RESEARCH ARTICLE

DnaJB6 is a RanGTP-regulated protein required for microtubuleorganization during mitosisMiquel Rosas-Salvans1, Jacopo Scrofani1, Aitor Modol1 and Isabelle Vernos1,2,3,*

ABSTRACTBipolar spindle organization is essential for the faithful segregation ofchromosomes during cell division. This organization relies on thecollective activities of motor proteins. The minus-end-directed dyneinmotor complex generates spindle inward forces and plays amajor rolein spindle pole focusing. The dynactin complex regulates manydynein functions, increasing its processivity and force production.Here, we show that DnaJB6 is a novel RanGTP-regulated protein.It interacts with the dynactin subunit p150Glued (also known asDCTN1) in a RanGTP-dependent manner specifically in M-phase,and promotes spindle pole focusing and dynein force generation. Ourdata suggest a novel mechanism by which RanGTP regulates dyneinactivity during M-phase.

KEY WORDS: DnaJB6, RanGTP, Spindle organization, Dynein,Microtubules

INTRODUCTIONSpindle assembly involves the organization of microtubules intotwo antiparallel arrays with interdigitating plus-ends, and twofocused minus-ends forming the spindle poles. This organization ishighly dependent on the collective activities of microtubule-dependent motor proteins. The plus-end-directed homotetramericmotor Eg5 (also known as KIF11) generates outward forces that,together with other mechanisms, separate the spindle poles (Blangyet al., 1995; Kapitein et al., 2005; Kashina et al., 1996, 1997;Whitehead and Rattner, 1998). Indeed, interfering with Eg5 activityresults in the assembly of monopolar spindles with unseparatedcentrosomes (Mayer et al., 1999). Interestingly, impairing dyneinactivity rescues spindle bipolarity in Eg5-inhibited cells (Ferenzet al., 2009; Mitchison et al., 2005; Raaijmakers et al., 2013;Tanenbaum et al., 2008). This suggests that a fine balance betweenoutward forces generated by Eg5 and inward forces generated bydynein defines spindle length and bipolarity (Ferenz et al., 2009;Florian and Mayer, 2012; Gaglio et al., 1997, 1996; Heald et al.,1996; Tanenbaum et al., 2008; Walczak et al., 1998).In addition to force generation within the spindle, dynein is also

essential for focusing microtubule minus-ends into the spindle poles(Echeverri et al., 1996; Gaglio et al., 1996; Heald et al., 1996;

Mitchison et al., 2005; Tan et al., 2018; Walczak et al., 1998;Wittmann and Hyman, 1999). This process involves NuMA (alsoknown as NUMA1), which recruits the dynein–dynactin complex tothe microtubule minus-ends, promoting the transport and focusingof the microtubule minus-ends (Compton and Cleveland, 1993;Gaglio et al., 1995; Haren et al., 2009; Hueschen et al., 2017;Khodjakov et al., 2003; Kisurina-Evgenieva et al., 2004; Merdeset al., 2000, 1996; Silk et al., 2009). At the poles, dynein is alsoinvolved in centrosome coupling to the spindle. Indeed, interferingwith dynein activity induces centrosome positioning defects ordetachment from the spindle pole (Heald et al., 1997; Gaglio et al.,1997; Morales-Mulia and Sholey, 2005; Goshima et al., 2005).During mitosis, dynein is also recruited to the cell membranethrough interaction with different protein complexes. Corticaldynein localization and dynein-dependent force generation onastral microtubules regulate spindle positioning and orientation(di Pietro et al., 2016; Lu and Johnston, 2013).

During M-phase, RanGTP triggers the nucleation, stabilizationand organization of microtubules in the vicinity of thechromosomes, playing an essential role in bipolar spindleassembly (Carazo-Salas et al., 1999; Cavazza and Vernos, 2015;Clarke and Zhang, 2008). The general mechanism involvesthe release of a specific group of nuclear localization signal(NLS)-containing proteins from inhibitory interactions withimportins (Görlich et al., 1995, 1996). Several RanGTP-regulatedproteins have been identified and shown to play essential roles inspindle assembly (Cavazza and Vernos, 2015). Recently, we useda mass spectrometry-based proteomic approach to identify thefull proteome associated with RanGTP-induced microtubules inXenopus egg extracts and validated a few selected proteins with noreported function in mitosis (Rosas-Salvans et al., 2018). One ofthem, DnaJB6 (also known as MRJ, HSJ2 or MSJ1), had also beenidentified in mitotic spindle proteomes (Bonner et al., 2011; Raoet al., 2016; Sauer et al., 2005) although its function in mitosis hadnot been addressed. DnaJB6 is an heat-shock protein 40 kDa(HSP40) family protein with two alternatively spliced isoforms(in humans). These two isoforms share the first 239 amino acids butdiffer at their C-terminus. The short isoform (DnaJB6-S, 241 aminoacids) is cytoplasmic, whereas the long isoform (DnaJB6-L, 326amino acids) has a nuclear localization signal at its C-terminus(Mitra et al., 2008), and has been shown to localize to the nucleus(Cheng et al., 2008; Mitra et al., 2008). The co-chaperon activity ofthe short isoform has been demonstrated in vitro (Chuang et al.,2002). Additionally, DnaJB6 prevents huntingtin aggregationin vitro independently of its DnaJ domain (and HSP70), althoughin vivo this domain seems to be required (Chuang et al., 2002;Hageman et al., 2010). DnaJB6-L expression is associated withsuppression of tumorigenesis and metastasis in breast cancer (Menget al., 2016). Interestingly, DnaJB6 is highly expressed in humantestis, ovary, liver and placenta (Seki et al., 1999), and its expressionis increased during mitosis in HeLa cells (Dey et al., 2009; SekiReceived 24 October 2018; Accepted 16 April 2019

1Centre for Genomic Regulation (CRG), Barcelona Institute of Science andTechnology, Dr Aiguader 88, 08003 Barcelona, Spain. 2Universitat PompeuFabra (UPF), 08003, Barcelona, Spain. 3ICREA, Passeig de Lluis Companys 23,08010 Barcelona, Spain.

*Author for correspondence (Isabelle.Vernos@crg.eu)

M.R., 0000-0002-5397-121X; I.V., 0000-0003-1469-9214

This is an Open Access article distributed under the terms of the Creative Commons AttributionLicense (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use,distribution and reproduction in any medium provided that the original work is properly attributed.

1

© 2019. Published by The Company of Biologists Ltd | Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

et al., 1999). We have previously shown that DnaJB6 localizes to thespindle poles and its silencing affects microtubule aster formation incells undergoing microtubule regrowth, suggesting that it is a novelRanGTP-regulated spindle assembly factor (SAF) (Rosas-Salvanset al., 2018).Here, we show that DnaJB6 interacts with the dynactin complex

component p150Glued (also known as DCTN1) in a RanGTP-dependent manner in pulldown experiments. Moreover, DnaJB6silencing induces multiple spindle defects that are all consistent withdefective dynein function. These data suggest that dynein could beregulated during mitosis through a novel mechanism involvingDnaJB6 and RanGTP.

RESULTSDnaJB6 is a RanGTP-regulated protein that localizesto the spindleDnaJB6 has two alternatively spliced isoforms in humans that differat their C-terminus. To determine the localization of these proteinsin interphase and in mitosis, we performed immunofluorescencestudies in cells expressing the long or the short isoform of DnaJB6tagged with FLAG. In agreement with previous reports, DnaJB6-Slocalized to the cytoplasm in interphase, whereas DnaJB6-Llocalized to the nucleus (Cheng et al., 2008; Mitra et al., 2008).In mitotic cells, DnaJB6-S did not show any specific localization.By contrast, DnaJB6-L localized to the spindle in metaphase cells(Fig. 1A).We then used a Xenopus egg extract system to directly test

whether the long isoform of DnaJB6 (DnaJB6-L; xDnaJB6-L refersto the Xenopus form), which contains an NLS, could be regulated byRanGTP during mitosis. GST–xDnaJB6-L-coated beads wereincubated in cytostatic factor (CSF)-arrested egg extractsupplemented or not with RanGTP. Western blot (WB) analysisof the proteins associated with the beads showed that importin-βwasspecifically pulled down by GST–xDnaJB6-L in the absence butnot in the presence of RanGTP (Fig. 1B). This result suggests thatDnaJB6 is regulated by RanGTP during mitosis as is seen for otherSAFs (Cavazza and Vernos, 2015). The high degree of conservationof the NLS sequence of the Xenopus and human DnaJB6 proteins(Fig. 1C) suggests that the interaction with importin-β could beconserved in human.We then used immunofluorescence microscopy to characterize

the subcellular localization of endogenous DnaJB6 during the cellcycle, using an anti-DnaJB6 antibody that recognizes both isoformsin western blotting experiments. The anti DnaJB6 signal waspredominant in the nucleus during interphase with some signal inthe cytoplasm.Mitotic cells showed a clear signal for DnaJB6 on thespindle microtubules with an accumulation at the spindle poles(Fig. 1D).

