2820 Research Article Opposing effects of Ndel1 and α1orα2 on cytoplasmic through competitive binding to Lis1

Chong Ding1, Xujun Liang2, Li Ma1, Xiaobing Yuan2 and Xueliang Zhu1,* 1Laboratory of Molecular Cell Biology, Institute of Biochemistry and Cell Biology and 2State Key Laboratory of Neurobiology, Institute of Neuroscience, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yue Yang Road, Shanghai 200031, China *Author for correspondence ([email protected])

Accepted 26 May 2009 Journal of Cell Science 122, 2820-2827 Published by The Company of Biologists 2009 doi:10.1242/jcs.048777

Summary Lis1 is an essential whose insufficiency causes aberrant dynein for Lis1 binding in a dose-dependent manner. neuronal positioning during neocortical development. It is Overexpression of α2 in developing rat brain repressed the believed to regulate both cytoplasmic dynein, a radial migration of neurons and mitotic progression of minus-end-directed motor, through direct interaction, and neuroprogenitors. By contrast, a Lis1-binding-defective point platelet-activating factor acetylhydrolase (PAF-AH) Ib by mutant, α2E39D, was ineffective in the above assays. These results complexing with the catalytic subunits α1 and α2. Although α1 indicate an antagonistic effect of α1, α2 and Ndel1 for Lis1 and α2 are highly expressed in brain, their deficiencies fail to binding, probably to modulate dynein functions in vivo. They cause brain abnormality. Here, we show that overexpression of also help to explain why brain development is particularly α2 or α1 results in inactivation of dynein characterized by Golgi sensitive to a decrease in Lis1 levels. and endosome dispersion and mitotic delay. Further overexpression of Lis1 or Ndel1, a Lis1- and dynein-binding protein that is also crucial for dynein function, restored Golgi Key words: Nudel, Cytoplasmic dynein, Mitosis, Neuronal migration, and endosome distribution. Biochemical assays showed that α1 Platelet-activating factor acetylhydrolase (PAF-AH) Ib complex, and especially α2, were able to compete against Ndel1 and Vesicle transport

Introduction functions as well through direct interactions with both DHC and

Journal of Cell Science Haploinsufficiency of Lis1 in humans causes type I , Lis1 (Liang et al., 2004; Liang et al., 2007; Ma et al., 2009; Sasaki a severe congenital disease characteristic of smooth brain surface et al., 2005; Shen et al., 2008; Shu et al., 2004; Stehman et al., owing to deficient neuronal migration during the development of 2007; Yan et al., 2003b). the central nervous system (CNS) (Gupta et al., 2002; Reiner et al., Lis1 also serves as the non-catalytic subunit of platelet-activating 1993). Lis1-deficient mice are embryonic lethal soon after factor acetylhydrolase (PAF-AH) Ib (Hattori et al., 1994b). PAF- implantation. Heterozygous mice exhibit delayed neuronal AHs are enzymes that inactivate PAF by removing its acetyl group. migration and abnormal cortical organization, whereas compound At least three types of PAF-AH: PAF-AH Ib, PAF-AH II and PAF- mice with further reduction of Lis1 manifest severe defects in AH plasma, have been characterized in mammals (Karasawa et al., cortical development (Gambello et al., 2003; Hirotsune et al., 1998). 2003). PAF-AH plasma and PAF-AH II are both monomeric Therefore, the neuronal function of Lis1 is sensitive to its dosage. polypeptides with ~41% sequence identity. By contrast, PAF-AH Why this occurs, however, is not clear. Ib is a heterotrimeric protein complex enriched in brain and testis Lis1, Nde1 (also called NudE), Ndel1 (also called Nudel for (Karasawa et al., 2003; Tjoelker and Stafforini, 2000). α1 (also NudE-like) and cytoplasmic dynein form an evolutionarily named Pafah1b3) and α2 (Pafah1b2), two homologous conserved genetic pathway in eukaryotes (Wynshaw-Boris and sharing ~63% sequence identity, are catalytic subunits of PAF-AH Gambello, 2001). Cytoplasmic dynein is a microtubule (MT)-based Ib, present either in the form of heterodimers or homodimers, and minus-end-directed motor composed of two heavy chains whereas Lis1 (Pafah1b1) is the regulatory subunit (Hattori et al., (DHC) and several intermediate chains (DIC), light intermediate 1994b; Karasawa et al., 2003; Tjoelker and Stafforini, 2000). α2 chains and light chains. It is widely involved in mitosis, intracellular is ubiquitously expressed, with highest expression levels in brain, trafficking, and cell migration (Dujardin et al., 2003; Gomes et al., whereas α1 is predominantly expressed in embryonic brain 2005; Hirokawa, 1998; Hook and Vallee, 2006; Karki and Holzbaur, (Koizumi et al., 2003; Manya et al., 1998). Unlike Lis1, however, 1999). Ndel1, Nde1 and Lis1 probably serve as positive regulators neither α1- nor α2-deficient mice show apparent neurological of dynein. Lis1 binds to dynein and is involved in dynein functions defects. Instead, α2-null or double-knockout mice are sterile in mitosis (Faulkner et al., 2000; Hebbar et al., 2008; Tai et al., because of defective spermatogenesis, whereas α1-null mice have 2002) and cell migration (Dujardin et al., 2003; Tsai et al., 2007). normal fertility possibly because of functional compensation by α2 Lis1 also interacts directly with the N-terminus of Nde1 or Ndel1 (Koizumi et al., 2003; Yan et al., 2003a). (Feng et al., 2000; Niethammer et al., 2000; Sasaki et al., 2000). Since both dynein and the PAF-AH Ib complex share Lis1 as Ndel1 (presumably Nde1 as well) is crucial for a variety of dynein their regulatory factor or subunit, they might exhibit functional Modulation of dynein function 2821

