Neurodegeneration mutations in dynactin impair -dependent nuclear migration

Jeffrey K. Moorea, David Septb, and John A. Coopera,1

aDepartment of Cell Biology and Physiology, and bDepartment of Biomedical Engineering and Center for Computational Biology, Washington University, Saint Louis, MO 63110

Edited by J. Richard McIntosh, University of Colorado, Boulder, CO, and approved January 30, 2009 (received for review October 27, 2008) Neurodegenerative disease in humans and mice can be caused by opportunity to identify functions that are linked to disease. In mutations affecting the motor dynein or its biochem- mouse models, a homozygous G59S mutation results in embryonic ical regulator, dynactin, a multiprotein complex required for dy- lethality, similar to complete loss of p150glued (6). Heterozygous nein function (1–4). A single amino acid change, G59S, in the G59S-mutant mice are viable, and adults exhibit cellular abnor- conserved cytoskeletal-associated glycine-rich (CAP-Gly) malities in lower motor neurons that are reminiscent of changes domain of the p150glued subunit of dynactin can cause motor seen with human diseases, including ALS (6–8). The G59S phe- neuron degeneration in humans and mice, which resembles ALS (2, notype appears to be a dominant-negative effect rather than the 5–8). The molecular mechanism by which G59S impairs the function result of haploinsufficiency, because animals bearing a single wild- of dynein is not understood. Also, the relevance of the CAP-Gly type gene for p150glued do not display these phenotypes (6). The domain for dynein motility has not been demonstrated in vivo. motor neurons in adult G59S-mutant mice accumulate large inclu- Here, we generate a mutant that is analogous to G59S in budding sions that contain p150glued, raising the possibility that the disease yeast, and show that this mutation produces a highly specific pathology could either be due to the toxic accumulation of mis- phenotype related to dynein function. The effect of the point folded p150glued, altered dynein function, or a combination of these mutation is identical to that of complete loss of the CAP-Gly effects (7, 8). In vitro, the G59S mutation disrupts the interaction domain. Our results demonstrate that the CAP-Gly domain has a of p150glued with and EB1; however, fibroblasts from critical role in the initiation and persistence of dynein-dependent G59S patients exhibit only a mild dynein-like phenotype in main- movement of the mitotic spindle and nucleus, but it is otherwise taining Golgi distribution (5). CELL BIOLOGY dispensable for dynein-based movement. The need for this func- The key questions at this point are what is the role of the tion appears to be context-dependent, and we speculate that CAP-Gly domain in dynein motility, and how does the G59S CAP-Gly activity may only be necessary when dynein needs to mutation impair that function? Here, we demonstrate that the overcome high force thresholds to produce movement. CAP-Gly domain of Nip100, the budding yeast homologue of p150glued, makes a specific contribution to dynein motility. The ͉ microtubules ͉ motility ͉ nucleus function of dynein in yeast is to draw the mitotic spindle and nucleus into the bud neck by pulling on cytoplasmic microtubules from sites he cytoplasmic dynein motor is an ancient ATPase that powers at the cell cortex (19–21). The CAP-Gly domain of Nip100 is Tdirectional motility along microtubule tracks. Cells use dynein well-conserved, and we show that a mutation analogous to G59S to organize the cytosol by manipulating the position of various abolishes the function of this domain. Mutations that disrupt the cargoes with respect to the microtubule , and the predicted binding interface or that completely remove the CAP-Gly mitotic spindle and the microtubule organizing centers (MTOCs) domain have phenotypes identical to those of G59S. The CAP-Gly with respect to regions of the cell cortex. The latter function is domain appears to be involved specifically in the initial movement particularly important during asymmetric cell divisions and cell of the spindle and nucleus into the bud neck. The specificity of the migration. phenotype is remarkable in that these CAP-Gly mutations have no Cytoplasmic dynein is a large multisubunit complex, and every effect in several other assays of dynein function in vivo. Most known function of cytoplasmic dynein requires a second multisub- notably, the CAP-Gly domain is completely dispensable for the unit complex, dynactin (9, 10). Dynactin includes a short -like dynein-dependent sliding of free microtubules along the cell cortex. filament composed of the actin-related protein, Arp1, overlaid by These data support a model in which the CAP-Gly domain of a shoulder-sidearm complex composed of the p24, dynamitin/p50, dynactin is important for dynein function in scenarios that require and p150glued subunits (11). The amino terminus of the p150glued maximal force production by the motor. subunit contains a cytoskeletal-associated protein glycine-rich Results (CAP-Gly) domain, which can bind to microtubules and the Residue G59 of human p150glued is conserved in CAP-Gly domains, microtubule-binding , CLIP-170 and EB1 (12–14). In vitro, and it corresponds to G45 of yeast Nip100 (Fig. 1A). Based on antibodies that bind to the amino-terminal region of p150glued crystal structures of human p150glued, G59 is embedded in a ␤ sheet, diminish the run length of dynein–dynactin in motility assays; and does not appear to participate in interactions with binding however, the manner in which this domain contributes to the partners (Fig. 1B; see refs. 13, 14). We created homology models for function of dynein–dynactin remains unclear (15, 16). In fact, recent the WT and G45S forms of Nip100 using Rosetta (22). These work has shown that loss of the CAP-Gly domain of p150glued does structures were similar to the CAP-Gly domain of human p150glued, not affect the transport of several cargoes along microtubules in S2 cells (17) or the organization of Golgi in HeLa cells

(18), suggesting that the CAP-Gly domain may not be required for Author contributions: J.K.M., D.S., and J.A.C. designed research; J.K.M. and D.S. performed all dynein–dynactin functions. research; J.K.M. contributed new reagents/analytic tools; J.K.M., D.S., and J.A.C. analyzed A point mutation in the CAP-Gly domain of human p150glued, data; and J.K.M. and J.A.C. wrote the paper. G59S, has been linked to a slowly progressive form of lower motor The authors declare no conflict of interest. neuron disease that resembles ALS (2). Although neurons use the This article is a PNAS Direct Submission. dynein motor for several distinct tasks, the connection between 1To whom correspondence should be addressed. E-mail: [email protected]. dynein dysfunction and neuronal degeneration is not well under- This article contains supporting information online at www.pnas.org/cgi/content/full/ stood. The specific nature of the G59S mutation provides an 0810828106/DCSupplemental.