DnaJB6has a role inmicrotubule organization duringmitosisTo determine whether DnaJB6 has any role during mitosis,we monitored mitosis progression by performing time-lapsefluorescence microscopy in control and DnaJB6-silenced HeLacells expressing H2B–eGFP and α-tubulin–mRFP. DnaJB6silencing efficiently reduced the levels of the protein (Fig. 2A) andinduced a delay in mitosis. Indeed, DnaJB6-silenced cells needed35.5% more time than control cells to reach anaphase onset fromnuclear envelope breakdown (NEB) (Fig. 2B). To determine thecauses for this delay, we examined microtubule organization andchromosome positioning in fixed control andDnaJB6-silencedHeLacells by immunofluorescence microscopy. The quantification of thedistribution of the mitotic phases showed a significant decrease in the

Fig. 1. DnaJB6 is a RanGTP-regulated protein associated to the mitoticspindle. (A) Immunofluorescence on HeLa cells transfected with Flag-taggedlong and short DnaJB6 isoform-expressing constructs. DnaJB6-L (long)localizes in the nucleus during interphase and to the spindle poles inmitosis. DnaJB6-S (short) shows diffuse cytoplasmic localization bothduring interphase and mitosis. (B) Western blot analysis of a GST–xDnaJB6-Lpulldown experiment in Xenopus laevis egg extract. Importin-β associateswith GST–xDnaJB6-L and is released by addition of RanGTP to the extract(top). The lower panel shows that similar levels of GST–xDnaJB6 wereused for pull downs in the presence or absence of RanGTP. (C) Amino acidsequence alignment of the putative NLS in human (top) and Xenopus laevis(bottom) DnaJB6 long isoforms. Asterisks highlight identical amino acids.(D) Immunofluorescence images of HeLa cells showing the localization ofDnaJB6 in different cell cycle stages using an in-house generated antibody.DnaJB6 accumulates in the nucleus during interphase and localizes to thespindle during mitosis with an accumulation at the spindle poles in metaphase.Tubulin is shown in red, DnaJB6 in green and DNA in blue. Scale bars: 10 μm.

2

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

proportion of DnaJB6-silenced cells in metaphase (Fig. 2C).Concomitantly, the percentage of cells with spindle defectsincreased significantly. Additionally, an increase of the percentageof pro-metaphase cells was observed, although this was notstatistically significant (Fig. 2C).One major phenotype was a significant increase of the proportion

of cells with multipolar spindles in DnaJB6-silenced cells comparedto controls (30.5% versus 5.6%) (Fig. 2D). This phenotype was notassociated with centrosome amplification. Indeed, 88% of themultipolar spindles assembled in DnaJB6-silenced cells had onlytwo centrin-positive poles (Fig. 2E). Upon closer examinationbipolar spindles assembled in DnaJB6-silenced cells also showed arange of defects. They were significantly longer than thoseassembled in control cells (10.4 μm in DnaJB6-silenced cellsversus 9.9 μm in control cells) (Fig. 3A) and, in some cases (11.5%in DnaJB6-silenced cells and 1.1% in control cells), small ectopicmicrotubule clusters were in their proximity, suggesting defectiveincorporation of some microtubules to the main spindle body(Fig. 3B). Overall, there was also an increase of spindles not

correctly oriented with respect to the substrate in DnaJB6-silencedcells (Fig. 3C). Moreover, bipolar spindles assembled in DnaJB6-silenced cells had pole focusing defects with a significantly widerpolewidth than control bipolar spindles (Fig. 3D). In addition, somecentrosomes were loosely attached to the spindle pole (Fig. 3E) andthe centrosomes axis was altered showing a deviation with respect tothe spindle axis (Fig. 3E).

Taken together, these results suggest a role for DnaJB6 inmicrotubule organization during mitosis.

DnaJB6 has a direct role in microtubule organizationin M-phaseTo gain further support for the hypothesis that DnaJB6 is involvedin microtubule organization during mitosis, we monitoredmicrotubule organization in control and DnaJB6-silenced cellsundergoing regrowth after nocodazole washout. Cells were fixed atdifferent time points and processed for immunofluorescence(Fig. 4A). The efficiency of bipolar spindle organization wasmeasured by quantifying the number of microtubule asters at the

Fig. 2. DnaJB6 is required for bipolar spindle assembly. (A) Western blot analysis of control and DnaJB6-silenced cell lysates showing the efficiency ofDnaJB6 silencing. Cells were lysed 48 h post transfection and 40 μg of protein were loaded per lane. DnaJB6 silencing efficiency was 90% for DnaJB6-L andmore than 90% for DnaJB6-S. Tubulin is blotted as protein loading control. (B) Box-and-whisker plot showing the time from nuclear envelope breakdown (NEB) toanaphase onset for control and DnaJB6-silenced cells. Data are from three independent experiments in which a total of 281 control and 250 DanJB6silenced cells were analyzed. ***P<0.0001 (Mann–Whitney test). (C) Bar graph showing the mean±s.d. distribution of mitotic phases in control (blue) orDnaJB6-silenced HeLa cells (green). Prophase (P), prometaphase (PM), metaphase (M), anaphase (A) and aberrant spindles (Asp) are quantified. Monopolar,multipolar spindles and more disorganized structures were included into the aberrant spindle category (Asp). Data from three independent experiments inwhich a total of 454 control and 498 DnaJB6-silenced cells were analyzed. *P<0.05, ***P<0.0001 (ANOVA test). (D) Multipolar spindle are highly representedin DnaJB6-silenced cells. Left: representative image of a DnaJB6-silenced HeLa cell with amultipolar spindle. Tubulin is in red andDNA in blue. Scale bar: 10 μm.Right: quantification (mean±s.d.) of the percentage of multipolar spindles in control (blue) and DnaJB6-silenced (green) cells. The percentages are indicatedon top of the columns. Data are from three independent experiments in which a total of 765 control and 778 DnaJB6-silenced cells were analyzed. ***P<0.0001(Fisher’s exact test). (E) DnaJB6-silenced cells present extra poles that are mainly not centrosomal. Left: immunofluorescence images of a DnaJB6-silencedHeLa cell with a multipolar spindle showing tubulin (red), centrin (green) and DNA (blue). The arrowhead points to a centrin negative pole. Scale bar: 10 μm.Right: quantification (mean±s.d.) of the percentage of multipolar spindles with poles negative for centrin in control (blue) and DnaJB6-silenced (green) HeLacells. Thepercentages are indicated on top of the columns. Data are from two independent experiments inwhich a total of 126 control and 116DnaJB6-silenced cellswere analyzed. ***P<0.0001 (ANOVA test).

3

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

different time points (Fig. S1). In most control cells, the number ofmicrotubule asters was initially high due to the activity of thecentrosomal and chromosomal pathways (Meunier and Vernos,2011), and reduced over time so that finally only two microtubuleasters defined the poles of the bipolar spindle. In contrast, thepercentage of cells with only two microtubule asters wassignificantly reduced in DnaJB6-silenced cells at all the timepoints examined (Fig. 4B) and the difference with the controlincreased with time. At 60 min after nocodazole washout, 57.7% ofthe control cells had two asters organized into a bipolar spindle,whereas this percentage was reduced to 24.8% in DnaJB6-silenced

cells (Fig. 4C). These data suggest that the role of DnaJB6 inmicrotubule organization is associated with microtubule asterclustering and spindle pole focusing.

Since DnaJB6 is a member of the Hsp40 co-chaperon family, itsabsence could potentially lead to an accumulation of protein foldingdefects that could be responsible for the phenotypes that weobserved in mitosis. To rule out this possibility, we used theXenopus egg extract system. Cycled spindle assembly experimentswere performed in control and DnaJB6-depleted egg extracts andmaltose-binding protein (MBP) or recombinant MBP–xDnaJB6-Lwere added to the control and depleted extracts for rescue

Fig. 3. See next page for legend.

4

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

experiments (Fig. 5A). Samples were collected 60 min after cyclingthe extract into mitosis and processed for immunofluorescence.Although bipolar spindles assembled with similar efficiency incontrol and DnaJB6-depleted extracts (data not shown), the absenceof DnaJB6 resulted in spindle pole defects (Fig. 5B). While 86.8%of the control spindles had focused poles, spindles assembled inDnaJB6-depleted extracts had a statistically significant reduction offocused spindle poles to 64% and a high proportion of open spindlepoles, sometimes even generating two sub-poles. This phenotypewas rescued by addition ofMBP–xDnaJB6-L to the depleted extractpromoting the formation of focused spindle poles in 81% of samples(Fig. 5C). These results showed that the spindle pole focusingdefects can be entirely attributed to the absence of DnaJB6specifically during mitosis. Consistently, the addition of an excess

of MBP-xDnaJB6-L to the egg extracts as they were cycled intomitosis resulted in an increase of the percentage of very tightlyfocused spindle poles (Fig. 5D).

Together, these results indicate that DnaJB6 plays a direct role inmicrotubule focusing at the spindle poles during M-phase.

DnaJB6 interacts with p150Glued in a RanGTP-dependentmanner during mitosis and is required for its spindlelocalizationThe defects in spindle organization that we observed upon DnaJB6silencing in HeLa cells or depletion in Xenopus egg extracts(including spindle misorientation, centrosome detachment, spindlepole focusing defects and spindle length increase) are all compatiblewith altered dynein activity (Kardon and Vale, 2009; Raaijmakersand Medema, 2014). Pulldown experiments in egg extracts usingMBP–xDnaJB6-L- orMBP-coated beads in the presence or absenceof RanGTP showed that the dynactin subunit p150Glued wasspecifically pulled down by MBP–xDnaJB6 and only in thepresence of RanGTP (Fig. 6A). Therefore DnaJB6 interacts withp150Glued in a RanGTP-dependent manner during M-phase.