interplay by competing for Lis1, especially when wild-type Lis1 Disruption of dynein activity therefore results in their dispersion levels are attenuated by its . In this study, we or fragmentation (Burkhardt et al., 1997; Harada et al., 1998; Liang addressed such questions mainly by overexpressing α1 or α2 to et al., 2004). Indeed, whereas most WGA-positive vesicles were alter the stoichiometry of α-subunits and Lis1. We showed that concentrated at the MTOC in untransfected or GFP-expressing cells excess α1 or α2 can disrupt dynein activity through competitive (Fig. 1A, panels 1-2), they were dispersed in cells overexpressing interaction with Lis1 against dynein and Ndel1 in cultured GFP-α2 (Fig. 1A, panels 3-4). Similar effects were observed for mammalian cells. Moreover, overexpression of α2 in rat brain lysosomes labeled with Lysotracker (data not shown). By contrast, impaired neuronal migration during CNS development. overexpression of a Lis1-binding-defective point mutant (E39D) of Considering the presence of high α1 and α2 levels in brain, such α2 (Yamaguchi et al., 2007) failed to cause vesicle dispersion (Fig. an antagonistic interplay between dynein and PAF-AH Ib might 1A, panels 5-6), indicating a requirement for the Lis1-binding help to explain why Lis1 haploinsufficiency tends to disrupt activity of α2. Similarly, overexpression of GFP-α1 in Cos7 cells neuronal migration. also led to vesicle dispersion, although in a less-potent way (Fig. 1A, panels 7-8). These results indeed suggest inhibition of dynein- Results mediated vesicle transport by excess α1 and α2 in a Lis1-binding- Overexpression of α2 or α1 impairs retrograde vesicle dependent manner. transport α2 overexpression also led to disorganization of MT arrays in To investigate whether the catalytic subunits of PAF-AH Ib were Cos7 cells. Whereas clear radial MT arrays were visible in able to repress dynein functions by sequestering Lis1, we 67.8±2.5% of GFP-positive cells, the value was only 8.7±2.0% overexpressed α2 in Cos7 cells and examined the distribution of in cells overexpressing GFP-α2 (Fig. 1C-D). Nevertheless, the membrane organelles labeled with TRITC-conjugated wheatgerm vesicle dispersions were not fully correlated with MT agglutinin (WGA). WGA recognizes glycoproteins at the plasma organizations because WGA-positive vesicles were also found to membrane, endosome, and trans-Golgi cisternae (Liang et al., 2004; be dispersed in GFP-α2-expressing cells with radial MT arrays Raub et al., 1990; Virtanen et al., 1980). As endosomes and Golgi (Fig. 1E), possibly owing to different sensitivities of the MT cisternae are subjected to dynein-mediated transport towards MT organization and vesicle trafficking to dynein inactivation. As the minus-ends, they are usually enriched around the MT-organizing dynein motor is important for MT organization (Echeverri et al., center (MTOC) (Hirokawa, 1998; Karki and Holzbaur, 1999). 1996; Malikov et al., 2004; Quintyne et al., 1999), the influence Journal of Cell Science

Fig. 1. Effects of α2 or α1 overexpression on vesicle distribution and MT organization. (A,B) Incidence of vesicle dispersion in Cos7 cells transiently overexpressing GFP or the indicated GFP-tagged proteins. Cells were fixed in methanol and labeled with TRITC-WGA. Vesicles labeled with WGA are indicated by arrows in transfectants or arrowheads in untransfected cells. The insets in panels 3 and 7 show nuclear staining of corresponding transfectants. Error bars show s.d. **P<0.01 in Student’s t-tests. (C,D) MT organization in Cos7 cells overexpressing GFP or GFP-α2. Arrows indicate MT arrays in transfectants. Inset shows nuclear staining. Error bars indicate s.d. **P<0.01 in Student’s t-test. (E) Vesicle dispersion is seen in GFP-α2- positive cells with either abnormal (large arrows) or normal (small arrows) MT organization. Arrowheads indicate MT and vesicle organization in untransfected cells. 2822 Journal of Cell Science 122 (16)

