A transgenic zebrafish for monitoring in vivo ... - Wiley Online Library

DEVELOPMENTAL DYNAMICS 239:2695–2699, 2010

a

TECHNIQUES

A Transgenic Zebrafish for Monitoring In Vivo Microtubule Structures

Developmental Dynamics

Kazuhide Asakawa and Koichi Kawakami*

The microtubule (MT) cytoskeleton plays crucial roles in brain development by regulating the proliferation of neuronal progenitor cells, neuronal migration and axon guidance. Methods for monitoring MT in the intact brain, however, have been limited in vertebrates. Here, we report a transgenic zebrafish line for monitoring MT in vivo. This reporter line carries a transgene encoding the green fluorescent protein (GFP) -tagged tubulin gene linked to the upstream activating sequence (UAS), the recognition sequence of the yeast Gal4 transcriptional activator. By crossing this reporter line with appropriate transgenic Gal4 driver lines, we induced the GFP-tagged tubulin in various cell types from the embryonic stages to the adult stage. In larvae expressing the modified tubulin, individual MT filaments and other MT structures, including the mitotic spindles in proliferating neuronal progenitor cells, were clearly visualized. Therefore, the transgenic UAS reporter line should be useful for directly monitoring MTs in the intact brain. Developmental Dynamics 239:2695–2699, 2010. V 2010 Wiley-Liss, Inc. C

Key words: tubulin; hindbrain; Gal4-UAS; mitotic spindle; live imaging; vertebrate Accepted 16 July 2010

INTRODUCTION The development of the brain is dependent on the microtubule (MT) cytoskeleton, which plays key roles in the proliferation of neuronal progenitor cells, neuronal migration and axon guidance (Keays et al., 2007; Poirier et al., 2007; Jaglin et al., 2009; Tischfield et al., 2010). In vertebrates, the expression of the fluorescently labeled tubulin by microinjection into fertilized eggs has been used to directly visualize MTs in the intact central nervous system, and revealed the tubulin dynamics in the axons of sensory and motor neurons (Takeda et al., 1995). However, the application of this approach was restricted to earlier developmental stages, because the signals of the fluo-

rescent tubulin were transient. Thus, to understand the functional and regulatory mechanisms of MT in the brain, it is crucial to develop a reporter construct in which MTs are reproducibly labeled in a population of cells at any developmental stage. We have been developing a GAL4UAS method for gene and enhancer trapping in zebrafish and established transgenic fish lines expressing Gal4FF, a variant of the yeast Gal4 transcriptional activator, in a cell-type specific manner (Brand and Perrimon, 1993; Asakawa and Kawakami, 2008, 2009; Asakawa et al., 2008). In the present study, we used this technique to express a genetically encoded fluorescent tubulin (Carminati and

Stearns, 1997; Grieder et al., 2000). We constructed a transgenic reporter line in which a green fluorescent protein (GFP) -tagged tubulin gene was placed downstream of the upstream activating sequence (UAS), the target sequence of Gal4. When the UAS reporter line was crossed with a transgenic line expressing Gal4FF, the GFP-tagged tubulin was induced in Gal4FF-expressing cells during embryonic, larval and adult stages. We demonstrated that, by the expression of GFP-tagged tubulin, single MT filaments and other MT structures, including the mitotic spindles in proliferating neuronal progenitor cells, could be directly monitored in real time.

Additional Supporting Information may be found in the online version of this article. Division of Molecular and Developmental Biology, National Institute of Genetics, Department of Genetics, Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan Grant sponsor: the Ministry of Education, Culture, Sports, Science, and Technology of Japan. *Correspondence to: Koichi Kawakami, Division of Molecular and Developmental Biology, National Institute of Genetics, Department of Genetics, Graduate University for Advanced Studies (SOKENDAI), 1111 Yata, Mishima, Shizuoka 411-8540, Japan. E-mail: [email protected] DOI 10.1002/dvdy.22400 Published online 24 August 2010 in Wiley Online Library (wileyonlinelibrary.com).