Since DnaJB6 is a co-chaperon, we reasoned that its interactionwith p150Glued could have a role in the assembly or stabilization ofthe dynactin complex. To test this idea, we performed sucrosedensity gradient analysis of cell lysates in the presence of thechaotropic agent potassium iodide (KI), which has previously beenshown to promote the dissociation of the dynein complex (Kinget al., 2002). Control and DnaJB6-silenced HeLa cells weresynchronized in interphase and mitosis, and the cell lysatesincubated with KI before loading on top of an 8–20% sucrosedensity gradient. In control cells, western blot analysis showed thatp150Glued migrated to lighter fractions of the gradient in thepresence of KI. This indicated that KI induced the dissociation ofthe p150Glued complex both in interphase and mitosis, as describedfor the dynein complex. DnaJB6 silencing did not change thedissociation pattern of the p150Glued complex in interphase(Fig. S4). In contrast, it promoted a shift of p150Glued to lowersucrose concentrations in mitosis (Fig. 6B). These results suggestthat DnaJB6 is required for the stability of a p150Glued complexspecifically in mitosis. This complex may be the dynactin complexitself or a larger complex including dynein.

We then checked whether p150Glued spindle localizationwas affected in the absence of DnaJB6 both in egg extracts and incells. Spindles assembled in egg extracts showed that p150Gluedlocalized all along the spindle microtubules with some enrichmentat the spindle poles. In the absence of DnaJB6, p150Gluedaccumulated at the spindle poles (Fig. S2A,B). This was not relatedto changes in p150Glued protein stability since there was no changein the general levels of p150Glued in control and DnaJB6-depletedextracts (Fig. S2C). Rescue experiments showed that p150Gluedaccumulation was reverted to control levels at the spindle poles uponaddition of MBP–xDnaJB6-L to the depleted extract (Fig. S2A,B).These results indicate that DnaJB6 plays a role in p150Gluedspindle pole localization in egg extracts.

Consistent with the above experiments, p150Glued alsoaccumulated at the spindle poles in DnaJB6-silenced cells(Fig. 6C). We then checked dynein localization and found thatthe dynein intermediate chain and heavy chain (DIC and DHC,respectively) family proteins also localized more prominently to thespindle poles in DnaJB6-silenced cells. (Fig. S3B,C). Neither DICproteins, DHC proteins nor p150Glued total protein levels wereaffected by the silencing of DnaJB6 in HeLa cells (Fig. S3A).Finally, we also found spindle pole enrichment of NuMA

Fig. 3. DnaJB6 is required for spindle orientation and spindle polesorganization. (A) Box-and-whisker plot showing the spindle length incontrol and DnaJB6-silenced HeLa cells. The spindle length of 267 controland 294 DnaJB6-silenced cells from three independent experiments wasmeasured. A statistically significant increase of the length was detectedin DnaJB6-silenced cells. ***P<0.0001 (Mann–Whitney test). (B) DnaJB6-silenced cells have ectopic microtubule clusters in metaphase. Left:immunofluorescence images of a representative DnaJB6-silenced metaphasecell with an ectopic microtubule cluster (white arrowhead). Tubulin is shownin red and DNA in blue. Scale bar: 10 μm. Right: bar graph showing thepercentage of bipolar spindles containing ectopic microtubule clusters incontrol and DnaJB6-silenced HeLa cells. 182 control and 192 DnaJB6-silenced cells in metaphase were examined. ***P<0.001 (Fisher’s exact test).(C) DnaJB6-silenced cells present spindle orientation defects. Left:representative images of a control cell with a correctly oriented spindle inwhich the two centrosomes are aligned with the substrate (top image) and aDnaJB6-silenced cell with a misoriented spindle (bottom image; in this case,only one centrosome is visible due to the location of the two poles on twodifferent focal planes). Tubulin is in red, centrin in green and DNA in blue.Scale bar: 10 μm. Right: quantification of the mean±s.d. percentage ofmetaphase HeLa cells that present misoriented spindles. Spindles werescored as not correctly oriented with the substrate when the two centrosomeswere not on the same focal plane. Anti-centrin antibodies were used todetermine centrosome position. The percentages are indicated on topof the columns. Data from four independent experiments in which a total of600 cells were analyzed for condition. ***P<0.001 (Student’s t-test).(D) DnaJB6-silenced cells show severe spindle pole focusing defects.Left: immunofluorescence images of control and DnaJB6-silenced HeLacells showing tubulin (red), centrin (green) and DNA (blue). Spindles from acontrol and a DnaJB6-silenced are shown as examples of focused andopen spindle poles, respectively. Double arrows exemplify poles widthmeasurement. Scale bar: 10 μm. Right: box-and-whisker plot showingmeasures of spindle pole width in control and DnaJB6-silenced HeLa cells.Bipolar metaphase spindles with correctly aligned chromosomes wererandomly selected. The tubulin signal was thresholded and a line was drawncrossing the centrosome (centrin staining) and connecting the two closestspindle borders. Graph shows the results of a representative experimentout of two in which a total of 90 poles were measured in each condition.***P<0.001 (Student’s t-test). (E) DnaJB6-silenced cells show centrosomemispositioning relative to the spindle pole position. Left: representative imagesof a control and two DnaJB6-silenced cells. Centrosomes are laterallydisplaced or detached from the spindle poles in DnaJB6-silenced cells.Tubulin is in red, centrin is green and DNA is blue. Scale bar: 10 μm. Right:quantification of the centrosome position defects. When the centrosomesmove laterally the centrosome axis is deviated. The left-hand box-and-whiskerplot represents the centrosome axis deviation in control and DnaJB6-silencedcells. This deviation is calculated by measuring the α and subtracting this from90, where α is the angle between the centrosome axis and the metaphaseplate, as shown in the schematic representation. When the centrosomes aredetached from the poles a midpoint is fixed (yellow cross) and distance fromthe metaphase plate (h) measured and plotted (shown on the right). The plotsshow data from a representative experiment where more than 30 cells wereanalyzed for each condition. ***P<0.001 (Student’s t-test).

5

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

(Fig. S3D), recently shown to recruit p150Glued to the microtubuleminus-ends in mitosis (Hueschen et al., 2017), although to a lesserextent than dynactin or dynein.Taken together, our data indicates that DnaJB6 may be required

for dynein function during mitosis possibly though an interactionwith the dynactin complex in a RanGTP-dependent manner.

DnaJB6 favors the generation of inward forces withinthe spindleThe dynactin complex is the major adaptor for dynein, promoting itsprocessivity and force generation (Cianfrocco et al., 2015; Grotjahnet al., 2018; Jha and Surrey, 2015). The phenotypes associated withthe absence of DnaJB6 during mitosis are consistent with thosedescribed for interference with dynein in mitosis, including spindlepole focusing defects and multipolarity (McKinley and Cheeseman,2017; Raaijmakers and Medema, 2014). Moreover, our sucrosegradient results are compatible with DnaJB6 being required for thestability of the dynactin or dynein–dynactin complexes. Thissuggests that DnaJB6 silencing could impair dynein activity duringmitosis. It has been previously shown that interfering with dyneinactivity rescues the formation of bipolar spindles in Eg5-deficientcells (Ferenz et al., 2009; Mitchison et al., 2005; Raaijmakers et al.,2013; Tanenbaum et al., 2008). We therefore asked whetherDnaJB6 silencing could affect dynein force generation and rescuespindle bipolarity in HeLa cells treated with S-trityl-l-cysteine[STLC, and Eg5 inhibitor (Mayer et al., 1999)]. Control andDnaJB6-silenced HeLa cells were incubated with STLC for 2 h,fixed and processed for immunofluorescence. As expected, amajority of control cells had monopolar spindles (88%).Interestingly, DnaJB6-silenced cells had a significant increase inthe proportion of cells with bipolar spindles to 28%, instead of 10%in the control cells (Fig. 6D). To test whether DnaJB6 couldfunction through Hsp70 for inward force regulation, we testedwhether a general HSP70 inhibitor (VER155008, Massey et al.,

2010; Schlecht et al., 2013) could produce a similar bipolar spindleassembly rescue in STLC-treated cells. We did not observe anyrescue by inhibiting HSP70, indicating a direct role for DnaJB6 inthe regulation of inward force production during spindle assembly(Fig. S5A,B). The lower level of inward spindle forces suggestedin DnaJB6-silenced cells treated with STLC is indeed consistentwith the increase in spindle length in DnaJB6-silenced HeLacells (Fig. 3A), a correlation also shown in previous publications(Morales-Mulia and Scholey, 2005; Goshima et al., 2005;Mitchison et al., 2005).

Taken together, our results suggest that DnaJB6 is required fordynein-dependent spindle inward force generation during mitosis.

DISCUSSIONHere, we showed that the nuclear protein DnaJB6 interacts withdynactin p150Glued in a RanGTP-dependent way during mitosis byperforming pulldown experiments. Moreover, all the phenotypesthat we observed upon silencing or depleting DnaJB6 in cells andegg extracts are consistent with defects in dynein function. Ourwork therefore suggests a novel mechanism by which the smallGTPase Ran, in its GTP-bound form, regulates the activity of thedynein–dynactin complex during M-phase to promote inward forceproduction and spindle pole focusing.

