Sandra A. Jablonski

HEF1-dependent Aurora A activation induces disassembly of the primary cilium

The mammalian cilium protrudes from the apical/lumenal surface of polarized cells and acts as a sensor of environmental cues. Numerous developmental disorders and pathological conditions have been shown to arise from defects in cilia-associated signaling proteins. Despite mounting evidence that cilia are essential sites for coordination of cell signaling, little is known about the cellular mechanisms controlling their formation and disassembly. Here, we show that interactions between the prometastatic scaffolding protein HEF1/Cas-L/NEDD9 and the oncogenic Aurora A (AurA) kinase at the basal body of cilia causes phosphorylation and activation of HDAC6, a tubulin deacetylase, promoting ciliary disassembly. We show that this pathway is both necessary and sufficient for ciliary resorption and that it constitutes an unexpected nonmitotic activity of AurA in vertebrates. Moreover, we demonstrate that small molecule inhibitors of AurA and HDAC6 selectively stabilize cilia from regulated resorption cues, suggesting a novel mode of action for these clinical agents.

The RanGAP1-RanBP2 Complex Is Essential for Microtubule-Kinetochore Interactions In Vivo

RanGAP1 is the activating protein for the Ran GTPase. Vertebrate RanGAP1 is conjugated to a small ubiquitin-like protein, SUMO-1. This modification promotes association of RanGAP1 with the interphase nuclear pore complex (NPC) through binding to the nucleoporin RanBP2, also known as Nup358. During mitosis, RanGAP1 is concentrated at kinetochores in a microtubule- (MT) and SUMO-1-dependent fashion. RanBP2 is also abundantly found on kinetochores in mitosis. Here we show that ablation of proteins required for MT-kinetochore attachment (Hec1/Ndc80, Nuf2 ) disrupts RanGAP1 and RanBP2 targeting to kinetochores. No similar disruption was observed after ablation of proteins nonessential for MT-kinetochore interactions (CENP-I, Bub1, CENP-E ). Acquisition of RanGAP1 and RanBP2 by kinetochores is temporally correlated in untreated cells with MT attachment. These patterns of accumulation suggest a loading mechanism wherein the RanGAP1-RanBP2 complex may be transferred along the MT onto the kinetochore. Depletion of RanBP2 caused mislocalization of RanGAP1, Mad1, Mad2, CENP-E, and CENP-F, as well as loss of cold-stable kinetochore-MT interactions and accumulation of mitotic cells with multipolar spindles and unaligned chromosomes. Taken together, our observations indicate that RanBP2 and RanGAP1 are targeted as a single complex that is both regulated by and essential for stable kinetochore-MT association.

Mapping the assembly pathways that specify formation of the trilaminar kinetochore plates in human cells

We report the interactions amongst 20 proteins that specify their assembly to the centromere–kinetochore complex in human cells. Centromere protein (CENP)-A is at the top of a hierarchy that directs three major pathways, which are specified by CENP-C, -I, and Aurora B. Each pathway consists of branches that intersect to form nodes that may coordinate the assembly process. Complementary EM studies found that the formation of kinetochore trilaminar plates depends on the CENP-I/NUF2 branch, whereas CENP-C and Aurora B affect the size, shape, and structural integrity of the plates. We found that hMis12 is not constitutively localized at kinetochores, and that it is not essential for recruiting CENP-I. Our studies also revealed that kinetochores in HeLa cells contain an excess of CENP-A, of which ∼10% is sufficient to promote the assembly of normal levels of kinetochore proteins. We elaborate on a previous model that suggested kinetochores are assembled from repetitive modules (Zinkowski, R.P., J. Meyne, and B.R. Brinkley. 1991. J. Cell Biol. 113:1091–110).

The hBUB1 and hBUBR1 kinases sequentially assemble onto kinetochores during prophase with hBUBR1 concentrating at the kinetochore plates in mitosis

Abstract. The kinetochore binds an evolutionarily conserved set of checkpoint proteins that function to monitor whether chromosomes have aligned properly at the spindle equator. Human cells contain two related protein kinases, hBUB1 and hBUBR1, that appear to have evolved from a single ancestral BUB1 gene. We generated hBUB1- and hBUBR1-specific antibodies so that the localization patterns of these kinases could be directly compared. In the human U2OS osteosarcoma cell line, hBUB1 first appeared at kinetochores during early prophase before all kinetochores were occupied by hBUBR1 or CENP-F. Both proteins remained at kinetochores throughout mitosis but their staining intensity was reduced from anaphase onward. Kinetochores of unaligned chromosomes exhibited stronger hBUB1 and hBUBR1 staining. Immunoelectron microscopy showed that hBUBR1 appeared to be concentrated in the outer kinetochore plate and in some instances the inner plate as well. When chromosome spreads were examined by light microscopy, hBUB1 and hBUBR1 were coincident with CENP-E. This suggests that both kinases are concentrated near the surface of the kinetochore where they can monitor kinetochore-microtubule interactions.