C 2010 Wiley-Liss, Inc. V

2696 ASAKAWA AND KAWAKAMI


RESULTS AND DISCUSSION To construct the GFP-tagged tubulin gene, we chose the a-tubulin gene tuba2, which encodes one of the subunits of the tubulin heterodimer and highly conserved among species. The tuba2 cDNA fragment was fused to the EGFP gene at the NH2-terminus (Gtuba2) and placed downstream of five tandem UAS repeats (Fig. 1A). The resulting UAS:Gtuba2 construct was introduced into the Tol2 transposable element, generating the T2ZUAS:Gtuba2 cassette. To create a genomic insertion of T2ZUAS:Gtuba2, we first injected the plasmid carrying the T2ZUAS:Gtuba2 cassette into one-cell stage embryos with synthesized Tol2 transposase, and these embryos were raised to adulthood. Then, we crossed the injected fish with SAGFF73A fish, a transgenic fish expressing Gal4FF (Asakawa et al., 2008) ubiquitously from the zfand5b promoter (Asakawa and Kawakami, 2009) and investigated GFP expression in the resulting F1 embryos. We found that six of the seven injected fish produced F1 offspring expressing GFP throughout the whole body when crossed with SAGFF73A (Fig. 1B). Of interest, the intensity of the GFP signal varied among these GFP-positive F1 embryos. Therefore, we selected F1 embryos from a founder fish showing the strongest GFP fluorescence and raised them to adulthood. By southern blot hybridization, we found that each of these GFP-expressing F1 fish contained a single T2ZUAS:Gtuba2 (data not shown). Adaptor ligation polymerase chain reaction (PCR) revealed that, in the F1 fish, the T2ZUAS:Gtuba2 was integrated in an intergenic region on chromosome 2. To investigate whether the modified tubulin was functional, we performed several tests. First, we examined SAGFF73A;UAS:Gtuba2 larvae by confocal microscopy. We found that, in the skin cells of these larvae, filamentous structures highlighted by the GFP fluorescence were detected (Fig. 1C). Furthermore, when we injected a plasmid carrying mRFP1-tagged histone H2A linked to UAS into the SAGFF73A;UAS:Gtuba2 embryos, the bipolar mitotic spindles associated with the condensed chromosomes la-

beled with RFP were clearly detected (Fig. 1D). These observations indicated that Gtuba2 could be incorporated into MT filaments. The expression of UAS:Gtuba2 did not lead to instability of the MT network, as an elaborate MT network was observed in the SAGFF73A;UAS:Gtuba2 larvae comparable to that of the wild type, by immunolabling with anti-tubulin antibody (Supp. Fig. S1, which is available online). Second, we investigated the effect of Gtuba2 expression on development and viability. We found that early morphogenetic processes such as gastrulation and neurulation were not affected in SAGFF73A;UAS:Gtuba2 embryos, as judged from the Convergent-extension index (Angers et al., 2006) and the apical positioning of the centrosome, the major MT organizing center of the cell, in the neuronal progenitors (Hong et al., 2010; Supp. Fig. S2, panels A–H). The SAGFF73A;UAS:Gtuba2 embryos and larvae developed without any gross morphological defects, survived to

adulthood and were fertile (data not shown). Thus far, the live stock of SAGFF73A;UAS:Gtuba2 double transgenic fish has been stably maintained to the F5 generation. These observations indicate that Gtuba2 is functional as an a-tubulin and zebrafish can tolerate the whole body induction of the Gtuba2 protein. Next, we observed MTs in various cell types in SAGFF73A;UAS:Gtuba2 embryos or larvae. In the spinal cord, distinctive MTs bundles in the motor axons exiting the spinal cord and innervating the muscle were clearly observable (Fig. 1E, arrows). Furthermore, elaborate MT structures in the spinal cord extending along the dorsoventral axis were also observable (Fig. 1E). In the neuromasts, MT filaments were dense on the apical side of the hair cells and extended toward the basal side. (Fig. 1F,G). We also observed apical stereocilia of the hair cells (Fig. 1F,G, arrowheads). We found that SAGFF73A expressed Gal4FF maternally (Fig. 1H, inset). When the