Two isoforms of DnaJB6 are present in human cells but only thelong isoform localizes to the nucleus in interphase and to the spindlein mitosis. DnaJB6 also interacts with p150Glued in a RanGTP-dependent manner in M-phase egg extracts. Silencing experimentsreduced the levels of both the short and the long isoforms, but thedestabilization of the dynactin complex only occurred in mitosissuggesting that the interphase cytoplasmic short isoform plays norole in dynactin stabilization and function. Moreover, our data showthat the long isoform of DnaJB6 fully rescues the phenotypes of theDnaJB6 depletion in egg extracts and the distribution of p150Gluedin the spindle. Taken together, these data strongly suggest that

Fig. 4. DnaJB6 is involved in spindle microtubulesorganization. (A) Immunofluorescence images ofmitotic control and DnaJB6-silenced HeLa cellsundergoing microtubule regrowth after nocodazolewashout. Cells were fixed at the times indicatedafter nocodazole washout and processed forimmunofluorescence to visualize the microtubules(tubulin, shown in gray) and DNA (not shown). Thewhite arrowheads point to microtubule asters thatfail to incorporate into the main spindle poles. Scalebar: 10 μm. (B) Quantification of the percentage ofcells with two asters at different time points afternocodazole (NOC) washout as shown in A. The blueline corresponds to the control cells and the greenline to DnaJB6-silenced cells. Data are from threeindependent experiments in which more than300 cells were recorded for each time point.***P<0.001 (Fisher’s exact test). (C) Representativeimmunofluorescence image of a HeLa cell fixed at60 min after nocodazole washout with two microtubuleasters that do (top) or do not (bottom) organize abipolar spindle. Tubulin is in red, and DNA in blue.Scale bar: 10 μm. Quantification of the mean±s.d.percentage of cells having two microtubule astersorganized into a bipolar spindle 60 min afternocodazole washout. Data from two independentexperiments in which 503 control (blue) and 638DnaJB6-silenced cells (green) were analyzed. Theexact percentages are indicated on top of each bar.***P<0.001 (Fisher’s exact test).

6

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

the function in spindle assembly that we describe here is specific forthe long isoform of DnaJB6.DnaJB6 is a member of the HSP40 protein family. The

best-characterized role of several members of this family is to actas co-chaperons of HSP70 proteins. Some members of the HSP70family were previously shown to play a role in spindle assembly(Fang et al., 2016; O’Regan et al., 2015). Here, we found thatHSP70 inhibition does not rescue spindle bipolarity inEg5-inhibited cells nor does it generate spindle pole focusingdefects (data not shown). Moreover, the rescue experiments in egg

extracts support a direct role for DnaJB6 in mitosis since therecombinant protein that rescued the depletion phenotype wasadded to the depleted extract only as the extract was sent into mitosis(Fig. 5; Fig. S2). Taken together, our data strongly suggest thatthe role of DnaJB6 in spindle assembly that we describe here isindependent of HSP70. In agreement with this, HSP70-independentfunctions have been reported for some HSP40 family proteinsincluding DnaJB6 (Hageman et al., 2010).

Dynein is a major minus-end-directed motor that performs alarge number of functions in the cell. During mitosis it is essentialfor centrosome separation, chromosome alignment, spindle polefocusing, spindle positioning, centrosome coupling to spindlepoles and mitotic checkpoint silencing, in addition to generatingthe inward forces that balance those mediated by Eg5 (Kardonand Vale, 2009; Raaijmakers and Medema, 2014; Goshima,2005; di Pietro et al., 2016). All the phenotypes observed inDnaJB6-silenced cells are compatible with an impairment of dyneinfunction: centrosome mispositioning, spindle misorientation, spindlepole focusing defects and an increase in spindle length. The increasein spindlemultipolaritymay also result fromunbalanced forces withinthe spindle. Indeed a weakening of inward forces produced by dyneinmay result in dominant outward forces through Eg5, which couldresult in the disruption of the spindle pole integrity and the formationof multipolar spindles (Maiato and Logarinho, 2014). In agreementwith this idea, we found that DnajB6 silencing restores spindlebipolarity when Eg5 is inhibited.

The wide range of dynein functions raises the question of howthis motor is regulated. Several interacting proteins have been foundto provide specificity for dynein localization and/or regulation(Cianfrocco et al., 2015). One of the best-characterized generalregulators of dynein is dynactin, a large protein complex thatparticipates in most dynein functions. Dynactin was shown to act as adynein activator increasing its processivity and force production(Culver-Hanlon et al., 2006; Grotjahn et al., 2018;Kardon et al., 2009;King and Schroer, 2000; Ross et al., 2006; Tripathy et al., 2014).

We have shown that the long isoform of DnaJB6 pulls downthe dynactin subunit p150Glued in a RanGTP-dependent mannerduring mitosis. This interaction may play a role in dynactin complexstability, as indicated by its reduced sensitivity to KI-induceddisassembly in mitosis. However, and surprisingly, dynactinand dynein, as well as NuMA accumulate at the spindle poles inDnaJB6-silenced cells. These data are not very easy to reconcile,and additional work will be needed to understand preciselythe mechanism by which DnaJB6 affects dynein function andlocalization during mitosis. However, our data overall point to animportant role for DnaJB6 in modulating the functioning ofdynactin and dynein during spindle assembly.

In summary, we describe here the identification and functionalcharacterization of DnaJB6 as a novel RanGTP-regulated proteinrequired for microtubule organization during mitosis, and suggestthat it is important for dynein functions in the spindle.

MATERIALS AND METHODSHeLa cell culture, DnaJB6 silencing and drug treatmentsCells were grown at 37°C in a 5% CO2 humid atmosphere. HeLa cells weregrown in Dulbecco’s modified Eagle medium (DMEM) with 4.5 g/l glucoseand Ultraglutamine 1 (bioWhittaker, BE12-604F/U1), 10% fetal bovineserum (FBS; 102070-106, Invitrogen) and 100 units/ml penicillin and100 µg/ml streptomycin (15140-122, Invitrogen). HeLa cells stablyexpressing H2B–eGFP and α-tubulin–mRFP were a gift from PatrickMeraldi (ETH, Zurich, Switzerland) (McAinsh et al., 2006) and were grownin the presence of 400 µg/ml G418 and 20 µg/ml puromycin. Cells wereperiodically tested for mycoplasma contamination.

Fig. 5. DnaJB6 depletion from Xenopus egg extracts induces spindlepole organization defects. (A) Western blot analysis of control andxDnaJB6-depleted (Δ-DnaJB6) Xenopus egg extracts (EE) without (−) orwith (+) addition of the recombinant MBP–xDnaJB6-L. The depletion wasvery efficient and the recombinant protein was added at close to endogenousconcentrations. 1 μl of egg extract was loaded per lane. The endogenousand recombinant proteins were detected with an in-house generatedanti-xDnaJB6 antibody. (B) Fluorescence images of spindles assembled incontrol or DnaJB6-depleted egg extracts. Most spindles assembled incontrol-depleted extracts have focused spindle poles, whereas manyspindles assembled in DnaJB6-depleted extracts show spindle polefocusing defects defined as ‘Split poles’ or ‘Open poles’ as shown.The white arrowheads point to the different types of spindle poles. In themerge, tubulin is shown in red and DNA in blue. Scale bar: 10 μm. (C) Graphshowing the mean±s.d. percentage of focused spindle poles in controlextracts (blue), DnaJB6-depleted extracts (green) and DnaJB6-depletedextracts containing MBP–xDnaJB6 (orange). Data are from four independentexperiments in which 703, 607 and 358 spindle poles were examined,respectively. ***P<0.001 (ANOVA test). (D) Fluorescence image of a spindleassembled in egg extract supplemented with an excess of DnaJB6 (1 μM)showing tightly focused spindle poles (white arrowheads). Scale bar: 10 μm.Bar graph on the right shows the percentage of tightly focused poles in spindlesassembled in control extracts supplemented with MBP (blue) and in extractssupplemented with MBP–xDnaJB6 (purple). Data are from three independentexperiments in which 114 and 258 spindle poles were analyzed. ***P<0.0001(Fisher exact test).

7

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

Short interfering RNA (siRNA)-mediated interference oligonucleotidestargeting DnaJB6 (5′-CUAUGAAGUUCUAGGCGUG-3′) or scrambled(5′-CGUACGCGGAAUACUUCGAUU-3′) were transfected into HeLacells with Lipofectamine RNAmax (13778-150, Invitrogen) as detailed inthe manufacturer’s protocol; 4 μl of 20 μM siRNA were used for each wellof a six-well plate. All experiments were performed after an incubation at37°C for 48 h after transfection.

For microtubule regrowth experiments, cells were incubated for 3 h at37°Cwith 2 μMnocodazole (Sigma, M1404). Nocodazole was then washedout by three washes with 2 ml of pre-warmed 1× PBS and one of medium at37°C. Cells were incubated for different periods in a 37°C incubator, fixedin −20°C methanol for 10 min and processed for immunofluorescenceusing the DM1A antibody (see below).

For the spindle bipolarization rescue experiment, control andDnaJB6-silenced cells were incubated for 2 h at 37°C with 2 μM STLC(Sigma, 164739), fixed in −20°C methanol and processed forimmunofluorescence using the DM1A or anti-tubulin-β antibodies.