of α2 overexpression on MT organization is also consistent with α2 and α1 can compete against dynein for Lis1 inactivation of dynein. Nevertheless, unlike CHO cells To further corroborate the above results, we examined whether α2 (Yamaguchi et al., 2007), Cos7 and HEK293T cells overexpressing or α1 overexpression was indeed able to sequester Lis1 from dynein. either α2 or α1 did not form pleiomorphic nuclei (Fig. 1A,C,E, FLAG-Lis1 was coexpressed in HEK293T cells with GFP, GFP-α1, insets; also see Fig. 2A). GFP-α2 or GFP-α2E39D. Coimmunoprecipitation indicated that the associations of endogenous dynein and dynactin with FLAG-Lis1 Overexpression of α2 results in a mitotic delay were dramatically reduced in the presence of exogenous α2 (Fig. 3A, Dynein has several roles in M phase, including spindle organization, lanes 1-2 and 4-6). Similar levels of GFP-α1, however, showed less movement and inactivation of the spindle checkpoint competition ability (Fig. 3A, lanes 2-3 and 5-6). By contrast, GFP- (Banks and Heald, 2001; Cleveland et al., 2003; Heald, 2000; Sharp α2E39D was not effective (Fig. 3B), therefore the competition effect et al., 2000). Inhibition of dynein activity thus creates a defective of α2 is due to its physical interaction with Lis1. These results are spindle (Echeverri et al., 1996; Gaglio et al., 1997; Liang et al., also consistent with the effects of these proteins on vesicle distribution 2007; Ma et al., 2009) and causes mitotic delay (Echeverri et al., and mitosis (Figs 1,2). 1996; Howell et al., 2001; Liang et al., 2007). To investigate whether α2 overexpression was able to cause phenotypes resembling dynein Lis1 overexpression restores retrograde vesicle transport in inactivation in M phase, we examined HEK293T cells because their cells overexpressing α2 efficient transfection and rapid proliferation facilitate identification We then investigated whether increasing Lis1 levels would rescue of mitotic transfectants. Overexpression of GFP alone did not affect defects caused by α2 overexpression. When FLAG-Lis1 was spindle morphology: 89.8±3.2% of transfectants in late coexpressed with GFP-α2 in Cos7 cells, WGA-positive vesicles prometaphase or metaphase had normal bipolar spindles (Fig. 2A, were no longer dispersed (Fig. 4A, panels 1-3; Fig. 4B). Statistical panels 1-3; Fig. 2B). By contrast, mitotic transfectants of GFP-α2 analysis indicated that, whereas 79.8±4.2% of cells overexpressing often contained spindles that were multipolar, unfocused, or GFP-α2 alone in the same populations exhibited vesicle dispersions distorted in shape (Fig. 2A, panels 4-9), similarly to previous (Fig. 4C), the value was only 18.4±6.0% in cells overexpressing observations in CHO cells (Yamaguchi et al., 2007). Of cells in late both GFP-α2 and FLAG-Lis1 (Fig. 4C). By contrast, FLAG- prometaphase or metaphase, 85.9±6.5% showed abnormal spindles luciferase was not effective (Fig. 4A, panels 5-7; Fig. 4B): (Fig. 2A, panels 4-9; Fig. 2B). Overexpression of GFP-α2E39D, 86.3±0.8% of cells had dispersed vesicles when GFP-α2 was however, failed to cause abnormal spindles (Fig. 2A, panels 10-12; expressed alone (Fig. 4C) compared with 88.4±5.5% when both Fig. 2B). Flow cytometry assays indicated a moderate G2-M block proteins were coexpressed (Fig. 4C) in the same populations. with increased mitotic index in cells overexpressing GFP-α2 (Fig. 2C-D). Further examination indicated that mitotic cells Ndel1 competes with α2 for Lis1 and is also able to restore α2- overexpressing GFP-α2 exhibited a marked enrichment in induced vesicle dispersion prometaphase (Fig. 2E), suggesting a delay in prometaphase- Ndel1 interacts directly with both Lis1 and dynein (Niethammer et metaphase transition. al., 2000; Sasaki et al., 2000). Tarricone and colleagues have used Journal of Cell Science

Fig. 2. Effects of α2 overexpression on mitosis. (A) Typical spindles in HEK293T cells overexpressing the indicated proteins. (B) Incidence of abnormal spindles in late prometaphase or metaphase cells. Error bars show s.d. **P<0.01 in Student’s t-tests. (C) Mitotic index of cells overexpressing the indicated protein. *P<0.05. (D) Cell cycle distributions for cells overexpressing the indicated protein. (E) Statistics for percentage of mitotic cells in prophase, prometaphase, metaphase, anaphase and telophase. Modulation of dynein function 2823

Consistently with previous reports (Manya et al., 1998; Sasaki et al., 2000; Yan et al., 2003b), both Ndel1 and α2 were highly expressed in brain, spinal cord and testis (Fig. 5D). Ndel1 might thus be used to neutralize the negative effect of α2 on dynein and Lis1 in cells rich in α2.

Overexpression of α2 in rat brain impairs neuronal positioning To explore the effect of α2 overexpression in vivo, we examined whether excessive α2 affects neuronal migration during development of neocortex. Cortical neuroprogenitor cells at the ventricular zone (VZ) of rat brains at embryonic day 16.5 were transfected with plasmids through in utero electroporation (Chen et al., 2008) to express GFP only, GFP and α2, or GFP and α2E39D, Fig. 3. Competition of α1 or α2 against dynein for binding Lis1. HEK293T respectively. At postnatal day 3, distributions of transfected neurons cells were transfected to coexpress FLAG-Lis1 with GFP or the indicated GFP were examined in coronal sections at the striatum level in the fusion proteins. After co-immunoprecipitation (IP) with anti-FLAG M2 beads, dorsolateral area of the neocortex. Neuronal positioning was samples were subjected to 3-12% gradient SDS-PAGE. Immunoblotting was quantified from seven matched sections (one section per rat). In then performed with antibodies against FLAG, GFP or the indicated subunits control brain sections, 65.3±5.0% of GFP-positive neurons had of dynein or dynactin, respectively. (A) Effect of α1 or α2 on Lis1-dynein interaction; (B) α2E39D fails to affect Lis1-dynein interaction. arrived at layers II-III of the cortical plate (Fig. 6A-B). By contrast, in brain sections transfected to express α2 plus GFP, only 24.4±8.3% of GFP-positive neurons had arrived at layers II-III; the majority of GFP-positive neurons (50.8±13.7%) remained in the lower layers, the N-terminus of Ndel1 (residues 58-169) and size-exclusion i.e. intermediate zone (IZ) and white matter (WM) (Fig. 6A-B). chromatography to show that excess α2 can disrupt the preformed Overexpression of α2E39D with GFP in neuroprogenitor cells had Ndel158-169-Lis1 complex by titrating Lis1 (Tarricone et al., 2004). a similar phenotype to the GFP-only controls: 67.9±9.6% of GFP- To investigate whether α2 was able to compete against full-length positive neurons had reached layers II-III by postnatal day 3 (Fig. Ndel1, we examined the association of endogenous Lis1 with 6A-B). These results indicate that excess α2 can impair neuronal FLAG-tagged full-length Ndel1 in the presence or absence of GFP- positioning during development of neocortex through interaction α2 by coimmunoprecipitation. Overexpression of GFP-α2 in with Lis1. HEK293T cells (Fig. 5A, lanes 1-2) indeed significantly reduced the interaction between FLAG-Ndel1 and endogenous Lis1 (lanes Overexpression of α2 in rat brain causes a moderate 3-4). Moreover, overexpression of FLAG-Ndel1 in Cos7 cells was enrichment of mitotic neuroprogenitor cells able to counteract the effect of GFP-α2 on vesicle distribution (Fig. To investigate whether overexpression of α2 affected proliferation 5B). The percentage of cells showing vesicle dispersion was of neuroprogenitor cells, rat brains were labeled with BrdU 48 hours α Journal of Cell Science 78.6±5.0% in those positive for GFP- 2 or 27.8±4.5% in those after electroporation and were fixed at embryonic day 19.5 (Chen overexpressing both GFP-α2 and FLAG-Ndel1 (Fig. 5C). et al., 2008). Coronal sections were immunostained with anti-BrdU Therefore, Ndel1 is able to counteract α2, presumably by increasing and anti-phospho-histone H3 antibodies (Fig. 7A). Statistical the pool of Lis1 associated with dynein. analyses were performed from four matched sections (one section