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810828106 PNAS ͉ March 31, 2009 ͉ vol. 106 ͉ no. 13 ͉ 5147–5152 Downloaded by guest on October 2, 2021 Also, we excised codons 2-104 from the chromosomal NIP100 locus A p150 glued GHRGTVAYVGATLFATGKWVGVILDEAKGKNDGTVQGRKYFTCDEGH------GIFV  Nip100 EMKGRVKFIGETQFAKGIWYGIELDKPLGKNDGSANGIRYFDIDLKKANSNGGYYGLFC  to remove the entire CAP-Gly region of the protein. In this CLIP170 NKPGFIQFLGETQFAPGQWAGIVLDEPIGKNDGSVAGVRYFQCEPLK------GIFT  truncated protein, the first predicted coiled-coil is located at the Bik1 LGRGQLKYVGPVDTKAGMFAGVDLLANIGKNDGSFMGKKYFQTEYPQS------GLFI  amino terminus (Fig. S1). Mammalian p150glued contains a region B of basic residues between the CAP-Gly and coiled-coil domains, D which has been shown to bind microtubules and enhance the WT G45S∆CAP-GlyK54E,N55DK54A,N55AK54A N55A Nip100 processivity of dynein in vitro (24); however, Nip100 does not -myc appear to contain an analogous region. For each Nip100 mutant, expression was tested, and the level of soluble Nip100 was not Cdc3 decreased in any of the mutants, including the ⌬CAP-Gly truncate, for which the level was actually increased by 2-fold (Fig. 1D). E WT To determine whether and how the CAP-Gly domain might be dyn1∆ important for dynein–dynactin activity, we tested each mutant for nip100∆ several phenotypes indicative of loss of dynein function. First, we G45S used a simple assay, observing anaphase spindle position at a single 4 ∆CAP-Gly time-point in a population of asynchronous cells. The CAP-Gly C 3.5 WT G45S 3 K54E,N55D mutants were indistinguishable from wild-type cells and did not 2.5 K54A,N55A exhibit accumulation of spindles within the mother cell, a trait 2 K54A characteristic of dynein or dynactin null mutants (Fig. 1E). In 1.5

RMSF (Å) contrast, a larger truncation that removed the first coiled-coil 1 N55A

0.5 region of Nip100 along with the CAP-Gly domain was defective in 0 102030 0 % spindle position this assay, at a level comparable with that of nip100 and dyn1 null I67 I33 L69 F65 A39 A59 E25 K27 K31 E35 E47 P51 R29 D49 N55 R63 Q37 G41 G45S G53 G57 G61 W43 defect mutants (Fig. S1). In vertebrates, this coiled-coil region is required Fig. 1. The G59S mutation of human p150glued constructed in yeast. (A) for interaction with dynein (15). Sequence alignment of the CAP-Gly domains of human p150glued, yeast Next, we scored the position of preanaphase spindles in the Nip100, human CLIP170, and yeast Bik1. The conserved glycine at position 59 absence of Kar9, which helps position the spindle via an indepen- of p150glued is shown in black. Dashed lines indicate residues of human dent mechanism (25, 26). Loss of the CAP-Gly domain of Nip100 p150glued that are involved in interactions with binding partners, as deter- decreased the number of kar9⌬ cells in which the preanaphase mined by Weisbrich et al. (14). (B) Predicted structures of residues 25–83 for spindle was found in the bud (Fig. S1C), suggesting that the wild-type Nip100 and the Nip100-G45S mutant. The sequence of Nip100 was CAP-Gly domain is important for the translocation of short spin- glued threaded onto the structure of p150 (13) by using Rosetta, and both dles through the bud neck. Later in anaphase, the frequency of structures were analyzed by molecular dynamics simulations (see SI Materials and Methods). Regions that exhibited significantly altered dynamics in the mis-positioned spindles was normal, indicating that this defect G45S mutant are indicated in blue (G53-N55) and green (S58-G62). (C) Root- resolves over time, as the progresses (Fig. S1D). mean-square fluctuation plot for each residue of Nip100 (black) and Nip100- To examine dynein–dynactin function more carefully in the CAP- G45S (red), based on molecular dynamics simulations. The blue and green Gly mutants, we assayed the kinetics of spindle movement in synchro- regions here correspond to those in B.(D) Levels of soluble Nip100 in wild-type nized cells expressing GFP-. After release of cells from G1 and mutant strains. Mutations were introduced at the endogenous NIP100 arrest, we recorded time-lapse movies and measured the time until the locus, and each allele was tagged with the 13myc epitope. Endogenous Cdc3 movement of 1 spindle pole into the daughter cell. Cells expressing is shown as a loading control. Strain numbers: WT, yJC6056; nip100-G45S, either nip100-G45S or nip100⌬CAP-Gly required more time to move ⌬ yJC6300; nip100 CAP-Gly, yJC5811; nip100-K54E,N55D, yJC6301; nip100- the spindle across the bud neck than did wild-type cells (Fig. 2A). K54A,N55A, yJC6302; nip100-K54A, yJC6307; nip100-N55A, yJC6303. (E) An- aphase spindle position assay for dynein function. Percentage of cells with Therefore, mutations in the CAP-Gly domain appear to impair the abnormal spindle position is plotted for the mutant strains, with comparison efficiency of dynein function during spindle translocation. to WT, and nip100 and dyn1 null strains. Values are the mean of 5 sets of at To explore this hypothesis further, we developed a live-cell assay least 50 cells, and error bars are the SE of the mean. Strain numbers: WT, 5919; to monitor the movement of preanaphase spindles labeled with nip100⌬, yJC6047; dyn1⌬, yJC5603; nip100-G45S, yJC6295; nip100⌬CAP-Gly, GFP-tubulin. To eliminate the possibility that spindle pole move- yJC5795; nip100-K54E,N55D, yJC6296; nip100-K54A,N55A, yJC6297; nip100- ment could be influenced by spindle elongation, we arrested cells K54A, yJC6298; nip100-N55A, yJC6299. in S phase with short spindles by treatment with hydroxyurea (HU). In wild-type cells, we observed the spindle to move in both ␤ directions through the bud neck; these movements coincided with with G/S45 similarly positioned within the second strand of the lateral sliding of a cytoplasmic microtubule along the cell cortex sheet. We used molecular dynamics simulations to investigate the (Movie S1). These events require dynein–dynactin function; effect of changing Gly 45 to Ser (SI Materials and Methods). The nip100⌬ null mutants displayed frequent contacts between micro- hydroxyl group of the serine allowed the residue to make hydrogen tubule ends and the cell cortex, but no spindle movements or ␤ bonding interactions, disrupting the first 2 strands of the sheet, in microtubule sliding events occurred (Table S2 and Movie S2). A particular the strands from K31 to E35 and from I42 to S45. This kar9⌬ mutation dramatically enhanced dynein-dependent spindle loss of secondary structure elicited conformational changes movements in HU-arrested cells, consistent with previous findings ␤ throughout the sheet, and altered the dynamics of the regions (27). In this background, short preanaphase spindles oscillated between the strands (Fig. 1C). These regions of p150glued are known across the bud neck, moving completely between the mother and to be involved in protein–protein interactions, from structural bud (Table S2 and Movie S3). studies (14), so we hypothesized that the mutation might impair the To examine the contribution of the CAP-Gly domain of Nip100 functional interactions of the CAP-Gly domain in cells. in this setting, we collected timelapse movies of GFP-labeled To test the consequences of G45S in cells, we introduced the microtubules for each CAP-Gly mutation combined with kar9⌬. mutation into haploid yeast at the chromosomal NIP100 locus such The number of cells, in which the spindle moved from the mother that the mutant allele provided the only source of Nip100 (Table cell compartment through the neck, was significantly decreased in S1). For comparison, we mutated residues K54 and N55, which each CAP-Gly mutant (Fig. 2B; Movie S4). However, in CAP-Gly correspond to K68 and N69 in mammalian p150glued, and have been mutant cells where the spindle did move through the neck, the shown to be important for the CAP-Gly domain to interact with the frequency of transits of the spindle, back-and-forth through the EEY/F motifs found in ␣-tubulin, EB1, and CLIP170 (13, 14, 23). neck, was similar to that of wild-type cells (Fig. S2).