Cloning of Xenopus laevis DnaJB6Xenopus mRNAwas obtained from CSF Xenopus leavis egg extracts usingTRIzol reagent (15596-026, Thermo Fisher Scientific) according tomanufacturer’s instructions. cDNA was generated by RT-PCR usingan oligo-dT coupled with an adaptor sequence (5′-GGCCACGCGT-

CGACTAGTAC +17 ‘T’-3′). The 3′-end sequence of the mRNA wasobtained by PCR using the Phusion High-Fidelity DNA Polymerase(F530S; Thermo Fisher Scientific) according to the manufacturerinstructions, using a primer with the adaptor sequence and anothercorresponding to the central part of the xDnaJB6-S (NM_001095775.1)annotated sequence that is conserved between the long and the shortisoforms in human (5′-GGAGGTTTCCCTGCCTTTGGCCC-3′). In asecond step, another PCR was performed on the Xenopus egg cDNA usinga reverse primer corresponding to the 3′-end of the long isoform previouslyobtained and a forward primer containing the annotated start codon fromxDnaJB6-S (fwd 5′-ATGGTGGAGTATTACGAAGTTTTGGGAGTCC-3′;rev 5′-TTAGTAGATTGGTTTGGAAGACTTCTTTTTCTTG-3′). Thesequence for the xDnaJB6 long isoform was deposited into the GenBankdatabase (https://www.ncbi.nlm.nih.gov/nuccore/MK914533).

Protein productionThe full-length xDnaJB6-L sequence was cloned into the pGex-4T-2(Amersham Pharmacia Biotech) and pMal-c2 (New England Biolabs)vectors. MBP–xDnaJB6 was expressed in the E. coli BL21/RIL strain byinduction with 0.5 mM IPTG for 4 h at 25°C. The protein was purified onamylose resin beads (E8021S, NEB) according to the manufacturer’sinstructions. The purified protein was stored in 20 mMTris-HCl pH 7.4 with200 mM NaCl at −80°C.

Fig. 6. DnaJB6 regulates dynactin spindlelocalization and is required for dynein-dependentforce generation within the spindle. (A) Western blotanalysis of a MBP–xDnaJB6-L pulldown experiment inXenopus egg extract. MBP–xDnaJB6-L- or MBP-coatedDynabeads were incubated in egg extracts in thepresence or absence of RanGTP. p150Glued wasspecifically pulled down with MBP–xDnaJB6-L fromextracts containing RanGTP. The lower panel showsthat similar amounts of MBP–xDnaJB6-L was used forpulldown both in the presence and absence of RanGTP.(B) Western blot analysis showing the position ofp150Glued in 8–20% sucrose density gradients fromlysates of control and DnaJB6-silenced cells. Lysateswere prepared from mitotic cells. KI was added to thelysates and incubated for 1 h before running thegradients, at the concentrations indicated on the right. Ashift to in where p150Glued is detected in to fractionswith a lower percentage of sucrose is observed inmitoticDnaJB6 cell lysates treatedwith 150 mMKI compared tothe mitotic control cells lysates with the same treatment,as highlighted with asterisks. (C) p150Gluedaccumulates at spindle poles in DnaJB6-silenced cells.Left: representative immunofluorescence images fromcontrol and DnaJB6-silenced HeLa cells showing thelocalization of p150Glued in metaphase spindles. In themerge, tubulin is in red, dynein is in green and DNA is inblue. Scale bar: 10 μm. Right: box-and-whisker plotshowing the value of the of p150Glued signalnormalized to that of the tubulin signal (a.u., arbitraryunits) in each spindle pole in control and DnaJB6-silenced HeLa cells. More than 30 cells were analyzedfor each condition in one representative out of threeindependent experiments. (D) DnaJB6 silencingrescues spindle bipolarity in STLC-treated HeLa cells.Left: immunofluorescence images of representativemonopolar and bipolar spindles in DnaJB6-silencedHeLa cells incubated with STLC. Tubulin is shown ingreen and DNA in blue. Scale bar: 10 μm. Right: barsgraph showing the mean±s.d. percentage of bipolarspindles in control or DnaJB6-silenced HeLa cellsincubated with STLC. Data from three independentexperiments in which 937 control and 1057 DnaJB6-silenced cells were analyzed. ***P<0.001 (ANOVA test).

8

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

GST–xDnaJB6 was expressed in the E. coli BL21pLys strain byincubation with 1 mM IPTG overnight at 20°C. For protein purification,glutathione–Sepharose beads (17-5132-01, GE Healthcare) wereused according to manufacturer’s instructions. Protein was kept in 50 mMTris-HCl pH 8.0 with 200 mM KCl at −80°C.

AntibodiesAffinity-purified antibodies against human (h)DnaJB6 and GST werepreviously described (Rosas-Salvans et al., 2018). Antibodies againstxDnaJB6 were obtained by rabbit immunization with MBP–xDnaJB6-Land affinity purified. The working concentrations were as follows:anti-hDnaJB6 antibody [immunofluorescence (IF), 5 μg/ml; WB, 1 μg/ml] and anti-xDnaJB6 antibody (WB, 1 μg/ml).

The following primary commercial antibodies were used: DM1A (Sigma,T6199; IF and WB, 1 μg/ml), anti-tubulin-β (Abcam, ab6046; IF, 1:200),anti-p150Glued (BD Transduction Laboratories, 610474; IF, 5 μg/ml; WB,0.5 μg/ml), anti-centrin (Millipore, 04-1624; IF, 1:1000), anti-NuMA(Calbiochem, NA09L; IF, 1:500), anti-DIC (Millipore; MAB1618;IF, 1:500), anti-DHC (Santa Cruz Biotehcnology; SC514579; IF, 1:50),anti-Eg5 (BD Transduction Laboratories; 611186; IF, 1:1000),anti-importin-β/NTF97 (Abcam; ab36775-50; WB, 1:500) and genericIgGs (Sigma, I5006). The following secondary antibodies were used: AlexaFluor 488-conjugated goat anti-mouse-IgG (A11017; Invitrogen), AlexaFluor 488-conjugated goat anti-rabbit-IgG (A11034; Invitrogen),Alexa Fluor 568-conjugated anti-mouse-IgG (A11031; Invitrogen) andAlexa Fluor 568-conjugated goat anti-rabbit-IgG (A11036; LifeTechnologies) all at 0.2 μg/ml final concentration. For western blots, thefollowing secondary antibodies were used: anti-mouse-IgG conjugated toIRDye 800 (926-32212; Licor), goat anti-rabbit-IgG conjugated to IRDye800 (10733944; Thermo Fisher Scientific), goat anti-rabbit-IgG conjugatedto Alexa Fluor 680 (A-21109; Invitrogen) and goat anti-mouse-IgGconjugated to Alexa Fluor 680 (A21058; Life Technologies) diluted at1:10,000.

Xenopus egg extractsAll the experiments involving animals were performed according tostandard protocols approved by the ethical committee of the Parc de RecercaBiomedica de Barcelona. Xenopus laevis frogs (males and females)were purchased from Nasco and used at an age between 1 and 3 years.CSF-arrested egg extracts were prepared and used to perform cycled spindleassembly assays as previously described (Desai et al., 1999).

For depletion experiments, protein A-conjugated Dynabeads 280(10002D, Invitrogen) were coated with anti-xDnaJB6 antibody or genericIgGs (9 μg of antibody/30 μl of beads) and incubated in freshly preparedegg extract for 30 min on ice. Beads were recovered and protein depletionefficiency tested by western blot analysis. For add-back experiments,0.017 μM (estimated endogenous concentration of DnaJB6 in the eggextract) of MBP–xDnaJB6-L or MBP were added to the depleted eggextracts. To look at the effects of increasing the concentration of DnaJB6,we added 1 μM of MBP or MBP–xDnaJB6-L to egg extracts.

Pulldown experiments from egg extractsProtein A-conjugated Dynabeads 280 (10002D, Invitrogen) were coatedwith anti-GST or anti-MBP antibody (9 μg of antibody/30 μl of beads), andincubated with recombinant proteins (GST, GST–xDNAJB6-L, MBP orMBP–xDnaJB6-L). Protein-coated beads were then incubated in CSF eggextract (30 μl of beads/75 μl of egg extract) for 15 min at 20°C followed by30 min on ice. Recombinant RanQ69L-GTP was added at 15 μM. Beadswere recovered and washed three times in CSF-XB (10 mMHEPES pH 7.7,50 mM sucrose, 100 mM KCl, 0.1 mM CaCl2 and 5 mM EGTA; forchecking importin interaction) or three times in CSF-XB and one in PBScontaining NP40 (for checking p150Glued interaction). Proteins wereeluted from the beads by incubation in sample buffer for 10 min at roomtemperature and run on SDS-PAGE before western blotting.

ImmunofluorescenceCells grown on coverslips were fixed in −20°C methanol for 10 min. Theywere blocked and permeabilized for 30 min in IF-buffer (1× PBS, 0.1%

Triton X-100 and 0.5% BSA). Primary antibodies diluted in IF-buffer werethen applied for 1 h at room temperature. Samples were washed three timesin IF-buffer and secondary antibodies and Hoechst (Hoechst 33342; ref:H3570; Invitrogen; added at 1:1000) were applied for 45 min. Thecoverslips were washed with IF-buffer and twice with PBS beforemounting in Mowiol. For anti-DnaJB6 immunostaining, cells were pre-extracted for 6 s in 1× BRB80, 0.5% Triton X-100 and 1 mMdithiobis(succimidyl propionate) (DSP). Cells were then fixed in PFA 4%for 7 min and then processed for IF as described.