Fig. 1. Labeling of microtubules (MTs) with green fluorescent protein (GFP) -tagged tubulin in various cell types. A: Structure of the T2ZUASGtuba2 construct. The Tol2 sequence and UAS are shown in gray and purple, respectively. The tuba2 tagged with the EGFP gene is shown with a green arrow. B: The SAGFF73A; UAS:Gtuba2 embryo at 24 hours postfertilization (hpf). C: The epithelial cells of the head region of the SAGFF73A; UAS:Gtuba2 larva at 5 days postfertilization (dpf). D: The cells of the external enveloping layer cells of SAGFF73A; UAS:Gtuba2 embryo at 12 hpf. The embryo had been injected with pT2ZUASH2AR at the one cell stage. Gtuba2 and H2AR signals are shown in green and magenta, respectively. E: The lateral view of the trunk of the SAGFF73A; UAS:Gtuba2 larva including the spinal cord and the skeletal muscle at 5 dpf. Arrows indicate motor axons innervating the skeletal muscles. F,G: The neuromast of the SAGFF73A; UAS:Gtuba2 larva at 5 dpf. H: The embryo laid by the SAGFF73A;UAS:Gtuba2 female at the 32-cell stage; Inset: an egg laid by the SAGFF73A;UAS:GFP female. UAS, upstream activating sequence. Scale bars ¼ 500 bp in A; 250 mm in B,G; 10 mm in C,D,F; 20 mm in E; 50 mm in H. Fig. 2. Time lapse imaging of single microtubules (MTs). Serial images of a skin cell in the SAGFF73A;UAS:Gtuba2 larvae at 108 hours postfertilization (hpf). The green fluorescent protein (GFP) signal is shown in black. The MT that was present just before shrinking is indicated by the arrowhead. The shrinking end of the MT is indicated by the white arrow. The shrinking end is not evident in the next frame. The growing end of the other MT is indicated by the black arrow. The growth rate of this MT is in a range of 13.0–25.9 mm/min from t ¼ 128.2 to 139.8, which is slightly faster than that observed in mammalian LLCPK-1 cells (Rusan et al., 2001). The ovalshaped GFP-negative region (N) is the nucleus. The position of the putative MT-organizing center is indicated by the double arrowhead. The numbers in white indicate the time points in seconds. Scale bar ¼ 2 mm. Fig. 3. Time lapse imaging of mitosis in the developing hindbrain. A: The dorsal view of the developing cerebellar region in the hspGGFF27A:UAS:Gtuba2 embryo at 54 hours postfertilization (hpf). The anterior is to the left. White dotted lines demarcate the developing cerebellum. The yellow dotted line indicates the midline. The division of the progenitor cell in the yellow box is monitored by time lapse imaging. The a and b indicate the apical and basal side of the cerebellum, respectively. Scale bars ¼ 10 mm. B: Time lapse imaging of a progenitor cell in the yellow box in A. White arrowheads indicate the position of the centrosomes before separation. Yellow arrowheads indicate the separated centrosomes. The double arrowhead (t ¼ 46) indicates the position of the cleavage furrow. The positions of the centrosomes were unclear at t ¼ 26. The numbers in white indicate the time points in minutes. Scale bars ¼ 10 mm in A; 5mm in B.

ZEBRAFISH LINE FOR MONITORING MT IN VIVO 2697

female SAGFF73A;UAS:Gtuba2 fish was crossed with the male wild type fish, mitotic spindles were observable

in the resulting embryos from the cleavage stage (Fig. 1H). These observations indicated that the UAS:G-


Fig. 1.

Fig. 2.

Fig. 3.

tuba2 line can be used to visualize MT structures in various cell types. A major advantage of labeling MT in intact animals is to monitor MTs in real time. Therefore, we tested whether UAS:Gtuba2 can be used for time lapse imaging of the MTs. First, we tested whether the dynamic behavior of MT filaments can be visualized by Gtuba2. We chose the skin cells of SAGFF73A;UAS:Gtuba2 larvae to observe MT filaments (Fig. 1C), because the flat shape and relatively large size of these cells makes them suitable for live imaging. We found that most of the MT filaments were stable in these cells (Fig. 2). However, among the GFP-positive MT filaments, some were dynamically growing from and shrinking toward a putative centrosomal region (Fig. 2 and Supp. Movie S1), suggesting that MTs with different dynamics are present in these cells. These observations also suggest that the dynamic instability of a single MT can be rendered observable using Gtuba2. Next, we performed time lapse imaging of the proliferation of neuronal progenitor cells in the brain. By using the enhancer trap line hspGGFF27A (Asakawa et al., 2008), Gtuba2 was expressed in the hindbrain region, including the developing cerebellum (Fig. 3A). In the hspGG FF27A;UAS:Gtuba2 double transgenic larvae at 54 hours postfertilization (hpf), the typical elongated morphology of neuronal progenitor cells was visualized by GFP-positive MT arrays extending apicobasally. We