5148 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810828106 Moore et al. Downloaded by guest on October 2, 2021 A * * B 100 120 * 80 * 80 60 * * * 40 * * 40 20 cross the neck 0 % spindles that 0 ∆min to neck cross wt nip100 nip100 dyn1∆ NIP100: wt G45S ∆CAP-Gly K54E K54A K54A N55A G45S ∆CAP-Gly N55D N55A C

NIP100 kar9∆ 1µm CAP-Gly kar9∆ 0:00 2:00 4:00 6:00 8:00 10:00 12:00 0:001µm 2:00 4:00 6:00 8:00 10:00 12:00 nip100∆

14:00 16:00 18:00 20:00 22:00 24:00 14:00 16:00 18:00 20:00 22:00 24:00

D 4 4 2 NIP100 kar9∆ 2 nip100∆CAP-Gly kar9∆ 0 0 (µm) -2 (µm) -2 -4 -4 0 4 8 12162024 0 4 8 12162024 distance from neck time (min) distance from neck time (min)

Fig. 2. The CAP-Gly region of p150Glued/Nip100 contributes to dynein-dependent spindle movements. (A) Time elapsed between bud emergence and 1 spindle pole crossing the bud neck. Cells were arrested in G1 with 0.6 ␮M ␣factor, released into new media, and mounted on agarose pads for microscopy. Confocal images of GFP-tubulin were captured at 45-s intervals through a Z series of 7 planes separated by 0.8 ␮m. Asterisks indicate statistical significance (P Ͻ 0.05; CELL BIOLOGY compared with WT) as determined by t test. Strains: WT, yJC5920; nip100-G45S, yJC6446; nip100⌬CAP-Gly, yJC5798; dyn1⌬, yJC5603. (B) Percentage of HU-arrested kar9⌬ mutant cells in which the preanaphase spindle crosses the bud neck over the course of a 10-min timelapse movie. Neck crossing events were defined as at least 1 pole crossing the plane of the bud neck within a unidirectional movement. At least 90 cells were scored for each strain. Error bars are the SE of proportion. Asterisks indicate statistical significance (P Ͻ 0.05; compared with WT) as determined by Fisher’s exact test. Strain numbers: WT, yJC5802; nip100-G45S, yJC6308 and 6309; nip100⌬CAP-Gly, yJC5803; nip100-K54E,N55D, yJC6310; nip100-K54A,N55A, yJC6311; nip100-K54A, yJC6329; nip100-N55A, yJC6312. (C) Timelapse images of GFP-labeled microtubules in kar9⌬ mutants arrested with HU. Each image is a composite of 9 planes separated by 0.5 ␮m. Stacks were captured at 15-s intervals on a confocal microscope. The arrow marks the spindle pole that is tracked in D. Strains: WT, yJC5802; nip100⌬CAP-Gly, yJC5803. (D) The distance between one end of the spindle and the bud neck was plotted with respect to the long axis of division. The position of the spindle poles indicated in C was determined for each time point, and distances were calculated by using ImageJ.

