C ONFLICTOS SOCIALES DESARROLLADOS EN MÁS DE UN DEPARTAMENTO

94

TPX2 Regulates the Localization and Activity of Eg5 in the Mammalian Mitotic Spindle.

J. Titus1, N. Ma2, A. Gable3, J. Ross4, P. Wadsworth1; 1Biology, University of Massachusetts Amherst, 2Medical Oncology, Dana Farber Cancer Institute, Boston, MA, 3Virology, WuXi AppTec, Inc, Philadelphia, PA, 4Physics, University of Massachusetts, Amherst, MA

The assembly and function of the mitotic spindle requires the precise temporal and spatial regulation of numerous motor and non-motor proteins. In mammalian cells, the kinesin-5 family member, Eg5, cross-links and slides apart anti-parallel microtubules in vitro and is required for bipolar spindle formation in mitosis. Eg5 localizes to spindle microtubules and is enriched at spindle poles. In live cells, FRAP experiments show that at a population level, Eg5 is highly dynamic in all spindle regions, with a turnover half-time for fluorescence recovery of < 10 sec. Eg5 interacts with the Ran-regulated spindle-assembly factor, TPX2 and the proteins co-localize throughout mitosis. To determine if TPX2 regulates the localization and dynamics of Eg5, we examined spindle formation in cells expressing TPX2 lacking the Eg5 binding domain. In these cells, spindles were disorganized with multiple poles. The TPX2-Eg5 interaction was required for the formation of cold-stable kinetochore fibers and to localize the motor to spindle microtubules, but not spindle poles. FRAP showed that the interaction with Eg5 regulated TPX2 turnover, however Eg5 turnover was not changed in the absence of an interaction with TPX2. Microinjection of the Eg5 binding domain of TPX2 into metaphase cells resulted in spindle elongation, suggesting that TPX2 functions to inhibit Eg5 activity. To test this possibility, in vitro motility assays with purified proteins were performed. The results demonstrate that TPX2 reduces the velocity of Eg5-dependent microtubule gliding, inhibits microtubule sliding, and results in the accumulation of motors on microtubules. These results show that the localization and activity of Eg5 are regulated by interactions with the microtubule-associated protein TPX2.

95

The depolymerizing kinesins Kip3 (kinesin-8) and MCAK (kinesin-13) are catastrophe factors that destabilize microtubules by different mechanisms.

J. Howard1, M. Zanic1, M. Gardner2; 1Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany, 2Genetics, Cell Biology and Development, University of Minnesota

Microtubules are dynamic filaments whose plus ends alternate between periods of slow growth and rapid shortening as they explore intracellular space and move intracellular organelles. A key question is how regulatory proteins such as the depolymerizing kinesins modulate catastrophe, the conversion from growth to shortening. To study this process, we reconstituted microtubule dynamics in the absence and presence of the kinesin-8 Kip3 and the kinesin-13 MCAK. Surprisingly, we found that even in the absence of the kinesins, the microtubule catastrophe frequency depends on the length and age of the microtubule, indicating that catastrophe is a multistep process. Kip3 slowed microtubule growth in a length-dependent manner and increased the rate of accumulation of lattice destabilizing features that lead to catastrophe. In contrast, MCAK did not change the feature formation rate, but instead transformed catastrophe into a single step process. Thus, both kinesins are catastrophe factors, but differentially reshape the microtubule cytoskeleton.

96

MCAK activity controls microtubule dynamics and directed cell migration.

K. A. Myers1, C. M. Waterman1; 1Cell Biology and Physiology Center, NHLBI/NIH, Bethesda, MD

Cell polarization during directed cell migration is coordinated by (MT) growth toward the leading edge. Cells can locally control MT growth or shortening through the regulation of depolymerizing kinesins (MCAK), which catalyze the destabilization of MT plus ends, causing MTs to transition from growth to shortening (catastrophe). However, it is not known if MCAK regulation affects polarized MT growth dynamics that mediate cell migration. We hypothesized that MCAK is locally inhibited to establish preferential MT growth toward the leading edge to promote directed migration. To test this hypothesis, we performed high resolution imaging of fluorescently tagged EB3 that dynamically associates with growing MT plus-ends, coupled with automated image- based tracking of MT growth speeds and growth lifetimes (1/catastrophe frequency). We found that knockdown of MCAK (MCAK-KD) promoted increased MT growth lifetimes globally, abolishing the polarized differences in MT growth lifetime in the cell center and edge seen in control cells. As a result, directional cell migration was significantly reduced, suggesting that local regulation of MCAK activity is critical for establishing polarized MT growth and directing cell migration. To determine how local regulation of MCAK might be achieved, we investigated the role of Aurora-A kinase, which when activated by phosphorylation localizes to mitotic centrosomes and behaves as a phospho-inhibitor of MCAK depolymerase activity. Phospho- Aurora-A immunoprecipitation from cell extracts revealed interaction between MCAK and the phospho-active form of Aurora-A. Live-cell measurements of MT growth dynamics showed that overexpression of Aurora-A promotes long-lived MT growth similar to MCAK-KD, suggesting that Aurora-A regulates the depolymerase activity of cytoplasmic MCAK. To identify upstream activators of Aurora-A, we expressed either constitutively active- (CA-Rac1) or dominant negative-Rac1 (DN-Rac1) and measured changes in active Aurora-A via immunolabeling with a phospho-specific Aurora-A antibody. Compared to control, cells expressing CA-Rac1 displayed a 4-fold increase of phospho-active Aurora-A, while in cells expressing DN-Rac1, phospho- active Aurora-A levels were decreased by 3-fold. In addition, live-cell imaging revealed that like GFP-MCAK, GFP-Aurora-A tracks with a subset of growing MT plus-ends in CA-Rac1 cells, but

not in DN-Rac1 cells. Finally, while CA-Rac1 promoted fast and long-lived MT growth, this effect was lost by pharmacologic inhibition of Aurora-A. Together, these results suggest that Aurora-A- mediated regulation of MCAK depolymerase activity is achieved downstream of Rac1 signaling to elicit regional regulation of MT dynamics and thereby promote polarized MT growth and directional cell migration.