Spindles assembled in egg extracts were centrifuged onto coverslips aspreviously described, fixed in−20°Cmethanol for 10 min and processed forIF following the above protocol.

Sucrose gradientsSucrose gradient (8 to 20%) experiments were performed as described in(Jones et al., 2014). For preparing interphase cell lysates, cells werecollected in 5 ml of PBS using a scraper. For preparing mitotic lysates, cellswere incubated for 15 h in 2 μMnocodazole, washed 3× with PBS and oncewith medium (20 ml). They were then incubated at 37°C for 45 min whenthe majority of the cells were in metaphase. Cells were recovered by shakeoff. To prepare the lysates, cells were washed twice in 1× BRB80 andcentrifuged at 600 g for 5 min. The pellets were resuspended in 500 μl oflysis buffer (1× BRB80, 0.5% Triton-X100 with protease inhibitors) for15 min on ice and subsequently centrifuged at 16,100 g (in a 5415 REppendorf centrifuge) for 10 min at 4°C. The protein concentration of thesupernatants was measured by performing a Bradford assay. 100 μl ofcontrol or DnaJB6-silenced cell lysates were incubated with KI (0, 100 or150 mM) and incubated for 1 h on ice. 45 μl of each samplewas added to thetop of sucrose gradients in centrifuge tubes previously prepared at roomtemperature. Tubes were centrifuged at 38,000 rpm for 5 h at 4°C in a SW-55Ti rotor with an ultracentrifuge Beckman optima L-100K. 75 μl fractionswere recovered from the top and diluted in sample buffer. Samples were thenloaded on SDS-PAGE gels and subjected to western blotting.

Sucrose gradients were prepared in 800 µl (0.8 ml) UltraClear CentrifugeTubes, 5×41 mm (part Number 344090) and specific adaptors used forcentrifugation (Adapter, Split, Delrin, Tube, 5 mm diameter, qty. 2 halves,product number 356860). Sucrose solutions were prepared in BRB80containing protease inhibitors (1:4000), 0.1 mM ATP, 1 mM DTT, sucroseand KI at the selected concentrations (0, 100 or 150 µM KI). The gradientwas prepared by overlaying manually the different sucrose solutions (2%difference of sucrose between layers, seven layers from 20 to 8% of sucrose),freezing between each layer addition. The gradients were kept at −20°C andwarmed up to room temperature for 1 h before use.

MicroscopyNon-synchronized HeLa cells stably expressing H2B–eGFP and α-tubulin–mRFP were imaged every 4 min for 18 h with a 40× objective on a ZeissCell Observer HS inverted microscope equipped with a a Zeiss AxioCamMrX camera and temperature, humidity and CO2 control systems. Imageswere processed and analyzed with FIJI software.

Immunofluorescence images were obtained with an inverted DMI-6000-BLeica wide-field fluorescent microscope equipped with 40× and 63×objectives and a Leica DFC 360FX camera. Confocal images were obtainedon a Leica TCS SPE microscope equipped with a 63× objective and thefollowing laser lines: 405 nm (for DAPI and Hoechst), 488 nm (for GFP andFITC), 532 nm (for mRFP, DsRED, Cy3, TRITC and TexasRed), 635 nm(for Cy5 and DRAQ5).

Microscopy imaging quantificationsImages taken with a 63× objective on an inverted DMI-6000-B Leica wide-field fluorescent microscope, equipped with a Leica DFC 360FX camera,were analyzed using the FIJI program as follows.

Spindle profile quantificationsSpindles were individually rotated in order to orient them with a horizontalpole-to-pole axis. A rectangle of a fixed size containing both spindle polesand centered on the metaphase plate was drawn on each image. The averagepixel intensity was measured for sequential vertical pixel lines, obtaining a

9

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

list of intensities, which was exported to an independent file. The intensitylists for all the cells analyzed in any given condition were analyzed using thesoftware Prism and plotted together. A line graph was generated with a lineshowing the average intensities in each vertical pixel line and the standarddeviations. In order to compare the protein intensities at the spindle poles,the length of the spindles was normalized by subtracting the central spindlevalues of the larger spindles, when necessary. Protein intensities werenormalized using the intensity of the tubulin (obtained by the samemethod).The graphs obtained for DnaJB6-silenced cells and control cells werecompared with a one-way ANOVA test to obtain the significance of thedifferences between the two conditions.

Quantification of fluorescence signal at spindle polesMetaphase HeLa cells were randomly selected and subsequentmeasurements were made with ImageJ software. The tubulin signal wasthresholded (by a rolling ball method with radius of 50 pixels; ImageJthreshold option), a selection on eachmitotic spindle polewas generated andtubulin total intensity. On the same selection, the total fluorescence of theprotein of interest was measured, corrected by means of backgroundsubtraction and finally normalized to the overall tubulin intensity.

Spindle poles width quantificationMetaphase HeLa cells were randomly selected. Centrosomes were labeledusing an anti-centrin and anti-tubulin antibodies. The centrin signal wasused to define the centrosome position. Subsequent measurements weremadewith ImageJ software. The tubulin signal was thresholded (by a rollingball method with radius of 50 pixels; ImageJ threshold option) and thedistance between the two closest spindle borders crossing the centrosomeposition was measured.

Quantification of centrosome misplacementCentrosomes were labeled using an anti-centrin antibody. Metaphase HeLacells were randomly selected. Metaphase plate axis was defined based on thealigned chromosomes. Subsequent measurements were made with ImageJsoftware. The centrosome axis was defined as a line connecting the twocentrosomes. In control cells, the two lines had an angle close to 90°. Amarkwas also placed on the centrosome-to-centrosome mid-point that falls inclose proximity to the metaphase plate axis in control cells. To quantifycentrosome displacement in DnaJB6i cells the angle (α) between thecentrosome and the metaphase plate axis was measured. To representcentrosome axis deviation, the α angle was subtracted from 90° andrepresented in a box-and-whisker plot. To quantify centrosomedisengagement from the spindle, the distance between the centrosome-to-centrosome mid-point and the metaphase plate was measured and plotted ina box-and-whisker plot.

Statistical analysisPRISM software was used for the statistical analysis. Depending on thenature of the data, a one-way or two-way ANOVA test, a Mann–Whitneytest, a Fisher’s exact test or a Student’s t-test was used. Specifically, one-wayANOVAwas used in Fig. 6C, two-way ANOVAwas used in the rest of thecases applying a Tukey’s multiple comparisons test in Fig. 5C and a Sidak’smultiple comparisons test in the rest of the cases. Normality distribution wastested (D’Agostino–Pearson omnibus normality test) for each sample whenapplicable, and equality of variances between samples in comparedconditions was tested when necessary. In box-and-whisker plots, boxesshow the upper and lower quartiles (25–75%) with a line at the median,whiskers extend from the 10–90th percentiles, and dots outside of whiskerranges correspond to outliers. The + indicates the mean values. Sample sizewas determined based on previous work done in the laboratory. Forquantifications, the investigator was blind to the sample allocation at themoment of counting.

AcknowledgementsWe thank S. Meunier for discussions and initial supervision of M.R.-S. We thankN. Mallol and L. Avila for excellent technical support. We thank the CRGmicroscopyfacility for technical support.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsConceptualization: M.R.-S., I.V.; Methodology: M.R.-S., I.V.; Validation: M.R.-S.;Formal analysis: M.R.-S.; Investigation: M.R.-S., J.S., A.M.; Resources: I.V.; Writing- original draft: M.R.-S., I.V.; Writing - review & editing: M.R.-S., J.S., A.M., I.V.;Visualization: M.R.-S., I.V.; Supervision: I.V.; Project administration: I.V.; Fundingacquisition: I.V.

FundingWe acknowledge the Spanish Ministry of Economy, Industry and Competitiveness(Ministerio de Economıa y Competitividad) to the EMBL partnership and supportof the Spanish Ministry of Economy and Competitiveness, ‘Centro de ExcelenciaSevero Ochoa’ as well as support of the CERCA Programme/Generalitat deCatalunya. M.R. was supported by a FPI fellowship from the Spanish Ministry ofEconomy BES-2013-064601. Work in the Vernos laboratory was supported bygrants from the Spanish Ministry of Economy and Competitiveness I+D (grantsBFU2012-37163 and BFU2015-68726-P). Deposited in PMC for immediate release.

Data availabilityThe Xenopus laevis sequence for DnaJB6 long isoform mRNA (DnaJB6-A) hasbeen submitted to GenBank, accession number: MK914533.

Supplementary informationSupplementary information available online athttp://jcs.biologists.org/lookup/doi/10.1242/jcs.227033.supplemental

ReferencesBlangy, A., Lane, H. A., d’Herin, P., Harper, M., Kress, M. and Nigg, E. A. (1995).