Fig. 4. Negative correlation of α2-induced vesicle dispersion with Lis1 levels. (A) Cos7 cells co-overexpressing GFP-α2 and FLAG- Lis1 or FLAG-luciferase were labeled with TRITC-WGA and anti- FLAG antibody. Cells with typical distributions of WGA-positive vesicles are shown. (B) Immunoblotting showing proper expression of the indicated exogenous proteins. Cos7 cells transfected to overexpress the indicated FLAG- or GFP-tagged proteins were subjected to immunoblotting. α-tubulin was detected as loading control. (C) Percentage of cells showing vesicle dispersion. GFP- α2-positive cells with or without FLAG-tagged Lis1 or luciferase in the same populations were both scored. Error bars indicate s.d. **P<0.01 in Student’s t-test. 2824 Journal of Cell Science 122 (16)

Fig. 6. Overexpression of α2 impairs the radial migration of cortical neurons in vivo. (A) Matched coronal sections of P3 rat brains overexpressing the indicated proteins. To highlight the transfection marker GFP, sections were immunostained with anti-GFP antibody (green) and counterstained for DNA with DAPI (red). Cortical layers are marked on the right. IZ, Intermediate Fig. 5. Negative correlation of α2-induced vesicle dispersion with Ndel1 zone; WM, white matter. (B) Statistics showing distributions of transfected Journal of Cell Science levels. (A) α2 competes against Ndel1 for interaction with endogenous Lis1. neurons in different cortical zones. Error bars indicate s.d. ***P<0.001 in HEK293T cells were transfected to coexpress FLAG-Ndel1 with GFP or GFP- Student’s t-tests. α2 and then subjected to coimmunoprecipitation (IP) with anti-FLAG M2 beads. (B,C) Ndel1 represses α2-induced vesicle dispersion. WGA staining of a typical Cos7 cell overexpressing both GFP-α2 and FLAG-Ndel1 is shown. GFP-α2-positive cells with or without FLAG-Ndel1 in the same population those of dynein inactivation in vesicle transport (Burkhardt et al., were both scored. Error bars indicate s.d. **P<0.01 in Student’s t-test. 1997; Liang et al., 2004), spindle organization (Echeverri et al., (D) Expression levels of the indicated proteins in different tissues of adult 1996; Gaglio et al., 1997; Liang et al., 2007; Ma et al., 2009) and β mice. -actin was detected as loading control. M-phase progression (Echeverri et al., 1996; Howell et al., 2001; Liang et al., 2007). As the Lis1 binding-defective mutant α2E39D was not effective (Figs 1-2), we conclude that α2 exerts the effects per rat) for each construct. The percentage of BrdU-positive cells through interaction with Lis1. Since Glu39 is conserved in both α1 expressing GFP alone, GFP and α2, or GFP and α2E39D in brain and α2 (Tarricone et al., 2004), we expect that the E39D sections was similar (Fig. 7B). By contrast, when mitotic cells were in α1 would also abolish both Lis1 binding and the dominant- visualized by immunostaining of phospho-histone H3 (Hans and negative effects of α1. We further demonstrated that overexpression Dimitrov, 2001), the mitotic index of GFP-α2-overexpressing cells of either α1 or α2 repressed association of dynein with Lis1. Again, (3.7±1.1%) was ~threefold higher than that of cells overexpressing such an effect of α2 depended on its interaction with Lis1 (Fig. 3). GFP alone (1.1±0.7%) or GFP-α2E39D (1.6±1.0%) (Fig. 7C). Furthermore, overexpression of Lis1 was able to repress vesicle Therefore, overexpression of α2 does not alter the progression of dispersion induced by α2 overexpression (Fig. 4). Therefore, the neuroprogenitor cells into S phase. Rather, its overexpression dominant-negative effects of α2 and α1 upon overexpression are causes a delay in the M-phase progression of neuroprogenitor cells, attributed to their sequestration of Lis1 from dynein and the which is consistent with results in HEK293T cells (Fig. 2). consequent inactivation of the dynein motor. α1 was less potent than α2 in both inducing vesicle dispersion and repressing the Lis1- Discussion dynein interaction (Figs 1 and 3), possibly because of its lower Our results strongly suggest a functional interplay between the affinity to Lis1 than α2 (Koizumi et al., 2003). dynein motor and the PAF-AH Ib complex, which compete against Yamaguchi and colleagues have previously shown that each other for Lis1 binding. We showed that overexpression of overexpression of murine α2 in CHO cells results in spindle human α1 or α2 (Figs 1-2) led to phenotypes closely resembling abnormality, nucleus pleiomorphy and amplification Modulation of dynein function 2825

Fig. 7. Effects of α2 overexpression on proliferation of neuroprogenitor cells. (A) Representative brain sections of E19.5 rat embryos showing immunostaining of GFP, BrdU and phospho-histone H3 (H3P). Arrows indicate a typical triple-positive cell, which is enlarged in insets. (B,C) Statistics for transfected cells positive for BrdU or H3P. Error bars indicate s.d. *P<0.05 and **P<0.01 in Student’s t-tests.