To quantitate spindle movement, we tracked the position of Dynein-dependent spindle movements depend on microtubule slid- spindle poles over time (Fig. 2 C and D; for histograms of complete ing along the cell cortex; therefore, we examined microtubule-cortex datasets from these analyses, see SI Materials and Methods). Com- interactions in the timelapse movies. The CAP-Gly mutants all exhib- pared with wild-type Nip100, CAP-Gly mutants tended to produce ited a decrease in the frequency of microtubule sliding events (Table 1). shorter displacements (Table 1). In these abbreviated movements, Conversely, the frequency of unproductive microtubule-cortex interac- spindles often approached the neck, but did not pass through it, or tions, in which the microtubule plus end contacted the cortex, but did they made brief excursions into the neck. The instantaneous not transition into sliding, was increased (Table 1). Together, these data velocities observed between individual frames for CAP-Gly mu- indicate that the CAP-Gly domain of Nip100 promotes the initiation tants were similar to those observed for wild-type Nip100 (mean ϭ and persistence of the dynein-dependent microtubule sliding events that 2.6 Ϯ 0.1 ␮m/min; see Table 1). power spindle translocation through the bud neck.

Table 1. Spindle movements and microtubule-cortex interactions in CAP-Gly mutants Spindle movements Microtubule-cortex interactions

Neck crosses, Displacement, Velocity, Sliding events, Unproductive Sliding Strain minϪ1 ␮m ␮m⅐minϪ1 minϪ1 contacts, minϪ1 ratio

NIP100 kar9⌬ 0.23 Ϯ 0.01 (203) 3.4 Ϯ 0.2 (25) 2.6 Ϯ 0.1 0.30 Ϯ 0.02 (44) 0.44 Ϯ 0.03 0.41 Ϯ 0.02 nip100-G45S kar9⌬ 0.09 ؎ 0.01 (210) 1.8 ؎ 0.1 (19) 2.4 Ϯ 0.1 0.14 ؎ 0.03 (20) 1.02 ؎ 0.08 0.13 ؎ 0.02 nip100⌬CAP-Gly kar9⌬ 0.10 ؎ 0.01 (301) 1.9 ؎ 0.2 (28) 2.4 Ϯ 0.1 0.11 ؎ 0.01 (44) 0.79 ؎ 0.04 0.13 ؎ 0.02 nip100-K54E,N55D kar9⌬ 0.11 ؎ 0.01 (170) 1.9 ؎ 0.2 (20) 2.3 ؎ 0.1 0.15 ؎ 0.03 (20) 1.05 ؎ 0.11 0.14 ؎ 0.03 nip100-K54A,N55A kar9⌬ 0.09 ؎ 0.01 (141) 2.2 ؎ 0.2 (20) 2.5 Ϯ 0.1 0.13 ؎ 0.02 (20) 0.81 ؎ 0.07 0.15 ؎ 0.03 nip100-K54A kar9⌬ 0.11 ؎ 0.01 (133) 3 Ϯ 0.2 (19) 2.8 Ϯ 0.1 0.11 ؎ 0.03 (20) 0.78 ؎ 0.07 0.12 ؎ 0.03 nip100-N55A kar9⌬ 0.08 ؎ 0.01 (127) 2.0 ؎ 0.2 (20) 2.2 ؎ 0.1 0.12 ؎ 0.03 (20) 0.73 ؎ 0.08 0.15 ؎ 0.04 NIP100 bim1⌬ kar9⌬ 0.05 ؎ 0.01 (148) 1.8 ؎ 0.3 (24) 2.1 ؎ 0.1 0.08 ؎ 0.02 (20) 0.36 Ϯ 0.04 0.12 ؎ 0.38 nip100⌬CAP-Gly bim1⌬ kar9⌬ 0.06 ؎ 0.01 (169) 2.0 ؎ 0.3 (18) 2.6 Ϯ 0.1 0.06 ؎ 0.02 (20) 0.44 Ϯ 0.04 0.17 ؎ 0.45

Data were collected from 10-min movies of cells arrested in HU (see Materials and Methods). Mean values Ϯ SEM are shown. Boldface indicates statistical significance (P Ͻ 0.05) compared with NIP100 kar9⌬ , determined by t test. Numbers in parentheses are the number of cells scored. Strain numbers: WT, yJC5802; nip100-G45S, yJC6308 and 6309; nip100⌬ CAP-Gly, yJC5803; nip100-K54E,N55D, yJC6310; nip100-K54A,N55A, yJC6311; nip100-K54A, yJC6329; nip100-N55A, yJC6312; nip100⌬ CAP-Gly bim1⌬ , yJC6596 and 6597; bim1⌬ , yJC6599.

Moore et al. PNAS ͉ March 31, 2009 ͉ vol. 106 ͉ no. 13 ͉ 5149 Downloaded by guest on October 2, 2021 A tubulin Nup133 B tubulin Nup133

C D initial oscillating bni1∆ initial bni1∆ tubulin

Nup133 0123 x instantaneous velocity (µm • min -1 )

Fig. 3. Nuclear envelope position during spindle movements. (A) The nuclear envelope enters the neck with the spindle. (B) The outline of the nuclear envelope remains in the neck and moves rather little, whereas the spindle moves back-and-forth across the neck. Timelapse images of GFP-labeled microtubules and RFP-labeled Nup133 in kar9⌬ mutants arrested with HU. Images are composites of Z projections from 9 confocal sections at 0.5-␮m increments captured every 20 s. Strain: yJC6288. (Scale bar, 1 ␮m.) (C) Timelapse images of GFP-labeled microtubules and RFP-labeled Nup133 in bni1⌬kar9⌬ mutants arrested with HU. Images were collected as described in A and B. Strain: yJC6289. (D) Mean instantaneous velocity during initial spindle movement into the neck and subsequent spindle oscillations in kar9⌬ cells, and initial spindle movement in bni1⌬kar9⌬ cells. Initial movements began with the entire nucleus located within the mother cell and proceeded to bring at least 1 spindle pole across the bud neck. At least 25 cells were analyzed for each category. Error bars are the SE of the mean.