Phosphorylation by p34cdc2 regulates spindle association of human Eg5, akinesin-related motor essential for bipolar spindle formation in vivo. Cell 83,1159-1169. doi:10.1016/0092-8674(95)90142-6

Bonner, M. K., Poole, D. S., Xu, T., Sarkeshik, A., Yates, J. R., III and Skop, A. R.(2011). Mitotic spindle proteomics in Chinese hamster ovary cells. PLoS ONE 6,e20489. doi:10.1371/journal.pone.0020489

Carazo-Salas, R. E., Guarguaglini, G., Gruss, O. J., Segref, A., Karsenti, E. andMattaj, I.W. (1999).GenerationofGTP-boundRanbyRCC1 is required forchromatin-induced mitotic spindle formation. Nature 400, 178-181. doi:10.1038/22133

Cavazza, T. and Vernos, I. (2015). The RanGTP pathway: from nucleo-cytoplasmictransport to spindle assembly and beyond. Front Cell Dev Biol 3, 82. doi:10.3389/fcell.2015.00082

Cheng, X., Belshan, M. and Ratner, L. (2008). Hsp40 facilitates nuclear import ofthe human immunodeficiency virus type 2 Vpx-mediated preintegration complex.J. Virol. 82, 1229-1237. doi:10.1128/JVI.00540-07

Chuang, J. Z., Zhou, H., Zhu, M., Li, S. H., Li, X. J. and Sung, C. H. (2002).Characterization of a brain-enriched chaperone, MRJ, that inhibits Huntingtinaggregation and toxicity independently. J. Biol. Chem. 277, 19831-19838. doi:10.1074/jbc.M109613200

Cianfrocco, M. A., DeSantis, M. E., Leschziner, A. E. and Reck-Peterson, S. L.(2015). Mechanism and regulation of cytoplasmic dynein. Annu. Rev. Cell Dev.Biol. 31, 83-108. doi:10.1146/annurev-cellbio-100814-125438

Clarke, P. R. and Zhang, C. (2008). Spatial and temporal coordination of mitosis byRan GTPase. Nat. Rev. Mol. Cell Biol. 9, 464-477. doi:10.1038/nrm2410

Compton, D. A. and Cleveland, D. W. (1993). NuMA is required for the propercompletion of mitosis. J. Cell Biol. 120, 947-957. doi:10.1083/jcb.120.4.947

Culver-Hanlon, T. L., Lex, S. A., Stephens, A. D., Quintyne, N. J. and King, S. J.(2006). A microtubule-binding domain in dynactin increases dynein processivityby skating along microtubules. Nat. Cell Biol. 8, 264-270. doi:10.1038/ncb1370

Desai, A., Murray, A. W., Mitchison, T. J. and Walczak, C. E. (1999). The use ofXenopus egg extracts to study mitotic spindle assembly and function in vitro.Methods Cell Biol. 61, 385-412. doi:10.1016/S0091-679X(08)61991-3

Dey, S., Banerjee, P. and Saha, P. (2009). Cell cycle specific expression andnucleolar localization of human J-domain containing co-chaperone Mrj.Mol. Cell.Biochem. 322, 137-142. doi:10.1007/s11010-008-9950-y

di Pietro, F., Echard, A. and Morin, X. (2016). Regulation of mitotic spindleorientation: an integrated view. EMBO Rep. 17, 1106-1130. doi:10.15252/embr.201642292

Echeverri, C. J., Paschal, B. M., Vaughan, K. T. and Vallee, R. B. (1996).Molecular characterization of the 50-kD subunit of dynactin reveals function for thecomplex in chromosome alignment and spindle organization during mitosis.J. Cell Biol. 132, 617-633. doi:10.1083/jcb.132.4.617

Fang, C.-T., Kuo, H.-H., Pan, T. S., Yu, F.-C. and Yih, L. H. (2016). HSP70regulates the function of mitotic centrosomes. Cell. Mol. Life Sci. 73, 3949-3960.doi:10.1007/s00018-016-2236-8

Ferenz, N. P., Paul, R., Fagerstrom, C., Mogilner, A. and Wadsworth, P. (2009).Dynein antagonizes eg5 by crosslinking and sliding antiparallel microtubules.Curr. Biol. 19, 1833-1838. doi:10.1016/j.cub.2009.09.025

10

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

Florian, S. and Mayer, T. U. (2012). The functional antagonism between Eg5 anddynein in spindle bipolarization is not compatible with a simple push-pull model.Cell Rep 1, 408-416. doi:10.1016/j.celrep.2012.03.006

Gaglio, T., Saredi, A. and Compton, D. A. (1995). NuMA is required for theorganization of microtubules into aster-like mitotic arrays. J. Cell Biol. 131,693-708. doi:10.1083/jcb.131.3.693

Gaglio, T., Saredi, A., Bingham, J. B., Hasbani, M. J., Gill, S. R., Schroer, T. A.and Compton, D. A. (1996). Opposing motor activities are required for theorganization of the mammalian mitotic spindle pole. J. Cell Biol. 135, 399-414.doi:10.1083/jcb.135.2.399

Gaglio, T., Dionne, M. A. and Compton, D. A. (1997). Mitotic spindle poles areorganized by structural and motor proteins in addition to centrosomes. J. Cell Biol.138, 1055-1066. doi:10.1083/jcb.138.5.1055

Gorlich, D., Kostka, S., Kraft, R., Dingwall, C., Laskey, R. A., Hartmann, E. andPrehn, S. (1995). Two different subunits of importin cooperate to recognizenuclear localization signals and bind them to the nuclear envelope. Curr. Biol. 5,383-392. doi:10.1016/S0960-9822(95)00079-0

Gorlich, D., Pante, N., Kutay, U., Aebi, U. and Bischoff, F. R. (1996). Identificationof different roles for RanGDP and RanGTP in nuclear protein import. EMBO J. 15,5584-5594. doi:10.1002/j.1460-2075.1996.tb00943.x

Goshima, G., Wollman, R., Stuurman, N., Sholey, J. M. and Vale, R. D. (2005).Length control of the metaphase spindle.Curr. Biol. 22, 1979-1988. doi:10.1016/j.cub.2005.09.054

Grotjahn, D. A., Chowdhury, S., Xu, Y., McKenney, R. J., Schroer, T. A. andLander, G. C. (2018). Cryo-electron tomography reveals that dynactin recruits ateam of dyneins for processive motility.Nat. Struct. Mol. Biol. 25, 203-207. doi:10.1038/s41594-018-0027-7

Hageman, J., Rujano, M. A., van Waarde, M. A., Kakkar, V., Dirks, R. P.,Govorukhina, N., Oosterveld-Hut, H. M., Lubsen, N. H. and Kampinga, H. H.(2010).ADNAJBchaperonesubfamilywithHDAC-dependent activities suppressestoxic protein aggregation.Mol. Cell 37, 355-369. doi:10.1016/j.molcel.2010.01.001

Haren, L., Gnadt, N., Wright, M. and Merdes, A. (2009). NuMA is required forproper spindle assembly and chromosome alignment in prometaphase. BMCRes. Notes 2, 64. doi:10.1186/1756-0500-2-64

Heald, R., Tournebize, R., Blank, T., Sandaltzopoulos, R., Becker, P., Hyman, A.and Karsenti, E. (1996). Self-organization of microtubules into bipolar spindlesaround artificial chromosomes in Xenopus egg extracts. Nature 382, 420-425.doi:10.1038/382420a0

Heald, R., Tournebize, R., Habermann, A., Karsenti, E. and Hyman, A. (1997).Spindle assembly in Xenopus egg extracts: respective roles of centrosomes andmicrotubule self-organization. J. Cell Biol. 138, 615-628. doi:10.1083/jcb.138.3.615

Hueschen, C. L., Kenny, S. J., Xu, K. and Dumont, S. (2017). NuMA recruitsdynein activity to microtubule minus-ends at mitosis. Elife 6, R162. doi:10.7554/eLife.29328

Jha, R. and Surrey, T. (2015). Regulation of processive motion and microtubulelocalization of cytoplasmic dynein. Biochem. Soc. Trans. 43, 48-57. doi:10.1042/BST20140252

Jones, L. A,, Villemant, C., Starborg, T., Salter, A., Goddard, G., Ruane, P.,Woodman, P. G., Papalopulu, N. and Allan, V. J. (2014). Dynein lightintermediate chains maintain spindle bipolarity by functioning in centriolecohesion. J. Cell Biol. 207, 499-516. doi:10.1083/jcb.201408025

Kapitein, L. C., Peterman, E. J., Kwok, B. H., Kim, J. H., Kapoor, T. M. andSchmidt, C. F. (2005). The bipolar mitotic kinesin Eg5 moves on bothmicrotubules that it crosslinks. Nature 435, 114-118. doi:10.1038/nature03503

Kardon, J. R. and Vale, R. D. (2009). Regulators of the cytoplasmic dynein motor.Nat. Rev. Mol. Cell Biol. 10, 854-865. doi:10.1038/nrm2804

Kardon, J. R., Reck-Peterson, S. L. and Vale, R. D. (2009). Regulation of theprocessivity and intracellular localization of Saccharomyces cerevisiae dynein bydynactin.Proc.Natl. Acad.Sci.USA106, 5669-5674. doi:10.1073/pnas.0900976106

Kashina, A. S., Baskin, R. J., Cole, D. G., Wedaman, K. P., Saxton, W. M. andScholey, J. M. (1996). A bipolar kinesin. Nature 379, 270-272. doi:10.1038/379270a0

Kashina, A. S., Rogers, G. C. and Scholey, J. M. (1997). The bimC family ofkinesins: essential bipolar mitotic motors driving centrosome separation. Biochim.Biophys. Acta 1357, 257-271. doi:10.1016/S0167-4889(97)00037-2