Human Zw10 and ROD are mitotic checkpoint proteins that bind to kinetochores

Here we show that human Zeste White 10 (Zw10) and Rough deal (Rod) are new components of the mitotic checkpoint, as cells lacking these proteins at kinetochores fail to arrest in mitosis when exposed to microtubule inhibitors. Checkpoint failure and premature mitotic exit may explain why cells defective for hZw10 and hRod divide with lagging chromosomes. As Zw10 and Rod are not conserved in yeast, our data, combined with an accompanying study of Drosophila Zw10 and Rod, indicate that metazoans may require an elaborate spindle checkpoint to monitor complex kinetochore functions.

Re-expression of the tumor suppressor NF2/merlin inhibits invasiveness in mesothelioma cells and negatively regulates FAK.

The neurofibromatosis type 2 NF2 gene product, merlin, is a tumor suppressor frequently inactivated in malignant mesothelioma (MM). To investigate a possible correlation between merlin inactivation and MM invasiveness, we restored merlin expression in NF2-deficient MM cells. Re-expression of merlin markedly inhibited cell motility, spreading and invasiveness, properties connected with the malignant phenotype of MM cells. To test directly whether merlin inactivation promotes invasion in a nonmalignant system, we used small interfering RNA to silence Nf2 in mouse embryonic fibroblasts (MEFs) and found that downregulation of merlin resulted in enhanced cell spreading and invasion. To delineate signaling events connected with this phenotype, we investigated the effect of merlin expression on focal adhesion kinase (FAK), a key component of cellular pathways affecting migration and invasion. Expression of merlin attenuated FAK phosphorylation at the critical phosphorylation site Tyr397 and disrupted the interaction of FAK with its binding partners Src and p85, the regulatory subunit of phosphatidylinositol-3-kinase. In addition, NF2-null MM cells stably overexpressing FAK showed increased invasiveness, which decreased significantly when merlin expression was restored. Collectively, these findings suggest that merlin inactivation is a critical step in MM pathogenesis and is related, at least in part, with upregulation of FAK activity.

Synthetic Lethal Screen of an EGFR-Centered Network to Improve Targeted Therapies

A targeted RNAi screen reveals potential targets for combination approaches to cancer treatment. Rationally Designing Combination Therapy Drug resistance is a problem in cancer treatment, making combination therapies common. However, all too often, resistance also develops to empirically developed combination therapies, or those combinations are generally cytotoxic and not selective for the cancer cells. Astsaturov et al. developed a library of candidate genes centered on the epidermal growth factor receptor (EGFR) and targeted these genes with silencing RNAs to identify candidate proteins that could be inhibited to reduce cancer cell viability in the presence of EGFR inhibitors. Cotreatment with EGFR inhibitors and clinically available drugs that inhibit the candidate proteins reduced tumor size in xenografts and cell viability of multiple cancer cell lines. These results suggest that this network-centered approach may be fruitful for development of rationally designed combination therapies. Intrinsic and acquired cellular resistance factors limit the efficacy of most targeted cancer therapeutics. Synthetic lethal screens in lower eukaryotes suggest that networks of genes closely linked to therapeutic targets would be enriched for determinants of drug resistance. We developed a protein network centered on the epidermal growth factor receptor (EGFR), which is a validated cancer therapeutic target, and used small interfering RNA screening to comparatively probe this network for proteins that regulate the effectiveness of both EGFR-targeted agents and nonspecific cytotoxic agents. We identified subnetworks of proteins influencing resistance, with putative resistance determinants enriched among proteins that interacted with proteins at the core of the network. We found that clinically relevant drugs targeting proteins connected in the EGFR network, such as protein kinase C or Aurora kinase A, or the transcriptional regulator signal transducer and activator of transcription 3 (STAT3), synergized with EGFR antagonists to reduce cell viability and tumor size, suggesting the potential for a direct path to clinical exploitation. Such a focused approach can potentially improve the coherent design of combination cancer therapies.

