D Sharp and D Buster (2005) Cell Cycle 4: 1482-85

Microtubule Length Regulation by the Kinesin-8 Motor Protein: Capping and Polymerizing Enzymes.

L. Reese1, A. Melbinger2, E. Frey1; 1Arnold Sommerfeld Center for Theoretical Physics (ASC) and Center for NanoScience (CeNS), Ludwig-Maximilians-Universität München, Munich, Germany, 2Institut Pasteur, Paris, France

In the cell, size and dynamics of filaments are tightly controlled. One important player in microtubule length regulation is the microtubule depolymerizing motor protein kinesin-8. The accumulation of this protein along microtubules enables a differential regulation of microtubule length. Cellular mechanisms of length regulation however, are still obscure due to the complicated mechanochemical dynamics of microtubules and the presence of a multitude of microtubule associated proteins. We extend a simple mathematical model based on what is known for the microtubule depolymerizing motor kinesin-8, to account for microtubule growth factors, like the microtubule polymerizing enzyme XMAP215. We predict that microtubule polymerizing enzymes significantly enhance the range over which microtubule dynamics can be regulated. Further, our findings shed light on the functioning of microtubule depolymerization by kinesin-8. We show that depending on microscopic details of kinesin-8's depolymerization mechanism, it's length regulatory function is reminiscent of dynamic instability.

279

The novel Joubert syndrome protein KIF7 is involved in the regulation of microtubular dynamics, cellular polarity and cell cycle progression.

C. Dafinger1, M. C. Liebau1,2, T. Benzing1,3, B. Schermer1,3; 1Department II of Internal Medicine and Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany,

2_{Department of Pediatrics, University of Cologne, Cologne, Germany,} 3_{University of Cologne,} Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases, Cologne, Germany

Joubert syndrome (JBTS) is a rare mostly autosomal recessively inherited developmental disorder characterized by a specific brain malformation with various additional pathologies. JBTS results from mutations in at least 17 different genes, all of which play a role in the formation or function of primary cilia. Primary cilia are essential for vertebrate development, and mutations affecting this organelle underlie a large group of diseases referred to as ciliopathies. We recently identified mutations in KIF7 as a cause of JBTS in a consanguineous family. KIF7 belongs to the family of kinesin motor proteins and is a known regulator of Sonic Hedgehog signaling. Knockdown of KIF7 in retinal pigment epithelial (RPE) cells affects the structure of

cilia, centrosomes and the Golgi network, and is associated with alterations in cell shape, rendering the cells larger and no longer able to align in a parallel manner as wild type RPE cells do, suggesting defects in cell polarity. Consistently, KIF7 was found to interact with the polarity complex protein PAR3. We have now subsequently analyzed the cell biological function of KIF7 in more detail. In migration assays over 48 hours KIF7 knockdown cells presented with impaired directed cellular growth likely due to a defect in cell cycle progression. FACS analysis revealed a considerable G2/M arrest in U2OS cells. Additionally, we found KIF7 to interact with the histone deacetylase HDAC6 and observed increased tubulin acetylation after knockdown of KIF7, which may impair correct transport of polarity complex proteins and could be causative for the additional cellular phenotypes observed in KIF7 knockdown cells. Changes in microtubule stability might function as underlying disease mechanisms impacting ciliary function, cellular polarity and cell cycle progression.

280

Effect of Laulimalide on Microtubule Dynamics during Cell Division.

