Date of Award

Winter 12-2015

Embargo Period

4-20-2016

Degree Type

Dissertation (CMU Access Only)

Degree Name

Doctor of Philosophy (PhD)

Department

Biological Sciences

Advisor(s)

Brooke M. McCartney

Abstract

The multifunctional Adenomatous polyposis coli (APC) proteins negatively regulate Wnt signaling, stabilize microtubules, and indirectly regulate actin through effectors such as Asef and IQGAP. We have shown that both vertebrate APC (vAPC) and Drosophila APC1 through their basic domains can bundle and nucleate actin filaments, and collaborate with the formin Diaphanous (Dia) to efficiently nucleate actin assembly in vitro. In addition, Drosophila APC2 (lacking a basic domain) and Dia bind directly to each other, and are required for actin furrow extension in the embryo. In contrast, APC1 does not function in actin furrow extension. While these data suggest that APC-Dia collaborations are an evolutionarily conserved mode of actin filament assembly, significant gaps exist in our understanding of their mechanism and physiological relevance. We investigated the mechanism underlying the vAPC/APC1-Dia collaboration in detail, but how APC2 affects Dia without a basic domain is not known. Here we demonstrate that APC2 interacts with Dia through its ß-catenin binding 20 amino acid repeats (20Rs), and that 20R phosphorylation by GSK3ß regulates APC2’s actin furrow activity. In addition to its phosphor-regulation, we explored APC2’s minimal cortical localization mechanism to further understand the role of APC2 in actin furrow extension. Furthermore, we took advantage of the many coordinated actin-mediated processes that occur during Drosophila oogenesis to investigate the physiological role of APC1-mediated actin assembly. In stage 10B egg chambers, an array of cytoplasmic actin bundles (actin baskets) form around each nucleus in the nurse cells to secure them. Here we show that APC1, APC2 and Dia are required for the proper assembly of these actin baskets. In APC1 or dia mutants, the actin baskets fail to form at stage 10B, but surprisingly they are present at stage 11. This suggests that APC1 or Dia may be able to carry out actin assembly alone, but with reduced efficiency resulting in assembly delay. Additionally, we demonstrated that microtubules are required for proper actin cable formation and it is through the functions of microtubule-associated proteins (CLIP-190 and EB1) that this microtubule-actin interaction is mediated. The work I performed in the McCartney Lab suggests that actin assembly is multi-faceted and tightly regulated to control for the precise spatiotemporal assembly of actin structures in the cell. More importantly, my work showed for the first time that APC1 is an actin assembly factor in vivo. This work suggests that different proteins can modulate actin assembly and it is possible that many novel actin assembly mechanisms remain to be discovered. Additionally, my work validated that formins are modulated in vivo by several different proteins that can either enhance or inhibit formin function controlling the spatiotemporal assembly of actin. In conclusion, my work focused on investigating the multifaceted process of actin polymerization, allowing us to better understand how actin structures form in the cell and how several proteins collaborate in vivo to form these actin structures. Additionally, we know that microtubules and actin filaments coordinate to perform many basic cellular processes but how these two cytoskeletal networks coordinate is not well understood. Through my work, we found that actin assembly is mediated by microtubule +TIP binding proteins that interact with the actin polymerization machinery demonstrating that the microtubule cytoskeletal network directly influences actin assembly.

Share

COinS