Mitosis in yeast is closed, meaning that the nuclear envelope bules pull out of the spindle pole body (SPB) spontaneously as the does not disassemble. Electron micrographs reveal that the preanaphase spindle began to move (Fig. S3 and Movie S8). When diameter of the nucleus within the mother cell is far greater than a microtubule pulled out, the spindle snapped back into the mother, the diameter of the bud neck, and that the nuclear space is and the detached microtubule slid along the bud cortex at high greatly constricted during passage through the neck (28). To speed (Fig. S3). investigate how constriction of the nucleus might influence To allow for a quantitative analysis of free microtubule sliding, we spindle movement, we labeled the nuclear pore complex with enhanced the frequency of cytoplasmic microtubule detachment Nup133-RFP, and captured 2-color timelapse movies of HU- from the SPB by introducing a cnm67⌬ mutation, which destabilizes arrested kar9⌬ cells that also expressed GFP-tubulin. In these the cytosolic face of the SPB (30). In this background, the number movies, the initial movement of the nucleus into the bud neck of free cytoplasmic microtubules in mitotic cells was Ͼ10-fold coincided with a bud-directed microtubule-sliding event that higher than in CNM67 cells, whether cells expressed full-length or brought the spindle to the proximal edge of the nucleus (Fig. 3A). ⌬CAP-Gly Nip100 (Table S3). As expected, free microtubules in The nuclear membrane appeared to squeeze through the neck dyn1⌬ cnm67⌬ or nip100⌬ cnm67⌬ double mutants exhibited and, then, to expand as it entered the bud. After this point, the end-on contacts with the cell cortex, but did not slide. In these nucleus adopted a bilobed shape and remained lodged across the double mutants, the number of detached microtubules was de- bud neck. In cells where the spindle translocated back-and-forth creased compared with cnm67⌬ single mutants, indicating that the across the neck, the outline of the nuclear envelope changed very force generated by dynein contributes to detaching the microtubule little, displaying only transient disfigurations when one end of the from the SPB. spindle collided with it (Fig. 3B; Movie S5 and Movie S6). Thus, To measure the movement of free sliding microtubules, we first the deformation of the nuclear envelope and volume is coupled collected 2-color movies of cnm67⌬ cells expressing GFP-tubulin with the initial movement of the nucleus and spindle into the bud and RFP-Spc97, a component of the ␥-tubulin complex (31). The neck, and not subsequent spindle oscillations. In HU-arrested persistent association of Spc97 at 1 end of the free sliding micro- Nip100⌬CAP-Gly cells in which the spindle failed to enter the tubule confirmed the minus end-directed polarity of all sliding bud neck, we observed that the entire nucleus remained within events, and demonstrated that the minus end was stabilized by the the mother compartment (37/40 cells; Movie S7). ␥-tubulin complex (Fig. S4A, Movie S9, and Movie S10). We To compare these movements, we measured the speed of mitotic measured the movement of free microtubules by tracking the spindle translocation during initial nuclear movement into the neck position of their minus ends. During individual sliding events, we and subsequent spindle oscillations in these movies. The spindle observed changes in speed and occasional pauses, but no reversals entered the neck with a mean instantaneous speed of 1.7 Ϯ 0.1 of direction (Fig. S4B, Movie S11, and Movie S12). The distribution ␮m/min, which was significantly less than the speed of spindle of instantaneous velocities was similar for cells expressing full- movements that occurred after the spindle was positioned across length Nip100 (6.4 Ϯ 0.2 ␮m/min) and the ⌬CAP-Gly truncate the neck (3.0 Ϯ 0.1 ␮m/min; P Ͻ 0.01; see Fig. 3D). To test whether (6.2 Ϯ 0.2 ␮m/min) (Fig. S4C; P ϭ 0.56). These speeds are similar the constriction of the nucleus retards spindle movement into the to the microtubule gliding speeds observed in vitro for purified neck, we examined spindle movement in cells with wide necks; dynein (5.4 Ϯ 1.6 ␮m/min; see ref. 32). Also, the duration of bni1⌬ mutants have wide necks, thought to arise from defective individual sliding events was not affected by loss of the CAP-Gly septin organization (29). In bni1⌬ cells, the nuclear envelope was domain (Fig. S4D; P ϭ 0.16). wider at the neck, and spindles moved into the neck at greater To examine the character of these movements more closely, we speeds (2.2 Ϯ 0.1 ␮m/min; P Ͻ 0.01; see Fig. 3 C and D). This result separated each sliding event into segments in which the speed did supports the idea that constriction of the nucleus antagonizes not change. We found that segments of longer duration were dynein-induced movement of the spindle into the neck. associated with slower speeds, whereas the speeds of shorter To test the role of the CAP-Gly domain in dynein-dependent segments were greater and more varied (Fig. S4E). The results for motility further, we examined cases where movement of the cyto- this analysis were similar for full-length Nip100 and the ⌬CAP-Gly plasmic microtubule was uncoupled from movement of the nucleus. mutant. Overall, the CAP-Gly domain of Nip100 does not appear In wild-type cells, we occasionally observed cytoplasmic microtu- to influence the sliding of free microtubules along the cell cortex,