(Yamaguchi et al., 2007). We observed similar spindle phenotypes 2 and 7), it is not clear whether the improper neuronal positioning α Journal of Cell Science in HEK293T cells with human 2 and further showed that the (Fig. 6) is primarily due to defects in mitosis of neuronal progenitors prometaphase-metaphase transition is clearly delayed (Fig. 2). or migration. Nevertheless, these phenotypes are similar to those Nevertheless, we observed neither centrosome amplification in Cos7 of Lis1 insufficiency (Hirotsune et al., 1998; Pawlisz et al., 2008; cells nor clear changes in nuclear morphology in both Cos7 and Yingling et al., 2008). HEK293T cells overexpressing α2 (Figs 1-2; data not shown). Such Therefore, α2 and Ndel1 appear to serve as two competitive a difference might be due to the use of different cell lines. Moreover, reservoirs to pool Lis1. Their relative stoichiometries determine the our use of a GFP tag and human α2 might also contribute to ‘storage capacity’ of each reservoir. Increased Ndel1 levels tend to differences in certain phenotypes. Despite this, Yamaguchi and preserve more Lis1 to facilitate dynein functions, whereas increased colleagues also find that overexpression of α2 leads to more severe α2 levels tend to inactivate dynein by sequestering more Lis1. Such phenotypes than α1 (Yamaguchi et al., 2007). a model is not only compatible with results of current study, but α2 is also able to repress the Ndel1-Lis1 interaction. Using also helps to explain why Lis1 dosage is unusually important for column chromatography, Tarricone and colleagues show that α2 brain lamination (Gupta et al., 2002; Hirotsune et al., 1998): α1 can compete with Ndel158-169 for binding Lis1 (Tarricone et al., and α2 are highly expressed in brain (Koizumi et al., 2003; Manya 2004). We confirmed this using coimmunoprecipitation assays with et al., 1998) and are thus more likely to cause dynein inactivation full-length Ndel1 (Fig. 5A). Moreover, we further showed that when Lis1 levels become limited. Suppression of the hydrocephalus overexpression of Ndel1, similarly to Lis1 (Fig. 4), repressed vesicle phenotype in Lis1+/– Reln–/– mice by α2 deficiency (Assadi et al., dispersion induced by overexpression of α2 (Fig. 5), presumably 2008) and restoration of cortical MTs and dynein localization in by antagonizing α2 so that more Lis1 is recruited back to dynein. cells lacking Lis1 by Ndel1 (Yingling et al., 2008) are both Such an effect of Ndel1 probably explains why its expression levels consistent with the model. In addition to Lis1 insufficiency, the were positively correlated with α2 levels in many tissues (Fig. 5D): ‘overflow’ of Lis1, for example in the case of Lis1 overexpression, such a correlation might assure dynein functions and possibly other also has effects on MT dynamics and mitosis (Faulkner et al., 2000; PAF-AH-Ib-independent Lis1 functions in cells with high expression Sapir et al., 1997). Reduction of Lis1 levels has been shown to levels of α2 and/or α1. restore sterility in mice induced by α1 or α2 deficiency (Yan et al., Consistent with the above studies in cultured cells, we further 2003a), possibly by eliminating the overflow of Lis1. α1 is not as demonstrated that overexpression of α2, but not α2E39D, abolished potent as α2 (Figs 1 and 3) (Assadi et al., 2008; Yamaguchi et al., proper neuronal positioning during development of neocortex (Fig. 2007), probably because α2 has higher affinity to Lis1 and is more 6). As α2 overexpression also delayed M phase progression (Figs widely expressed (Koizumi et al., 2003; Manya et al., 1998; Yan 2826 Journal of Cell Science 122 (16)