A. Mazumdar1, G. Chan1; 1Experimental Oncology, Cross Cancer Institute, University of Alberta, Edmonton, AB, Canada

Small molecule inhibitors of microtubules are valuable both as tools for studying the biology of microtubules and in the development of anti-tumor agent. Laulimalide is one such recent molecule isolated from the sea-sponges and has taxol like properties. It inhibits cell-cycle proliferation and is highly effective in taxol resistant cells. In contrast to taxol, it binds to a different site at the exterior of β-tubulin near the charged c-terminal tail and stabilizes the microtubules in vitro. Cells treated with laulimalide show no evidence of microtubule bundling in interphase. Live cell analysis reveals that cells treated with laulimalide enter mitosis but chromosomes scatter from the metaphase plate. HeLa cells incubated with low dose of laulimalide have high mitotic index and lead to multiple ring shaped spindle poles along with defects in chromosome alignment and segregation. In order to understand the molecular pathways that are affected by laulimalide treatment, we did an extensive literature search looking for proteins, depletion of which creates multiple ring shaped spindle poles. We found that cells lacking microtubule associated proteins (MAPs), ch-TOG or HAUS, have multiple spindle poles that are ring-shaped. We found that ch-TOG and HAUS were both mislocalized in HeLa cells treated with laulimalide. This result implicates the involvement of ch-TOG and HAUS in the laulimalide phenotype. Immunofluorescence and RNAi studies reveal that these proteins are localized at the spindle poles and required for their maintenance. Also, we show that in cells treated with laulimalide, Eg5 kinesin motor protein is required to generate the multipolarity. Currently, we are using quantitative immunofluorescence and fluorescence recovery after photobleaching (FRAP) to study the differences in spindle microtubule organization caused by laulimalide versus taxol in interphase and mitotic PtK2 cells. This will allow us to understand the effect of laulimalide on microtubule dynamics in cells.

281

New Insights into the Inhibition of Microtubule Dynamics and Spindle Formation by the Novel Microtubule Stabilizer Taccalonolide AJ.

A. L. Risinger1, C. C. Rohena1, M. Lopus2, S. Riffle2, M. A. Jordan2, L. Wilson2, S. L. Mooberry1; 1_{University of Texas Health Science Center at San Antonio, San Antonio, TX,} 2_{University of} California at Santa Barbara, Santa Barbara, CA

Taccalonolide AJ causes cancer cells to arrest in mitosis due to the formation of abnormal, multipolar mitotic spindle asters. The morphology of these structures differs from those initiated

by paclitaxel. At the concentration that causes maximal G2/M accumulation, paclitaxel initiated

the formation of 2-3 relatively large diffuse spindle asters, while 5-7 compact spindle asters were observed in taccalonolide AJ-treated cells. Studies were conducted to determine if the mechanism of aberrant spindle aster formation differed between taccalonolide AJ and paclitaxel. High-content live-cell microscopy was used to investigate the initial formation and the fate of cells entering mitosis after treatment with each microtubule stabilizer. Two distinct mechanisms of multipolar spindle formation were observed: the slow fragmentation of an initial bipolar spindle or the initial formation of multiple spindle asters that seem to be nucleated from the cell cortex. While paclitaxel and taccalonolide AJ shared similar patterns of initial aster formation, dramatic differences were observed in the fate of these asters. Paclitaxel-treated cells contained multiple asters that appeared to fuse into 2-3 larger more diffuse asters. This is possibly the cell’s attempt to restore a bipolar spindle morphology. In contrast, taccalonolide AJ- treated cells contained multiple spindle asters that did not merge and instead persisted as more numerous and compact asters. Co-localization studies showed that while many microtubule associated proteins were eventually localized to mature taccalonolide AJ-induced asters, NuMa was unique in that it associated with microtubules very early in mitosis, even localizing to the asters nucleated at the cell cortex. Depletion experiments showed NuMa was not required for the cortical localization of microtubules, but that it may play an important role in aster maturation after drug treatment. Further mechanistic studies were conducted to determine whether taccalonolide AJ inhibited microtubule dynamics in a similar way to paclitaxel. These experiments demonstrated that sub-stoichiometric concentrations of taccalonolide AJ cause a dose-dependent suppression of microtubule dynamics. While taccalonolide AJ had no effect on the growth rate of microtubules, it significantly suppressed the shortening rates and catastrophe and rescue frequencies. We hypothesize that the ability of taccalonolide AJ to suppress microtubule dynamics is responsible for the aberrant spindle aster formation and resolution observed.

282

Neuropathic effects of eribulin, ixabepilone, paclitaxel, and vincristine on peripheral sensory neurons in vitro.