5150 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810828106 Moore et al. Downloaded by guest on October 2, 2021 which suggests that the role of the CAP-Gly domain in dynein kar9⌬ double mutant cells during HU arrest. Spindle movements motility is context-dependent. were rare in the presence of the bim1⌬ mutation (Table 1). This We evaluated several molecular mechanisms by which the CAP- effect may be partly due to suppression of microtubule dynamics, Gly domain of p150glued/Nip100 might enhance the dynein motility. and subsequent reduction in the frequency of total microtubule Dynactin is required for the localization of dynein to the cell cortex, contacts with the cortex (Table 1; see ref. 19). However, we found which is presumably necessary for microtubule sliding (33). To test that the ratio of microtubule-cortex interactions that transitioned whether the CAP-Gly domain contributes to this function, we into sliding events was low in these cells, similar to the level examined the localization of dynein heavy chain (Dyn1–3GFP, observed for the CAP-Gly mutants (Table 1). When sliding events expressed from the endogenous locus) in asynchronous and HU- were initiated in bim1⌬ kar9⌬ mutants, they typically produced arrested cells, expressing full-length Nip100 or the ⌬CAP-Gly short displacements of the spindle (Table 1). To determine whether truncate. Loss of the CAP-Gly domain did not decrease the number the CAP-Gly domain was still contributing to dynein function in of cortical dynein foci in either case (Fig. S5). In fact, the number bim1⌬ mutants, we analyzed triple mutants expressing of cortical dynein foci was slightly increased in Nip100⌬CAP-Gly Nip100⌬CAP-Gly in the bim1⌬ kar9⌬ background. These cells cells, by a statistically significant amount (P Ͻ 0.01). We also exhibited phenotypes that were nearly identical to those observed quantitated the fluorescence intensity of individual foci, and found for bim1⌬ kar9⌬ (Table 1). This result indicates that the loss of Bim1 that the amount of dynein per cortical focus was not altered (Fig. is epistatic to the disruption of the CAP-Gly domain of Nip100, S5; P ϭ 0.84). Therefore, the diminution of dynein function in consistent with the hypothesis that the CAP-Gly domain binds to CAP-Gly mutants was not due to defective anchoring of dynein to Bim1. the cortex. As an alternative, we considered whether decreased nuclear Discussion movement could be due to poor dynein–dynactin function at the In this study, we report that a mutation in the CAP-Gly domain of SPB. Dynein–dynactin has been shown to organize microtubule the dynactin complex, which can cause motor neuron degeneration minus-ends at the MTOC in other organisms (17, 34, 35). If the in humans, abolishes the dynein-based function of this domain in CAP-Gly domain were required for anchoring cytoplasmic micro- yeast. Although mutation of the CAP-Gly domain did not abrogate tubules at the yeast SPB, then an increased frequency of microtu- dynein–dynactin motility, we found that these mutants displayed bule detachment in CAP-Gly mutants could account for the failure defects in specific settings. CAP-Gly function was important for the to pull the nucleus through the neck. To the contrary, we found that initiation and persistence of dynein-dependent microtubule sliding,

the frequency of microtubule detachment during spindle move- which is directly coupled to the initial movement of the mitotic CELL BIOLOGY ments was less in CAP-Gly mutant cells (Table S3). Also, loss of the spindle and nucleus into the bud neck. Once the nucleus was CAP-Gly domain did not alter the number or length of cytoplasmic positioned across the neck, loss of CAP-Gly function had only a microtubules at any phase of the cell cycle (Table S4). Although mild effect on spindle movement. Also, the sliding of free micro- these data argue that the CAP-Gly domain is dispensable for tubules, ones that had detached from the spindle, along the cortex microtubule organization at the SPB, one cannot completely ex- was not affected by loss of the CAP-Gly domain. clude a role for the CAP-Gly domain at this site. Detachment of How might the movement of the spindle and nucleus into the bud cytoplasmic microtubules during spindle translocation appears to neck be distinct from dynein motility in other contexts? Several lines be caused by dynein pulling from the cell cortex (Movies S1–S8). of evidence indicate that this movement introduces a load burden Thus, it is possible that diminished activity of cortical dynein in the that antagonizes dynein motility. First, we observed that the velocity CAP-Gly mutants may mask a defect at the SPB. of dynein-dependent microtubule sliding was significantly reduced Last, we asked whether the CAP-Gly domain was necessary for when coupled with movement of the nucleus (Fig. 3D). Second, by the association of Nip100 with microtubules in the cell. First, we monitoring the morphology of the nuclear envelope during spindle examined Nip100 localization in cells with a 3X-GFP tandem fusion movements, we found that when the spindle was initially pulled into at the C terminus of Nip100, expressed from the chromosomal the neck, the nuclear envelope deformed, causing a substantial locus. In haploid cells expressing Nip100-⌬CAP-Gly-3GFP, the constriction of the nuclear volume as it extended into the bud (Fig. fluorescence distribution along cytoplasmic microtubules was qual- 3A). Once the nucleus was positioned across the neck, subsequent itatively similar to that of full-length Nip100–3GFP (Fig. S6). microtubule sliding events produced spindle movements that did Quantitation of the fluorescence intensity at microtubule plus ends not require a change in nuclear shape. Thus, nuclear constriction revealed similar values for the truncated and full-length versions of only occurs during the initial movement into the bud. Third, we Nip100, 117 Ϯ 16 and 85 Ϯ 18 (mean Ϯ SEM), respectively, in cells found that lessening the constriction of the nucleus, by enlarging the with short bipolar spindles. This result is consistent with our bud neck, increased the velocity of spindle and nuclear movement previous finding that the localization of Nip100 to microtubule ends into the neck. This velocity dependence is reminiscent of results is mediated by dynein (33). Next, we examined the microtubule- from in vitro assays of dynein motility, in which the rate at which pelleting behavior of Nip100 and Nip100⌬CAP-Gly from yeast cell the dynein motor moves along a microtubule is inversely propor- lysates. Truncated Nip100 pelleted with microtubules similar to tional to the magnitude of load burden applied to the motor by full-length Nip100 (Fig. S6). The CAP-Gly domain alone exhibited optical tweezers (37, 38). Thus, our results in cells suggest that little to no microtubule pelleting, indicating that the microtubule- attempting to move the nucleus into the neck imposes a greater load pelleting ability of Nip100 from the cell extract probably depends burden on dynein–dynactin than do subsequent spindle oscillations on other interactions with binding partners. Thus, the CAP-Gly or free microtubule sliding. domain of Nip100 is not the primary link between dynactin and The force required to move the nucleus may antagonize the microtubules. initiation and persistence of microtubule sliding events by disrupt- We considered the possibility that the CAP-Gly domain of ing either the interaction between dynein–dynactin and its cortical Nip100 may indirectly link dynein–dynactin to microtubules by receptor, or the processive association of dynein–dynactin with the interaction with CLIP-170 or EB1. The CLIP-170 homologue, Bik1, microtubule substrate. In our experiments, the CAP-Gly domain participates in targeting dynein–dynactin to microtubule plus ends, was not necessary for the association of dynein with the cortex (Fig. and disruption of BIK1 leads to a complete loss of dynein function S5). It is more likely that the CAP-Gly domain mediates association (Fig. S1; see refs. 33 and 36). This phenotype precludes the analysis with microtubules. Although our results provide no evidence of a of bik1⌬ null mutants in our spindle movement assay. The EB1 direct interaction between the Nip100 CAP-Gly domain and tubu- homologue, Bim1, is not necessary for dynein function. Therefore, lin, we found that loss of EB1/Bim1 elicited defects in microtubule we examined the movement of GFP-labeled spindles in bim1⌬ sliding that were similar to CAP-Gly mutants. Also, double mutants