Inhibitors of histone deacetylases alter kinetochore assembly by disrupting pericentromeric heterochromatin.

The kinetochore, a multi-protein complex assembled on centromeric chromatin in mitosis, is essential for sister chromosome segregation. We show here that inhibition of histone deacetylation blocks mitotic progression at prometaphase in two human tumor cell lines by interfering with kinetochore assembly. Decreased amounts of hBUB1, CENP-F and the motor protein CENP-E were present on kinetochores of treated cells. These kinetochores failed to nucleate and inefficiently captured microtubules, resulting in activation of the mitotic checkpoint. Addition of histone deacetylase inhibitors prior to the end of S-phase resulted in decreased HP1-? on pericentromeric heterochromatin in S-phase and G2, decreased pericentromeric targeting of Aurora B kinase, resulting in decreased pre-mitotic phosphorylation of pericentromeric histone H3(S10) in G2, followed by assembly of deficient kinetochores in M-phase. HP1-?, Aurora B and the affected kinetochore proteins all were present at normal levels in treated cells; thus, effects of the inhibitors on mitotic progression do not seem to reflect changes in gene expression. In vitro kinase activity of Aurora B isolated from treated cells was unaffected. We propose that the increased presence in pericentromeric heterochromatin of histone H3 acetylated at K9 is responsible for the mitotic defects resulting from inhibition of histone deacetylation.

Astrin regulates Aurora-A localization

Alterations in the expression and activity of the centrosomal kinase, Aurora-A/STK15, affect genomic stability, disrupt the fidelity of centrosome duplication, and induce cellular transformation. A mitotic spindle-associated protein, astrin/DEEPEST, was identified as an Aurora-A interacting protein by a two-hybrid screen. Astrin and Aurora-A co-express at mitosis and co-localize to mitotic spindles. RNAi-mediated depletion of astrin abolishes the localization of Aurora-A on mitotic spindles and leads to a moderate mitotic cell cycle delay, which resembles the mitotic arrest phenotypes in siAurora-A treated cells. However, depletion of Aurora-A does not affect astrin localization, and co-depletion of both astrin and Aurora-A causes a mitotic arrest phenotype similar to depletion of siAurora-A alone. These results suggest that astrin acts upstream of Aurora-A to regulate its mitotic spindle localization.

Screening of Conditionally Reprogrammed Patient-Derived Carcinoma Cells Identifies ERCC3–MYC Interactions as a Target in Pancreatic Cancer

Purpose: Even when diagnosed prior to metastasis, pancreatic ductal adenocarcinoma (PDAC) is a devastating malignancy with almost 90% lethality, emphasizing the need for new therapies optimally targeting the tumors of individual patients. Experimental Design: We first developed a panel of new physiologic models for study of PDAC, expanding surgical PDAC tumor samples in culture using short-term culture and conditional reprogramming with the Rho kinase inhibitor Y-27632, and creating matched patient-derived xenografts (PDX). These were evaluated for sensitivity to a large panel of clinical agents, and promising leads further evaluated mechanistically. Results: Only a small minority of tested agents was cytotoxic in minimally passaged PDAC cultures in vitro. Drugs interfering with protein turnover and transcription were among most cytotoxic. Among transcriptional repressors, triptolide, a covalent inhibitor of ERCC3, was most consistently effective in vitro and in vivo causing prolonged complete regression in multiple PDX models resistant to standard PDAC therapies. Importantly, triptolide showed superior activity in MYC-amplified PDX models and elicited rapid and profound depletion of the oncoprotein MYC, a transcriptional regulator. Expression of ERCC3 and MYC was interdependent in PDACs, and acquired resistance to triptolide depended on elevated ERCC3 and MYC expression. The Cancer Genome Atlas analysis indicates ERCC3 expression predicts poor prognosis, particularly in CDKN2A-null, highly proliferative tumors. Conclusions: This provides initial preclinical evidence for an essential role of MYC–ERCC3 interactions in PDAC, and suggests a new mechanistic approach for disruption of critical survival signaling in MYC-dependent cancers. Clin Cancer Res; 22(24); 6153–63. ©2016 AACR.