97

Axon injury triggers a microtubule-based pathway that protects dendrites.

L. Chen1, M. Stone1, J. Tao1, M. Rolls1; 1Department of Biochemistry and Molecular Biology, Pennsylvania State Univ, University Park, PA

In order for a single set of neurons to function for the entire lifetime of an animal, these cells need to be able to survive injury and repair damage. Using Drosophila dendritic arborization neurons, we have found that one unexpected response to axon injury is a global increase in the number of growing microtubules.

To understand the pathway that increases microtubule dynamics in response to injury, we performed a candidate screen of microtubule regulators. Reducing levels of microtubule nucleation by RNAi against the core nucleation protein gamma-tubulin 23C, or introducing one copy of a null allele of this gene, reduced the increase in microtubule dynamics in injured neurons but did not affect uninjured cells. Microtubule nucleation is thus required for the increase in number of growing microtubules after axon injury.

We hypothesized that the increased microtubule dynamics might be required for axon regeneration after injury. To test this hypothesis we assayed regeneration in backgrounds with reduced nucleation. No difference from control animals was found when gamma-tubulin 23C was targeted by RNAi. Thus the changes in microtubule dynamics induced by axon injury are likely to be required for an injury-response other than regeneration.

As microtubule dynamics increased in dendrites after a distant axon injury, we hypothesized that rebuilding of dendritic microtubules might act to protect dendrites from retraction or degeneration triggered by losing the axon. To test this hypothesis we made use of the finding that dendrites have an active degeneration program that clears them after they are severed from the cell body. To trigger increased microtubule dynamics we cut the axon 8 hours before dendrite severing. This resulted in a delay in dendrite degeneration. To determine whether the delay in dendrite degeneration was due to the increased microtubule dynamics, we introduced a single mutant copy of gamma-tubulin 23C or targeted the microtubule polymerase msps by RNAi. Both treatments blocked dendrite protection by prior axon injury confirming a role for microtubules in a dendrite protective pathway.

This study identifies a novel program that uses microtubules to protect dendrites after axon injury. This program likely plays an important role in maintaining intact circuits while axons attempt repair.

98

Structural Mechanism for Dynein Control by Lis1.

A. Roberts1,2, J. Huang1, A. Leschziner3, S. Reck-Peterson1; 1Department of Cell Biology, Harvard Medical School, Boston, MA, 2Faculty of Biological Sciences, University of Leeds, Leeds, United Kingdom, 3Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA

Cytoplasmic dynein, the large microtubule-based motor protein, is carefully controlled in cells, enabling it to be deployed to specific sites, collect cargoes, and transport them towards the microtubule minus end at set times. Dynein’s “engine” is evolved from ring-shaped AAA+ ATPases, and its microtubule-binding domain lies at the tip of a coiled-coil stalk, but how these elements might be acted upon to achieve control remains unknown. Here, using purified proteins from Saccharomyces cerevisiae, we dissect how cytoplasmic dynein is controlled by two of its ubiquitous regulators: Lis1/Pac1 and Nudel/Ndl1. By single-molecule microscopy, we find that Lis1 slows dynein velocity, prolongs its encounters with microtubules, and is tethered to dynein by Nudel. High-precision analysis shows that dynein-Lis1 undergoes frequent “anchored” cycles, during which ATP is consumed without the usual microtubule release and forward motion. Unexpectedly, the structural basis for these changes involves Lis1 binding at the interface between dynein’s ATPase domain and its microtubule-binding stalk. Lis1 is thus ideally situated to alter allosteric communication in cytoplasmic dynein, which could facilitate its control and function in cells. J. Huang and A. Roberts contributed equally to this work.

99

The crystal structure of tubulin tyrosine ligase offers insight into tubulin recognition.

A. Szyk1, A. Deaconescu2, G. Piszczek3, A. Roll-Mecak1,3; 1Cell Biology and Biophysics Unit, NINDS, National Institutes of Health, Bethesda, MD, 2Brandeis University, Waltham, MA,

3_{NHLBI, National Institutes of Health, Bethesda, MD}

Tubulin tyrosine ligase (TTL) adds a C-terminal Tyr to alpha-tubulin as part of a tyrosination/detyrosination cycle present in most eukaryotic cells. The C-terminal tyrosine in alpha-tubulin serves as an ON/OFF signal for the recruitment of microtubule dynamics regulators. TTL loss causes morphogenic abnormalities and is associated with aggressive cancer progression and poor prognosis. We present the first crystal structure of TTL, defining the structural scaffold upon which the diverse TTL-like (TTLL) family of tubulin-modifying enzymes is built. TTL uses a bipartite strategy to recognize tubulin. It engages the tubulin tail through low-affinity, high-specificity interactions, and co-opts what is otherwise a homo- oligomerization interface in structurally related enzymes to form a tight hetero-oligomeric complex with tubulin. Small-angle X-ray scattering and functional analyses reveal that TTL forms an elongated complex with the tubulin dimer and prevents incorporation of the dimer into microtubules by capping the tubulin longitudinal interface, thereby possibly modulating the partition of tubulin between its monomeric and polymeric forms.