Moore et al. PNAS ͉ March 31, 2009 ͉ vol. 106 ͉ no. 13 ͉ 5151 Downloaded by guest on October 2, 2021 that lacked both the CAP-Gly domain and EB1/Bim1 were nearly transport appears to be unaffected in the G59S heterozygous adult identical to the EB1/Bim1 null mutation alone. These data are mice, indicating that not all functions of dynein–dynactin are consistent with the notion that EB1/Bim1 is necessary for the impaired (7). Our results suggest that the G59S mutation may function of the CAP-Gly domain. It will important to further test impair the function of dynein under load, such as during the the roles of EB1/Bim1 and CLIP-170/Bik1 in this mechanism by reorganization of the microtubule cytoskeleton, and we speculate generating mutations that specifically disrupt the EEY/F motifs in that this defect may disrupt the homeostasis of adult neurons. In the each protein. future, our system may provide a framework for rapid testing of We hypothesize that the role of the CAP-Gly domain of dynactin protein-expression or small-molecule therapies aimed at restoring is to enhance the ability of dynein to produce movement under load dynein–dynactin function in these individuals. by helping the dynein–dynactin complex resist detachment from the microtubule and maintain directional motility. Our model is con- Materials and Methods sistent with the observation that dynactin increases the run length Chemicals and reagents were from Fisher Scientific and Sigma-Aldrich, unless of dynein in in vitro motility assays, and informed by results from stated otherwise. optical trap experiments, where purified dynein has been shown to detach or drift backwards on microtubules when subjected to Yeast Strains and Manipulation. General yeast manipulation, media, and trans- formation were performed by standard methods (42). Strains are listed in Table increasing load (15, 16, 37, 38). Based on this model, we predict that S1. To generate mutants of Nip100 expressed from the endogenous chromo- the CAP-Gly domain of dynactin provides a module to tune the somal locus, we used a site-specific genomic mutagenesis strategy (43). Further dynein motor for tasks that require greater force generation in details are provided in SI Materials and Methods. other cells, including neurons. In the context of adult motor neurons, what might be the function Microscopy and Data Analysis. Timelapse Z series images of spindle oscillations of the CAP-Gly domain of p150glued? Dynein–dynactin is required and free microtubule sliding were captured on an Olympus Bmax-60F microscope for retrograde transport of through the axon, and it is equipped with a 1.35NA 100ϫ UPlanApo objective, spinning disc Confocal Scan- important for the structure of the neuron. In Drosophila adults, ner Unit (CSU10), Picarro Cyan laser (488 nm), and a Stanford Photonics XR- dynein–dynactin stabilizes microtubule networks at neuromuscular Mega10 ICCD camera, by using QED software (Media Cybernetics). Image analysis junctions (39). During development in mammals, dynein–dynactin was preformed by using ImageJ. Two-color imaging of GFP-tubulin, and RFP- Nup133 was also carried out on this microscope, by using Picarro Cyan and Cobolt promotes axonal outgrowth by organizing microtubule networks in Jive (561 nm) lasers. the growth cone, and it positions the nucleus in migrating neuronal precursors (40, 41). To our knowledge, it is not known whether ACKNOWLEDGMENTS. We thank Drs. Brian Galletta, Jun Li, and Mark Longtine similar roles are needed to maintain the structure of adult motor for advice and suggestions, and Dr. Elmar Schiebel (Universitat Heidelberg, neurons. Transgenic mice homozygous for the p150glued G59S Heidelberg) for the gift of the mCherry-tubulin construct. This work was sup- ported by National Institutes of Health (NIH) Grants GM47337 (to J.A.C.) and mutation die during development, and heterozygotes slowly accu- GM67246 (to D.S.). J.K.M. was supported by a postdoctoral fellowship from the mulate defects in motor neuron morphology, Molecular Oncology Program of the Siteman Cancer Center at Washington organization, and neuromuscular junction structure (6–8). Axonal University, funded by NIH Grant T-32-CA113275.