et al., 2003a). Because Nde1 levels also affect phenotypes of Lis1 In utero electroporation insufficiency (Pawlisz et al., 2008), it might also function in a In utero electroporation was performed as described previously (Chen et al., 2008). Briefly, pCAGGS-IRES-EGFP, pCAGGS-α2-IRES-EGFP or pCAGGS-α2E39D- redundant manner with Ndel1 in forming a pool of Lis1. IRES-EGFP was mixed with pEYFP (Clontech) at a ratio of 2:1 before transfection Lis1 is in fact more likely to be a ‘content’of α1 and α2 reservoirs to enhance fluorescent signal. Multiparous Sprague-Dawley rats at 16.5 days of than the regulatory subunit for the enzyme activity of PAF-AH Ib. gestation were anesthetized intraperitoneally with 10% chloral hydrate (3.5 ml per kg body weight). Uteruses were exposed and then 15-20 μg of plasmid mixed with Although Lis1 has been considered to be the regulatory subunit, Fast Green (2 mg/ml; Sigma) was injected by trans-uterus pressure microinjection homodimers or heterodimers of α1 and α2 are sufficient for PAF- into the lateral ventricle of embryos. Electric pulses were generated by an AH activity (Hattori et al., 1994a; Manya et al., 1999). Moreover, ElectroSquireportator T830 (BTX) and applied to the cerebral wall at five repeats of stoichiometric amounts of Lis1 fail to significantly alter the catalytic 60 V for 50 milliseconds, with an interval of 100 milliseconds. Effects on neuronal migration were examined at postnatal day 3. To label proliferating cells, BrdU (20 activity of PAF-AH Ib (Manya et al., 1999). Consistently, structural mg/ml; Sigma) was injected intraperitoneally twice at a 30 minute intervals 48 hours analyses also suggest that Lis1 binding is unlikely to lead to major after in utero electroporation at 150 mg per kg body weight and examined after a conformational changes in the α2-α2 dimer (Tarricone et al., 2004). further 24 hours. Nevertheless, whether cells use α1 or α2 to dynamically regulate For histochemical analysis, brains were removed after transcardial perfusion, fixed in 4% paraformaldehyde, and then cryopreserved in OCT compound. Coronal cryostat dynein activity, for example, by acutely altering their expression sections of 30 μm thickness were cut on a freezing microtome and then subjected to levels, regional densities and/or affinities to Lis1, awaits further immunostaining as described (Chen et al., 2008). investigation. The authors thank Qiongping Huang and Lele Xie for technical Materials and Methods assistance, and Naihe Jing, Fubin Wang and Yidong Shen for critical Plasmids comments on experiment design. We are grateful to Orly Reiner (The To express FLAG-tagged or green fluorescent protein (GFP)-tagged human α1 or Weizmann Institute of Science, Israel) for kindly providing Lis1 α2, the full-length cDNAs were obtained by RT-PCR and subcloned into pUHD30F cDNA. This work was supported by National Science Foundation of (Zhu, 1999) and pEGFP-C1 (Clontech). Point mutant of α2 was obtained by PCR. China (30721065, 30830060, 30623003, and 30771076), Ministry of Plasmids for expressing FLAG-tagged Ndel1 and Lis1, and GFP-tagged Ndel1 were Science and Technology of China (2005CB522701, 2006CB943900, described previously (Liang et al., 2004; Shen et al., 2008; Yan et al., 2003b). Firefly luciferase cDNA was amplified by PCR from pGL3-Basic vector (Promega) and 2007CB914501, and 2007CB947100), Science and Technology inserted into pUHD30F to express FLAG-Luc as a control. For in utero electroporation, Commission of Shanghai Municipality (08XD14048 and 088014199), pCAGGS-IRES-EGFP (Chen et al., 2008) was used to construct plasmids to and Chinese Academy of Sciences (KSCX2-YW-R-108). coexpress GFP as a marker with wild-type or mutant α2 through the internal ribosome entry site (IRES). References Assadi, A. H., Zhang, G., McNeil, R., Clark, G. D. and D’Arcangelo, G. (2008). Pafah1b2 Antibodies and other staining reagents mutations suppress the development of hydrocephalus in compound Pafah1b1; Reln and α TRITC-WGA and mouse antibodies against -tubulin, FLAG M2, DIC, BrdU were Pafah1b1; Dab1 mutant mice. Neurosci. Lett. 439, 100-105. glued purchased from Sigma. Anti-p150 and p50 mAbs were from BD Biosciences (San Banks, J. D. and Heald, R. (2001). Chromosome movement: dynein-out at the kinetochore. Diego, CA). Rabbit polyclonal antibodies to DHC and GFP was from Santa Cruz Curr. Biol. 11, R128-R131. Biotechnology (Santa Cruz, CA). Rabbit antibodies to Lis1 and phospho-histone H3 Burkhardt, J. K., Echeverri, C. J., Nilsson, T. and Vallee, R. B. (1997). Overexpression (Ser10) and chicken anti-α2 IgY were from Abcam (Cambridge, UK). Chicken anti- of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent Ndel1 IgY was prepared as described (Liang et al., 2007). Secondary antibodies labeled maintenance of membrane organelle distribution. J. Cell Biol. 139, 469-484. with Alexa Fluor 488, 546 and 647 were purchased from Invitrogen (Carlsbad, CA). Chen, G., Sima, J., Jin, M., Wang, K. Y., Xue, X. J., Zheng, W., Ding, Y. Q. and Yuan,

Journal of Cell Science X. B. (2008). Semaphorin-3A guides radial migration of cortical neurons during Cell culture and transfection development. Nat. Neurosci. 11, 36-44. Human embryonic kidney (HEK) 293T and monkey kidney Cos7 cells were cultured Cleveland, D. W., Mao, Y. and Sullivan, K. F. (2003). Centromeres and kinetochores: in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen) supplemented with from epigenetics to mitotic checkpoint signaling. Cell 112, 407-421. 10% (vol./vol.) bovine serum (Sijiqing Company, Hangzhou, China) in an atmosphere Dujardin, D. L., Barnhart, L. E., Stehman, S. A., Gomes, E. R., Gundersen, G. G. and Vallee, R. B. (2003). A role for cytoplasmic dynein and LIS1 in directed cell containing 5% CO2. Cell transfection was performed using the conventional calcium phosphate method. movement. J. Cell Biol. 163, 1205-1211. 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 the complex in Fluorescence staining and microscopy chromosome alignment and spindle organization during mitosis. J. Cell Biol. 132, 617- Cells grown on glass coverslips were fixed in cold methanol for 5 minutes and then 633. kept in 75% ethanol. Proper antibody combinations were chosen for multicolor Faulkner, N. E., Dujardin, D. L., Tai, C. Y., Vaughan, K. T., O’Connell, C. B., Wang, immunostaining. Nuclear DNA was stained with 4,6-diamidino-2-phenylindole Y. and Vallee, R. B. (2000). A role for the lissencephaly gene LIS1 in mitosis and (DAPI). Trans-Golgi cisternae and endosomes were decorated with TRITC-WGA cytoplasmic dynein function. Nat. Cell Biol. 2, 784-791. (Liang et al., 2004). Images were captured using a cooled CCD camera (SPOT II, Feng, Y., Olson, E. C., Stukenberg, P. T., Flanagan, L. A., Kirschner, M. W. and Walsh, Diagnostic Instruments, MI) on an Olympus BX51 microscope or a Leica TCS SP2 C. A. (2000). LIS1 regulates CNS lamination by interacting with mNudE, a central laser confocal microscope. For confocal microscopy, sections in Z-series were scanned component of the centrosome. Neuron 28, 665-679. at 0.1-0.2 μm intervals. Z-stack images were then formed by maximal projection. Gaglio, T., Dionne, M. A. and Compton, D. A. (1997). Mitotic spindle poles are organized Statistical results were obtained from two or more independent experiments. by structural and motor proteins in addition to . J. Cell Biol. 138, 1055- 1066. Coimmunoprecipitation and immunoblotting Gambello, M. J., Darling, D. L., Yingling, J., Tanaka, T., Gleeson, J. G. and Wynshaw- Coimmunoprecipitation using anti-FLAG M2 affinity resin (Sigma) was performed Boris, A. (2003). Multiple dose-dependent effects of Lis1 on cerebral cortical as described previously (Yan et al., 2003b). Protein samples were resolved by SDS- development. J. Neurosci. 23, 1719-1729. PAGE and transferred to nitrocellulose membranes (Whatman Schleicher and Schuell, Gomes, E. R., Jani, S. and Gundersen, G. G. (2005). Nuclear movement regulated by Dassel, Germany). Immunoblots were then developed in Western Lightning Cdc42, MRCK, myosin, and actin flow establishes MTOC polarization in migrating Chemiluminescence Reagent Plus (PerkinElmer, Boston, MA) and exposed to X-ray cells. Cell 121, 451-463. Gupta, A., Tsai, L. H. and Wynshaw-Boris, A. (2002). Life is a journey: a genetic look films (Kodak, Rochester, NY). at neocortical development. Nat. Rev. Genet. 3, 342-355. Hans, F. and Dimitrov, S. (2001). Histone H3 phosphorylation and cell division. Oncogene Flow cytometry 20, 3021-3027. α α E39D HEK293T cells overexpressing GFP- 2 or GFP- 2 were fixed with 2% Harada, A., Takei, Y., Kanai, Y., Tanaka, Y., Nonaka, S. and Hirokawa, N. (1998). paraformaldehyde in PBS for 15 minutes 48 hours after transfection. Cells were Golgi vesiculation and lysosome dispersion in cells lacking cytoplasmic dynein. J. Cell resuspended in ice-cold 70% ethanol for 30 minutes and then stained with propidium Biol. 141, 51-59. iodide (50 μg/ml) in the presence of RNase A (200 μg/ml). Samples were analyzed Hattori, M., Adachi, H., Tsujimoto, M., Arai, H. and Inoue, K. (1994a). The catalytic using fluorescence-activated cell sorter (BD Biosciences). The distributions subunit of bovine brain platelet-activating factor acetylhydrolase is a novel type of serine of GFP-positive cells are presented. esterase. J. Biol. Chem. 269, 23150-23155. Modulation of dynein function 2827