Khodjakov, A., Copenagle, L., Gordon, M. B., Compton, D. A. and Kapoor, T. M.(2003). Minus-end capture of preformed kinetochore fibers contributes to spindlemorphogenesis. J. Cell Biol. 160, 671-683. doi:10.1083/jcb.200208143

King, S. J. and Schroer, T. A. (2000). Dynactin increases the processivity of thecytoplasmic dynein motor. Nat. Cell Biol. 2, 20-24. doi:10.1038/71338

King, S. J., Bonilla, M., Rodgers, M. E. and Schroer, T. A. (2002). Subunitorganization in cytoplasmic dynein subcomplexes. Protein Sci. 11, 1239-1250.doi:10.1110/ps.2520102

Kisurina-Evgenieva, O., Mack, G., Du, Q., Macara, I., Khodjakov, A. andCompton, D. A. (2004). Multiple mechanisms regulate NuMA dynamics at spindlepoles. J. Cell Sci. 117, 6391-6400. doi:10.1242/jcs.01568

Lu, M. S. and Johnston, C. A. (2013). Molecular pathways regulating mitotic spindleorientation in animal cells.Development 140, 1843-1856. doi:10.1242/dev.087627

Maiato, H. and Logarinho, E. (2014). Mitotic spindle multipolarity withoutcentrosome amplification. Nat. Cell Biol. 16, 386-394. doi:10.1038/ncb2958

Massey, A. J., Williamson, D. S., Browne, H., Murray, J. B., Dokurno, P., Shaw,T., Macias, A. T., Daniels, Z., Geoffroy, S., Dopson, M. et al. (2010). A novel,small molecule inhibitor of Hsc70/Hsp70 potentiates Hsp90 inhibitor inducedapoptosis in HCT116 colon carcinoma cells. Cancer Chemother. Pharmacol. 66,535-545. doi:10.1007/s00280-009-1194-3

Mayer, T. U., Kapoor, T. M., Haggarty, S. J., King, R. W., Schreiber, S. L. andMitchison, T. J. (1999). Small molecule inhibitor of mitotic spindle bipolarityidentified in a phenotype-based screen. Science 286, 971-974. doi:10.1126/science.286.5441.971

McAinsh, A. D., Meraldi, P., Draviam, V. M., Toso, A. andSorger, P. K. (2006). Thehuman kinetochore proteins Nnf1R and Mcm21R are required for accuratechromosome segregation.EMBOJ.25, 4033-4049. doi:10.1038/sj.emboj.7601293

McKinley, K. L. and Cheeseman, I. M. (2017). Large-scale analysis of CRISPR/Cas9 cell-cycle knockouts reveals the diversity of p53-dependent responses tocell-cycle defects. Dev. Cell 40, 405-420.e402. doi:10.1016/j.devcel.2017.01.012

Meng, E., Shevde, L. A. and Samant, R. S. (2016). Emerging roles and underlyingmolecular mechanisms of DNAJB6 in cancer. Oncotarget 7, 53984-53996.doi:10.18632/oncotarget.9803

Merdes, A., Ramyar, K., Vechio, J. D. and Cleveland, D. W. (1996). A complex ofNuMA and cytoplasmic dynein is essential for mitotic spindle assembly. Cell 87,447-458. doi:10.1016/S0092-8674(00)81365-3

Merdes, A., Heald, R., Samejima, K., Earnshaw, W. C. and Cleveland, D. W.(2000). Formation of spindle poles by dynein/dynactin-dependent transport ofNuMA. J. Cell Biol. 149, 851-862. doi:10.1083/jcb.149.4.851

Meunier, S. and Vernos, I. (2011). K-fibre minus ends are stabilized by a RanGTP-dependent mechanism essential for functional spindle assembly. Nat. Cell Biol.13, 1406-1414. doi:10.1038/ncb2372

Mitchison, T. J., Maddox, P., Gaetz, J., Groen, A., Shirasu, M., Desai, A.,Salmon, E. D. and Kapoor, T. M. (2005). Roles of polymerization dynamics,opposed motors, and a tensile element in governing the length of Xenopus extractmeiotic spindles. Mol. Biol. Cell 16, 3064-3076. doi:10.1091/mbc.e05-02-0174

Mitra, A., Fillmore, R. A., Metge, B. J., Rajesh, M., Xi, Y., King, J., Ju, J., Pannell,L., Shevde, L. A. and Samant, R. S. (2008). Large isoform of MRJ (DNAJB6)reduces malignant activity of breast cancer. Breast Cancer Res. 10, R22. doi:10.1186/bcr1874

Morales-Mulia, S. and Scholey, J. M. (2005). Spindle pole organization inDrosophila S2 cells by dynein, abnormal spindle protein (Asp), and KLP10A.Mol.Biol. Cell 16, 4462. doi:10.1091/mbc.e04-12-1110

O’Regan, L., Sampson, J., Richards, M. W., Knebel, A., Roth, D., Hood, F. E.,Straube, A., Royle, S. J., Bayliss, R. and Fry, A. M. (2015). Hsp72 is targeted tothe mitotic spindle by Nek6 to promote K-fiber assembly and mitotic progression.J. Cell Biol. 209, 349-358. doi:10.1083/jcb.201409151

Raaijmakers, J. A. andMedema, R. H. (2014). Function and regulation of dynein inmitotic chromosome segregation. Chromosoma 123, 407-422. doi:10.1007/s00412-014-0468-7

Raaijmakers, J. A., Tanenbaum, M. E. and Medema, R. H. (2013). Systematicdissection of dynein regulators in mitosis. J. Cell Biol. 201, 201-215. doi:10.1083/jcb.201208098

Rao, S. R., Flores-Rodriguez, N., Page, S. L., Wong, C., Robinson, P. J. andChircop, M. (2016). The clathrin-dependent spindle proteome. Mol. Cell.Proteomics 15, 2537-2553. doi:10.1074/mcp.M115.054809

Rosas-Salvans, M., Cavazza, T., Espadas, G., Sabido, E. and Vernos, I. (2018).Proteomic profiling of microtubule self-organization in M-phase. Mol. Cell.Proteomics 17, 1991-2004. doi:10.1074/mcp.RA118.000745

Ross, J. L., Wallace, K., Shuman, H., Goldman, Y. E. and Holzbaur, E. L. (2006).Processive bidirectional motion of dynein-dynactin complexes in vitro. Nat. CellBiol. 8, 562-570. doi:10.1038/ncb1421

Sauer, G., Korner, R., Hanisch, A., Ries, A., Nigg, E. A. and Sillje, H. H. W.(2005). Proteome analysis of the human mitotic spindle. Mol. Cell. Proteomics 4,35-43. doi:10.1074/mcp.M400158-MCP200

Schlecht, R., Scholz, S. R., Dahmen, H., Wegener, A., Sirrenberg, C., Musil, D.,Bomke, J., Eggenweiler, H. M., Mayer, M. P. and Bukau, B. (2013). Functionalanalysis of Hsp70 inhibitors. PLoS ONE 8, e78443. doi:10.1371/journal.pone.0078443

Seki, N., Hattori, A., Hayashi, A., Kozuma, S., Miyajima, N. and Saito, T. (1999).Cloning, tissue expression, and chromosomal assignment of human MRJ genefor a member of the DNAJ protein family. J. Hum. Genet. 44, 185-189. doi:10.1007/s100380050139

Silk, A. D., Holland, A. J. and Cleveland, D. W. (2009). Requirements for NuMA inmaintenance and establishment of mammalian spindle poles. J. Cell Biol. 184,677-690. doi:10.1083/jcb.200810091

Tan, R., Foster, P. J., Needleman, D. J. and McKenney, R. J. (2018). CooperativeAccumulation of dynein-dynactin at microtubule minus-ends drives microtubulenetwork reorganization. Dev. Cell 44, 233-247.e234. doi:10.1016/j.devcel.2017.12.023

Tanenbaum, M. E., Macurek, L., Galjart, N. and Medema, R. (2008). Dynein, Lis1and CLIP-170 counteract Eg5-dependent centrosome separation during bipolarspindle assembly. EMBO J. 27, 3235-3245. doi:10.1038/emboj.2008.242

11

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience

Tripathy, S. K., Weil, S. J., Chen, C., Anand, P., Vallee, R. B. and Gross, S. P.(2014). Autoregulatory mechanism for dynactin control of processive and diffusivedynein transport. Nat. Cell Biol. 16, 1192-1201. doi:10.1038/ncb3063

Walczak,C.E.,Vernos, I.,Mitchison,T.J.,Karsenti,E.andHeald,R. (1998).Amodelfor the proposed roles of different microtubule-based motor proteins in establishingspindle bipolarity. Curr. Biol. 8, 903-913. doi:10.1016/S0960-9822(07)00370-3

Whitehead, C. M. and Rattner, J. B. (1998). Expanding the role of HsEg5within the mitotic and post-mitotic phases of the cell cycle. J. Cell Sci. 111,2551-2561.

Wittmann, T. and Hyman, T. (1999). Recombinant p50/dynamitin as a tool toexamine the role of dynactin in intracellular processes. Methods Cell Biol. 61,137-143. doi:10.1016/S0091-679X(08)61978-0

12

RESEARCH ARTICLE Journal of Cell Science (2019) 132, jcs227033. doi:10.1242/jcs.227033

Journal

ofCe

llScience