1. Hafezparast M, et al. (2003) Mutations in dynein link motor neuron degeneration to 24. Culver-Hanlon TL, et al. (2006) A microtubule-binding domain in dynactin increases defects in retrograde transport. Science 300:808–812. dynein processivity by skating along microtubules. Nat Cell Biol 8:264–270. 2. Puls I, et al. (2003) Mutant dynactin in motor neuron disease. Nat Genet 33:455–456. 25. Miller RK, Rose MD (1998) Kar9p is a novel cortical protein required for cytoplasmic 3. Munch C, et al. (2004) Point mutations of the p150 subunit of dynactin (DCTN1) gene microtubule orientation in yeast. J Cell Biol 140:377–390. in ALS. Neurology 63:724–726. 26. Hwang E, Kusch J, Barral Y, Huffaker TC (2003) Spindle orientation in Saccharomyces 4. Munch C, et al. (2005) Heterozygous R1101K mutation of the DCTN1 gene in a family cerevisiae depends on the transport of microtubule ends along polarized actin cables. with ALS and FTD. Ann Neurol 58:777–780. J Cell Biol 161:483–488. 5. Levy JR, et al. (2006) A motor neuron disease-associated mutation in p150Glued 27. Yeh E, et al. (2000) Dynamic positioning of mitotic spindles in yeast: Role of microtu- perturbs dynactin function and induces protein aggregation. J Cell Biol 172:733–745. bule motors and cortical determinants. Mol Biol Cell 11:3949–3961. 6. Lai C, et al. (2007) The G59S mutation in p150(glued) causes dysfunction of dynactin in 28. Byers B, Goetsch L (1975) Behavior of spindles and spindle plaques in the cell cycle and mice. J Neurosci 27:13982–13990. conjugation of Saccharomyces cerevisiae. J Bacteriol 124:511–523. 7. Chevalier-Larsen ES, Wallace KE, Pennise CR, Holzbaur EL (2008) Lysosomal prolifera- 29. Gladfelter AS, Kozubowski L, Zyla TR, Lew DJ (2005) Interplay between septin orga- tion and distal degeneration in motor neurons expressing the G59S mutation in the nization, cell cycle and cell shape in yeast. J Cell Sci 118:1617–1628. p150Glued subunit of dynactin. Hum Mol Genet 17:1946–1955. 8. Laird FM, et al. (2008) Motor neuron disease occurring in a mutant dynactin mouse 30. Hoepfner D, Brachat A, Philippsen P (2000) Time-lapse video microscopy analysis model is characterized by defects in vesicular trafficking. J Neurosci 28:1997–2005. reveals astral microtubule detachment in the yeast spindle pole mutant cnm67. Mol 9. Gill SR, et al. (1991) Dynactin, a conserved, ubiquitously expressed component of an Biol Cell 11:1197–1211. activator of vesicle motility mediated by cytoplasmic dynein. J Cell Biol 115:1639–1650. 31. Kollman JM, et al. (2008) The Structure of the {gamma}-Tubulin Small Complex: 10. Schroer TA, Sheetz MP (1991) Two activators of microtubule-based vesicle transport. Implications of Its Architecture and Flexibility for . Mol Biol Cell J Cell Biol 115:1309–1318. 19:207–215. 11. Schroer TA (2004) Annual Review of Cell and Developmental Biology. Dynactin Annu 32. Reck-Peterson SL, et al. (2006) Single-molecule analysis of dynein processivity and Rev Cell Dev Biol 20:759–779. stepping behavior. Cell 126:335–348. 12. Waterman-Storer CM, Karki S, Holzbaur EL (1995) The p150Glued component of the 33. Moore JK, Li J, Cooper JA (2008) Dynactin function in mitotic spindle positioning. dynactin complex binds to both microtubules and the actin-related protein centractin Traffic 9:510–527. (Arp-1). Proc Natl Acad Sci USA 92:1634–1638. 34. Quintyne NJ, et al. (1999) Dynactin is required for microtubule anchoring at centro- 13. Honnappa S, et al. (2006) Key interaction modes of dynamic ϩTIP networks. Mol Cell somes. J Cell Biol 147:321–334. 23:663–671. 35. Burakov A, et al. (2008) Cytoplasmic dynein is involved in the retention of microtubules 14. Weisbrich A, et al. (2007) Structure-function relationship of CAP-Gly domains. Nat at the in interphase cells. Traffic 9:472–480. Struct Mol Biol 14:959–967. 36. Sheeman B, et al. (2003) Determinants of S. cerevisiae dynein localization and activa- 15. King SJ, Schroer TA (2000) Dynactin increases the processivity of the cytoplasmic dynein tion: Implications for the mechanism of spindle positioning. Curr Biol 13:364–372. motor. Nat Cell Biol 2:20–24. 37. Mallik R, et al. (2004) Cytoplasmic dynein functions as a gear in response to load. Nature 16. Ross JL, et al. (2006) Processive bidirectional motion of dynein-dynactin complexes in 427:649–652. vitro. Nat Cell Biol 8:562–570. 38. Gennerich A, Carter AP, Reck-Peterson SL, Vale RD (2007) Force-induced bidirectional 17. Kim H, et al. (2007) Microtubule binding by dynactin is required for microtubule stepping of cytoplasmic dynein. Cell 131:952–965. organization but not cargo transport. J Cell Biol 176:641–651. 39. Eaton BA, Fetter RD, Davis GW (2002) Dynactin is necessary for synapse stabilization. 18. Dixit R, et al. (2008) Regulation of dynactin through the differential expression of Neuron 34:729–741. p150glued isoforms. J Biol Chem 283:33611–33619. 40. Grabham PW, et al. (2007) Cytoplasmic dynein and LIS1 are required for microtubule 19. Adames NR, Cooper JA (2000) Microtubule interactions with the cell cortex causing nuclear movements in Saccharomyces cerevisiae. J Cell Biol 149:863–874. advance during growth cone remodeling and fast axonal outgrowth. J Neurosci 20. Carminati JL, Stearns T (1997) Microtubules orient the mitotic spindle in yeast through 27:5823–5834. dynein-dependent interactions with the cell cortex. J Cell Biol 138:629–641. 41. Tsai JW, Bremner KH, Vallee RB (2007) Dual subcellular roles for LIS1 and dynein in 21. Lee WL, Kaiser MA, Cooper JA (2005) The offloading model for dynein function: radial neuronal migration in live brain tissue. Nat Neurosci 10:970–979. Differential function of motor subunits. J Cell Biol 168:201–207. 42. Amberg DC, Burke D, Strathern JN (2005) Methods in Yeast Genetics (Cold Spring 22. Das R, Baker D (2008) Macromolecular modeling with rosetta. Annu Rev Biochem Harbor Lab, NY). 77:363–382. 43. Gray M, Piccirillo S, Honigberg SM (2005) Two-step method for constructing unmarked 23. Slep KC, Vale RD (2007) Structural basis of microtubule plus end tracking by XMAP215, insertions, deletions and allele substitutions in the yeast genome. FEMS Microbiol Lett CLIP-170, and EB1. Mol Cell 27:976–991. 248:31–36.

5152 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0810828106 Moore et al. Downloaded by guest on October 2, 2021