Hattori, M., Adachi, H., Tsujimoto, M., Arai, H. and Inoue, K. (1994b). Miller-Dieker Reiner, O., Carrozzo, R., Shen, Y., Wehnert, M., Faustinella, F., Dobyns, W. B., Caskey, lissencephaly gene encodes a subunit of brain platelet-activating factor acetylhydrolase C. T. and Ledbetter, D. H. (1993). Isolation of a Miller-Dieker lissencephaly gene [corrected]. Nature 370, 216-218. containing G protein beta-subunit-like repeats. Nature 364, 717-721. Heald, R. (2000). Motor function in the mitotic spindle. Cell 102, 399-402. Sapir, T., Elbaum, M. and Reiner, O. (1997). Reduction of microtubule catastrophe events Hebbar, S., Mesngon, M. T., Guillotte, A. M., Desai, B., Ayala, R. and Smith, D. S. by LIS1, platelet-activating factor acetylhydrolase subunit. EMBO J. 16, 6977-6984. (2008). Lis1 and Ndel1 influence the timing of nuclear envelope breakdown in neural Sasaki, S., Shionoya, A., Ishida, M., Gambello, M. J., Yingling, J., Wynshaw-Boris, stem cells. J. Cell Biol. 182, 1063-1071. A. and Hirotsune, S. (2000). A LIS1/NUDEL/cytoplasmic dynein heavy chain complex Hirokawa, N. (1998). Kinesin and dynein superfamily proteins and the mechanism of in the developing and adult nervous system. Neuron 28, 681-696. organelle transport. Science 279, 519-526. Sasaki, S., Mori, D., Toyo-oka, K., Chen, A., Garrett-Beal, L., Muramatsu, M., Hirotsune, S., Fleck, M. W., Gambello, M. J., Bix, G. J., Chen, A., Clark, G. D., Miyagawa, S., Hiraiwa, N., Yoshiki, A., Wynshaw-Boris, A. et al. (2005). Complete Ledbetter, D. H., McBain, C. J. and Wynshaw-Boris, A. (1998). Graded reduction loss of Ndel1 results in neuronal migration defects and early embryonic lethality. Mol. of Pafah1b1 (Lis1) activity results in neuronal migration defects and early embryonic Cell. Biol. 25, 7812-7827. lethality. Nat. Genet. 19, 333-339. Sharp, D. J., Rogers, G. C. and Scholey, J. M. (2000). Microtubule motors in mitosis. Hook, P. and Vallee, R. B. (2006). The dynein family at a glance. J. Cell Sci. 119, 4369- Nature 407, 41-47. 4371. Shen, Y., Li, N., Wu, S., Zhou, Y., Shan, Y., Zhang, Q., Ding, C., Yuan, Q., Zhao, F., Howell, B. J., McEwen, B. F., Canman, J. C., Hoffman, D. B., Farrar, E. M., Rieder, Zeng, R. et al. (2008). Nudel binds Cdc42GAP to modulate Cdc42 activity at the leading C. L. and Salmon, E. D. (2001). Cytoplasmic dynein/dynactin drives kinetochore protein edge of migrating cells. Dev. Cell 14, 342-353. transport to the spindle poles and has a role in mitotic spindle checkpoint inactivation. Shu, T., Ayala, R., Nguyen, M. D., Xie, Z., Gleeson, J. G. and Tsai, L. H. (2004). Ndel1 J. Cell Biol. 155, 1159-1172. operates in a common pathway with LIS1 and cytoplasmic dynein to regulate cortical Karasawa, K., Harada, A., Satoh, N., Inoue, K. and Setaka, M. (2003). Plasma platelet neuronal positioning. Neuron 44, 263-277. activating factor-acetylhydrolase (PAF-AH). Prog. Lipid Res. 42, 93-114. Stehman, S. A., Chen, Y., McKenney, R. J. and Vallee, R. B. (2007). NudE and NudEL Karki, S. and Holzbaur, E. L. (1999). Cytoplasmic dynein and dynactin in cell division are required for mitotic progression and are involved in dynein recruitment to and intracellular transport. Curr. Opin. Cell Biol. 11, 45-53. kinetochores. J. Cell Biol. 178, 583-594. Koizumi, H., Yamaguchi, N., Hattori, M., Ishikawa, T. O., Aoki, J., Taketo, M. M., Tai, C. Y., Dujardin, D. L., Faulkner, N. E. and Vallee, R. B. (2002). Role of dynein, Inoue, K. and Arai, H. (2003). Targeted disruption of intracellular type I platelet dynactin, and CLIP-170 interactions in LIS1 kinetochore function. J. Cell Biol. 156, activating factor-acetylhydrolase catalytic subunits causes severe impairment in 959-968. spermatogenesis. J. Biol. Chem. 278, 12489-12494. Tarricone, C., Perrina, F., Monzani, S., Massimiliano, L., Kim, M. H., Derewenda, Z. Liang, Y., Yu, W., Li, Y., Yang, Z., Yan, X., Huang, Q. and Zhu, X. (2004). Nudel S., Knapp, S., Tsai, L. H. and Musacchio, A. (2004). Coupling PAF signaling to dynein functions in membrane traffic mainly through association with Lis1 and cytoplasmic regulation: structure of LIS1 in complex with PAF-acetylhydrolase. Neuron 44, 809- dynein. J. Cell Biol. 164, 557-566. 821. Liang, Y., Yu, W., Li, Y., Yu, L., Zhang, Q., Wang, F., Yang, Z., Du, J., Huang, Q., Tjoelker, L. W. and Stafforini, D. M. (2000). Platelet-activating factor acetylhydrolases Yao, X. et al. (2007). Nudel modulates kinetochore association and function of in health and disease. Biochim. Biophys. Acta 1488, 102-123. cytoplasmic dynein in M phase. Mol. Biol. Cell 18, 2656-2666. Tsai, J. W., Bremner, K. H. and Vallee, R. B. (2007). Dual subcellular roles for LIS1 Ma, L., Tsai, M. Y., Wang, S., Lu, B., Chen, R., Iii, J. R., Zhu, X. and Zheng, Y. (2009). and dynein in radial neuronal migration in live brain tissue. Nat. Neurosci. 10, 970- Requirement for Nudel and dynein for assembly of the lamin B spindle matrix. Nat. 979. Cell Biol. 11, 247-256. Virtanen, I., Ekblom, P. and Laurila, P. (1980). Subcellular compartmentalization of Malikov, V., Kashina, A. and Rodionov, V. (2004). Cytoplasmic dynein nucleates saccharide moieties in cultured normal and malignant cells. J. Cell Biol. 85, 429- to organize them into radial arrays in vivo. Mol. Biol. Cell 15, 2742-2749. 434. Manya, H., Aoki, J., Watanabe, M., Adachi, T., Asou, H., Inoue, Y., Arai, H. and Wynshaw-Boris, A. and Gambello, M. J. (2001). LIS1 and dynein motor function in Inoue, K. (1998). Switching of platelet-activating factor acetylhydrolase catalytic subunits neuronal migration and development. Dev. 15, 639-651. in developing rat brain. J. Biol. Chem. 273, 18567-18572. Yamaguchi, N., Koizumi, H., Aoki, J., Natori, Y., Nishikawa, K., Natori, Y., Manya, H., Aoki, J., Kato, H., Ishii, J., Hino, S., Arai, H. and Inoue, K. (1999). Takanezawa, Y. and Arai, H. (2007). Type I platelet-activating factor acetylhydrolase Biochemical characterization of various catalytic complexes of the brain platelet- catalytic subunits over-expression induces pleiomorphic nuclei and centrosome activating factor acetylhydrolase. J. Biol. Chem. 274, 31827-31832. amplification. Genes Cells 12, 1153-1161. Niethammer, M., Smith, D. S., Ayala, R., Peng, J., Ko, J., Lee, M. S., Morabito, M. Yan, W., Assadi, A. H., Wynshaw-Boris, A., Eichele, G., Matzuk, M. M. and Clark, and Tsai, L. H. (2000). NUDEL is a novel Cdk5 substrate that associates with LIS1 G. D. (2003a). Previously uncharacterized roles of platelet-activating factor and cytoplasmic dynein. Neuron 28, 697-711. acetylhydrolase 1b complex in mouse spermatogenesis. Proc. Natl. Acad. Sci. USA 100,