J. A. Smith1, A. Rifkind1, G. Smiyun1, J. Meinert1, S. Feinstein1, L. Wilson1, M. Jordan1; 1_{University of California, Santa Barbara, Santa Barbara, CA}

Peripheral neuropathy can be a serious dose-limiting toxicity associated with microtubule- targeting anticancer drug treatment. We examined the neuropathic effects of four microtubule- targeting drugs in vitro, eribulin mesylate (ERB), ixabepilone (IXA), paclitaxel (PAC), and vincristine (VCR) and using an immortalized rat peripheral sensory neuronal cell line, 50B11. The goal is to investigate a possible correlation between drug effects on microtubule assembly dynamics and neuronal cell damage. As an assay for cell viability we determined the ability of each drug to reduce total cellular ATP levels. We found that IXA, which suppresses microtubule dynamics and enhances microtubule polymer mass, was 20-fold weaker than PAC, which has a similar effect on microtubules (EC50, IXA, 6.7 nM; PAC, 0.33 nM). VCR and ERB, which

suppress microtubule dynamics but decrease microtubule mass, induced cell death with EC50

values of 0.36 nM and 0.82 nM, respectively. Thus, the order of potency for induction of 50B11 cell death is PAC=VCR>ERB>IXA. After 24 h incubation, the drugs induced a loss of neurites (neurite retraction) with potencies of: VCR>ERB>IXA> PAC. At 10 nM VCR, neurite number was reduced from 0.79 neurites/cell in control to 0.09 neurites/cell. Nearly complete neurite retraction occurred with ERB only at ≥ 100 nM (to 0.08 neurites/cell). Neither PAC nor IXA induced complete loss of neurites at ≤ 1000 nM. The microtubule-depolymerizing effects of VCR and ERB may be responsible for their stronger effects on neurite retraction. We measured drug effects on α-tubulin protein expression and tubulin acetylation by in-cell western.

Increased acetylation of α-tubulin is a correlate of increased microtubule stability. VCR strongly reduced α-tubulin protein levels, in agreement with Huff et al., 2010. α-tubulin was reduced by 60% at 100 nM VCR and by 38% at 100 nM ERB. No reduction in α-tubulin protein occurred with PAC or IXA, and at high concentrations both drugs increased tubulin expression, by 15% and 28% for 100 nM PAC and IXA, respectively. All four drugs increased the acetylation of tubulin at low nanomolar concentrations; at 4 nM, tubulin acetylation was increased by 107% for PAC, 75% for IXA, 70% for ERB and 61% for VCR. However, VCR uniquely reduced tubulin acetylation at high concentrations (by 88% at 100 nM, possibly due to microtubule destabilization). The relative effects on tubulin expression and acetylation by the four drugs correlate well with their relative potencies on neurite retraction, indicating that changes in microtubule stability may play a significant role in drug-induced neurite retraction. These results indicate that the specific functional perturbations of 50B11 sensory neurons by microtubule- targeting anticancer drugs vary substantially from one drug to another. Experiments to further assess effects of these drugs on sensory neuron damage in relation to their effects on microtubule assembly dynamics are underway. This work is funded by Eisai Inc.

283

Novel method for examining the effects of beta tubulin mutations on drug resistance or sensitivity by creating isogenic U2OS cell lines expressing codon-customized mutant TUBB ORFs.

K. Sagane1, K. Ogawa-Mitsuhashi1, K. Takase1, K. Kubara1, K. L. Agarwala1, K. Tsukahara1; 1_{Eisai Product Creation Systems, Eisai Co., Ltd., Tsukuba, Japan}

