large part by potently inhibiting cell cycle progression at any point in G1 phase to control the proliferation of a variety of cell lineages. Loss of sensitivity to TGF-ß1-induced cell cycle arrest is a crucial event during early tumorigenesis. Indeed, cancer cells of almost all tumor types display insensitivity to TGF-ß1 inhibition. As such, the pursuit of the molecular details underlying the TGF-ß1 growth arrest pathway is important for our understanding of cell cycle regulation, and significantly, how disruption of these mechanisms contributes to TGF-ß1 insensitivity and tumorigenesis.

TGF-ß1 inhibition of the cell cycle in G1 phase has been shown to involve two main transcriptionally based molecular events, including the induction of cyclin-dependent kinase (CDK) inhibitors and the suppression of the c-Myc protein. Both mechanisms contribute to the maintenance of the retinoblastoma (Rb) protein in its hypophosphorylated and antiproliferative form, thus preventing progression through the cell cycle. However, this type of regulation does not offer answers to all of the questions regarding TGF-ß1 arrest. While these transcriptional mechanisms provide explanations for TGF-ß1 arrest throughout most of G1, inhibition late in G1 by TGF-ß1 however, does not require any acute regulation of transcription. In addition, the chance to utilize canonical TGF-ß1 arrest mechanisms at this time has already passed (i.e. Rb is already hyperphosphorylated by late-G1). Previous work from our group shows instead that late-G1 TGF-ß1 cell cycle arrest requires an intact direct interaction between the N-terminus of Rb (RbN) and the C terminus of Mcm7, a subunit of the Cdc45-MCM-GINS (CMG) replicative helicase. Our studies show that TGF-ß1 exposure in late-G1 prevents the disassociation of Rb with fully assembled helicases, which remain inactive. In addition, it was found that early-G1 treatment with TGF-ß1 also targets CMG components, namely MCM protein accumulation (and therefore hexamer formation) in G1 is blocked. However, the residue(s) of RbN involved as well as the molecular mechanisms Rb utilizes for late-G1 TGF-ß1 arrest are not described, nor is it evident from this work if TGF-ß1 affects other genes involved in CMG assembly and/or activation. In the following study we explore these unanswered questions for TGF-ß1 growth arrest as a means to understand novel aspects of cell cycle regulation that must be abrogated during tumorigenesis. Our hypothesis is that CMG helicase control on some level is critical for all TGF-ß1-induced inhibition of cell cycle progression throughout the entire G1 phase.

In Chapter 2 herein we have investigated the details and mechanistic implications of the Rb/RbN inhibitory-interaction with the CMG helicase that is required for late-G1 TGF-ß1 arrest. We show that N-terminal exons of Rb that are lost in partially penetrant hereditary retinoblastomas inhibit DNA replication and elongation using a bipartite mechanism. Specifically, Rb exon 7 is necessary and sufficient to inhibit CMG helicase activation, while an independent loop domain within RbN that forms a projection blocks DNA polymerase α (Pol-α) and Ctf4 recruitment without affecting polymerases δ and ε or the CMG helicase. Individual disruption of exon 7 or the projection in RbN or Rb, as occurs in inherited cancers, partially impairs the ability of Rb/RbN to inhibit DNA replication and block G1-S cell cycle transit. Importantly, their combined loss abolishes these functions of Rb. Thus, TGF-ß1 cell cycle arrest in late-G1 requires the growth suppressive role of Rb in which replicative complexes are blocked directly via independent and additive N-terminal domains. TGF-ß1-induced arrest in late-G1 also requires the presence of Smad3 and Smad4, suggesting that a novel transcription-independent role may exist for Smad signaling proteins in blocking cell cycle transit directly in Rb-CMG inhibitory complexes.

TGF-ß1 is thought to require a functional Rb protein to inhibit the cell cycle at any point in G1 phase. Intriguingly, while cells lacking Rb (and the inhibitory N-terminal domains) lose sensitivity to TGF-ß1 arrest in late-G1, these same cells remain sensitive to TGF-ß1 inhibition in early-G1. This Rb-independent TGF-ß1 growth arrest also occurs in the absence of c-Myc and MCM suppression, as well as without CyclinE-Cdk2 inhibition, but requires Smad3 and Smad4 respectively. Here (Chapter 3) we have identified the mechanism by which TGF-ß1 achieves Smad-dependent G1 arrest in the absence of these common mediators. TGF-ß1 inhibits the assembly of CMG replicative helicases by suppressing the recruitment of the MCM complex to chromatin. Accordingly, the entire heterohexamer fails to load onto DNA. Cdc6 phosphorylation in its amino terminus is known to be required for Cdt1-dependent loading of the MCM complex. We show that in Rb-lacking cells early-G1 TGF-ß1 treatment blocks the phosphorylation of Cdc6 at serine 54, without affecting total Cdc6 protein levels, to prevent MCM heterohexamer formation on DNA. Consistent with TGF-ß1 signals targeting this recruitment and loading step, Cdt1 overexpression promotes S-phase entry in the presence of TGF-ß1, circumventing the need for Cdc6 phosphorylation. Importantly, Cdt1 requires an intact C-terminal MCM-binding domain in order to overcome this TGF-ß1-induced cell cycle arrest mechanism. These data indicate that early-G1 TGF-ß1 arrest can occur by perturbing Cdc6 phosphorylation to block Cdt1-mediated MCM recruitment and loading, leading to inhibition of CMG assembly and S-phase entry despite the lack of Rb and normal c-Myc and CyclinE-Cdk2 activities.

We conclude that the main event governing TGF-ß1-induced cell cycle arrest at any point in G1 is the inhibition of the assembly and/or activation of the replicative CMG helicase. However, TGF-ß1 growth arrest has a temporal dependence on the presence of the Rb protein. In normal cells containing Rb, the accumulation of MCM subunit proteins is blocked by TGF-ß1 in early-G1 and accordingly MCM heterohexamers are unable to form. However, if cells are allowed to transit to late-G1 when MCM complexes have already assembled on origins, but before functional CMG helicases have formed at G1-S, exposure to TGF-ß1 signaling prevents CMG activation via interactions with critical inhibitory domains within RbN. Cells lacking Rb (and these residues) are not sensitive to TGF-ß1 in late-G1. Surprisingly, these cells remain sensitive to TGF-ß1 early in G1 phase despite a lack of c-Myc/MCM protein suppression and CyclinE-Cdk2 inhibition. In these cells the recruitment and loading of the MCM complex is blocked to facilitate a TGF-ß1-mediated G1 arrest. It is only when this mechanism is overcome by Cdt1 overexpression that TGF-ß1 is unable to elicit cell cycle arrest in these cells. These data provide molecular explanations for studies reporting instances of TGF-ß1 arrest without canonical effectors, such as Rb, c-Myc loss, or CDK inhibitors. Additionally, this work argues for the development of novel cancer therapeutics targeting CMG helicase assembly or activation, the regulation of which is likely lost in a variety of TGF-ß1-insensitive and/or Rb-deficient malignancies. Indeed, reintroduction of these tumor suppressive pathways has shown efficacy in blocking growth of tumors or cancer cells lacking the same mechanisms. Our studies of Rb/RbN inhibition of DNA replication also provide proof of principle for this type of therapy, as well as the framework for how the CMG might be targeted by exploring further and perhaps mimicking Rb exon7-mediated CMG inhibition biochemically.

Share

COinS