Journal of Cell Science Pawlisz, A. S., Mutch, C., Wynshaw-Boris, A., Chenn, A., Walsh, C. A. and Feng, Y. 7189-7194. (2008). Lis1-Nde1-dependent neuronal fate control determines cerebral cortical size and Yan, X., Li, F., Liang, Y., Shen, Y., Zhao, X., Huang, Q. and Zhu, X. (2003b). Human lamination. Hum. Mol. Genet. 17, 2441-2455. Nudel and NudE as regulators of cytoplasmic dynein in poleward protein transport along Quintyne, N. J., Gill, S. R., Eckley, D. M., Crego, C. L., Compton, D. A. and Schroer, the mitotic spindle. Mol. Cell. Biol. 23, 1239-1250. T. A. (1999). Dynactin is required for microtubule anchoring at centrosomes. J. Cell Yingling, J., Youn, Y. H., Darling, D., Toyo-Oka, K., Pramparo, T., Hirotsune, S. and Biol. 147, 321-334. Wynshaw-Boris, A. (2008). Neuroepithelial stem cell proliferation requires LIS1 for Raub, T. J., Koroly, M. J. and Roberts, R. M. (1990). Endocytosis of wheat germ precise spindle orientation and symmetric division. Cell 132, 474-486. agglutinin binding sites from the cell surface into a tubular endosomal network. J. Cell Zhu, X. (1999). Structural requirements and dynamics of mitosin-kinetochore interaction Physiol. 143, 1-12. in M phase. Mol. Cell. Biol. 19, 1016-1024.