Tubulin is one of the oldest, most highly validated targets for cancer chemotherapy. A large number of mutations in tubulin genes have been identified from cancer patients. However, the effects of such mutations on drug susceptibility have not been adequately characterized. This is because there is no appropriate assay system, due to high expression levels and isoform complexity of endogenous tubulins. To overcome this problem, we have established a new method of tubulin mutation analysis. Our system is comprised of 3 parts, 1) a codon-customized synthetic type 1 beta-tubulin (TUBB) ORF, which is resistant to siRNA, 2) siRNA targeting of the endogenous TUBB sequence, and 3) integration of mutant TUBB ORF-expression cassettes into the specific genomic locus of the host cells using the Jump-InTM TITM Gateway® Targeted Integration System (Invitrogen). Several drug-resistant mutations of the TUBB gene were tested in our system, and the effects of each TUBB mutation upon sensitivity to or resistance against tubulin inhibitors, such as paclitaxel, vinblastine, vincristine and eribulin, were successfully determined. In conclusion, our versatile assay system is useful for characterizing the effects of TUBB gene mutations on drug resistance or sensitivity.

284

Effects of the Chemotherapeutic Agent Taxol at Nanomolar Concentrations Observed by TIRF Microscopy.

B. D. Charlebois1, S. M. McCubbin1; 1Biomedical Engineering, University of Michigan, Ann Arbor, MI

Microtubules are intracellular polymers that assemble from heterodimeric αβ-tubulin subunits. Polymerizing microtubules exhibit dynamic instability, a GTP-hydrolysis-driven phenomenon in which they alternate abruptly between phases of growth and relatively rapid shortening. The kinetics of tubulin subunit exchange at the microtubule tip, which are crucial to processes such

as mitosis, are affected by the chemotherapeutic drug taxol, but the precise mechanism of action at therapeutic concentrations remains unclear.

We observe microtubules polymerized at 4.5 micromolar tubulin and at 1-480 nanomolar taxol using total internal reflection fluorescence microscopy (TIRFM), achieving improvements in spatial and temporal resolution of, respectively, 1 and 2 orders of magnitude compared to previous taxol studies. We measure microtubule length changes to within 25 nM at 8 Hz, and we detect changes in the structure of the microtubule tip. We find that taxol potently affects microtubule growth at concentrations as low as 10 nM. At these concentrations we observed abrupt changes in the growth rate not observed under control conditions, with phases lasting on the order of 1 – 2 minutes, suggesting multiple tip states and kinetic rate constants. In many of these cases, microtubules switched between a state of normal growth and a “paused” state, where the net growth rate is small or negligible. Further, we have found that 10 nM taxol almost completely suppresses rapid shortening, and beginning at 100 nM taxol, the tubulin on and off rates gradually decrease, though the microtubule net growth rate (excluding pauses in taxol- microtubules) remains constant.</p>

285

Fluorescent Probes for Carbonyl-Containing Proteins.

O. Dilek1, K. Mukherjee1, S. Bane1; 1Chemistry, State University of New York at Binghamton, Binghamton, NY

Carbonyl groups can be introduced into proteins deliberately for bioorthogonal labeling or by endogenous processes such as oxidative stress. Chemical probes for detection and labeling protein aldehydes are primarily hydrazides or 2,4-dinitrophenylhydrazine, but the kinetics of the labeling reactions are slow at neutral pH. Aromatic hydrazines and amines are better candidates for performing the labeling reactions under biocompatible conditions. We have synthesized fluorophores that possess aromatic amino and hydrazinyl groups and studied their reactive and photochemical properties with a model aldehyde-containing protein. Of particular interest are fluorescein, rosamine and BODIPY-based fluorophores in which the nitrogen is bonded to the pendant aromatic ring. In most cases, the parent amine or hydrazine is weakly fluorescent due to intramolecular photoinduced electron transfer from the nitrogen. Imine or hydrazone formation can remove the intramolecular quenching, and therefore the fluorescence can “turn on” as a result of covalent bond formation. The reversible imine bond can be hydrolyzed to regenerate the carbonyl or reduced to form a stable covalent bond. Applications of these fluorophores to protein and cell biochemistry vary with the specific fluorophore. For example, cell impermeable fluorophores can label oxidized proteins on the cell surface, while cell permeable probes can react with proteins on the interior of the cell. Some of these reactions occur specifically in live cells.

Centrosomes I

286

How Flies Build a Mitotic Centrosome

P. T. Conduit1, M. Pratt1, J. Baumbach1, J. W. Raff1; 1Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

Centrosomes are the main microtubule organising centres in many cells. They comprise a pair of centrioles, which can recruit a large number of proteins collectively known as the Pericentriolar Material (PCM). The PCM harbours the complexes that facilitate microtubule nucleation and large amounts of PCM are recruited during mitosis in a process known as centrosome maturation. Several proteins have been implicated in this process, but a clear understanding of how the mitotic PCM is assembled has remained elusive. Here we show that two PCM components, Cnn and DSpd-2, are constantly recruited into a zone of PCM immediately surrounding the centrioles by the centriolar protein Asl, and then move outwards to fill the rest of the PCM. Thus, they exhibit a ‘central to peripheral flow’. In contrast, all of the other PCM components we have analysed are recruited evenly throughout the PCM, suggesting that they bind to a preformed PCM ‘scaffold’. Mutant analysis in somatic cells shows that centrosomes can partially mature and nucleate MTs in the absence of either Cnn or DSpd-2. In the absence of both proteins, however, centrosome maturation and MT nucleation is completely abolished. We propose that Asl, Cnn and DSpd-2 cooperate to form a PCM scaffold around the centrioles onto which other PCM components can bind. Without this scaffold, other PCM components are unable to accumulate around the centrioles and the centrosomes cannot function to nucleate microtubules.

287

Quantitative Mass Spectrometry-based Proteomics of Human Centrosomes after Cullin- RING E3 ligase and Proteasome Inactivation.

K. Vanselow1, K. M. Larsen1, L. Jakobsen1, J. M. Schrøder1, K. B. Schou1, E. Lundberg2, J. S. Andersen1; 1Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense, Denmark, 2School of Biotechnology, Alba Nova University Center, Royal Institute of Technology, Stockholm, Sweden

Ubiquitin-ligases are key regulators of centrosome function during the cell cycle. The identity and substrates of these regulators are, however, not well established. To identify novel ubiquitin-ligases and substrates with a role in centrosome biology, we applied mass spectrometry-based proteomics to quantify proteins differentially associated with the centrosome before and after treatment of cells with the proteasome inhibitor bortezomib and the NEDD8-activating enzyme inhibitor MLN4924. Upon proteasome inhibition, we observed that a large number of proteins accumulate at the centrosome. This is consistent with previous reports where fluorescence microscopy experiments have revealed that proteins accumulate at the centrosome, most likely due to a failure of degrading poly-ubiquitylated proteins. We were able to quantify those changes for the majority of known centrosomal proteins and observed major differences in the degree of accumulation of proteins involved in distinct processes such as centriole duplication, elongation and cohesion. Inactivation of Cullin-RING E3 ligases by the NEDD8-activating enzyme inhibitor MLN4924 followed by quantitative proteomic analysis revealed more subtle and selective changes in stabilization of proteins at the centrosome. To identify novel proteins associated with centrosomes after proteasome inhibition, we applied our PCP-SILAC method (Jakobsen, EMBO J., 2011) to measure enrichment profiles of proteins in fractions collected after sucrose gradient centrifugation of centrosomes. Cluster analysis of

profiles for 3822 proteins revealed a distinct cluster with 150 known centrosomal proteins and 60 additional components representing novel candidate proteins and possibly ubiquitylated substrates associated with the centrosome. This group of proteins comprised several E3 ligases of interest for further analysis. Initial immunostaining experiments confirmed centrosome localization for 12 proteins not previously shown to be associated with the centrosome. Changes in the localization patterns indicated cell cycle dependent function for these proteins. 288

Dissecting the interactions of the centrosome.

B. J. Galletta1, C. J. Fagerstrom1, R. X. Guillen1, K. C. Slep2, N. M. Rusan1; 1Cell Biology and Physiology Center, National Institutes of Health, Bethesda, MD, 2Biology, University of North Carolina, Chapel Hill, NC

The major microtubule-organizing (MTOC) center of many eukaryotic cells is the centrosome, a