2698 ASAKAWA AND KAWAKAMI

observed that a progenitor cell with its nucleus on the apical side began to develop a single dot of Gtuba2 (Fig. 3B, t ¼ 0). Twelve minutes later, the Gtuba2 dot was separated into two parts and these were segregated to each side of the nucleus (t ¼ 12–19). Then, GFP-positive filaments emanating from those two dots formed a bipolar spindle (t ¼ 19–30), indicating that the GFP dots were centrosomes. We noted that, until t ¼16, the region between the centrosomes, which corresponds to the nucleus, remained almost devoid of the GFP fluorescence. However, after t ¼ 19, such GFP-negative region disappeared, suggesting that the nuclear envelope break down occurred between t ¼ 16–19. The pole-to-pole distance remained constant for 12 min (t ¼ 30–42) and spindle elongation occurred abruptly (t ¼ 46). The spindle structure disappeared soon after the spindle elongation and the subsequent horizontal cytokinesis occurred (t ¼ 46–52). These observations indicate that UAS:Gtuba2 can be used to perform time-laps imaging of mitotic spindle in the developing brain. In the present study, we developed a method for monitoring MTs in vivo by means of a transgenic reporter line carrying the a-tubulin fused to GFP. While genetically encoded GFPtagged tubulins are extensively used to visualize MTs during animal development in C elegans (Strome et al., 2001) and Drosophila (Grieder et al., 2000), this is the first application in vertebrates. We found that zebrafish can tolerate whole body induction of GFP-tagged a-tubulin not only during the embryonic and larval stages, but also in the adult stage. Therefore, the toxicity due to the overexpression of the GFP-tagged tubulin appears to be minimal, and MT structures under near-physiological conditions can be visualized by this method. We have demonstrated that UAS:Gtuba2 fish can be used to monitor the dynamic instability of a single MT in skin cells. In neuronal progenitor cells, however, we found it was impossible to observe single MTs in real time due to the relatively small size of this cell type. It remains to be determined whether single MTs can be rendered observable with Gtuba2 in other cell types. We also demonstrated that the UAS:Gtuba2 fish is applicable for

time lapse imaging of the mitotic spindle in neuronal progenitor cells in the developing cerebellum. Separation and segregation of the centrosome before metaphase were clearly monitored by GFP fluorescence in the progenitor cells undergoing mitosis (Fig. 3). This indicates that the UAS:Gtuba2 line enables the detailed analyses of cell cycle progression that have been difficult to achieve by method such as visualizing chromosomes with fluorescent-tagged proteins such as histones (Pauls et al., 2001). It would also be advantageous to apply this system to investigate the function of regulators of mitotic spindle in proliferating neuronal progenitor cells, which could be done by inactivating such regulators through an injection of morpholino antisence oligonucleotide or genetic mutations and investigating the effect on the mitotic spindle. Because a large number of tissue-specific Gal4 lines are now available in zebrafish (Davison et al., 2007; Scott et al., 2007; Asakawa et al., 2008) the UAS:Gtuba2 reporter line will allow for detailed analysis of MTs in cell proliferation, migration, and morphogenesis during development of the brain, as well as other organs.

EXPERIMENTAL PROCEDURES Construction of the UAS:Gtuba2 plasmid To create the plasmid containing the T2ZUASGtuba2 construct, the EGFP fragment was introduced into the BglII site in the pT2MUASMCS, which contains a synthetic multicloning site downstream of five tandem repeats of the Gal4 binding sequence (5xUAS) and a TATA sequence in the Tol2 transposable element (Asakawa et al., 2008), generating T2MUA SMCSGN. The cDNA of tuba2 (NM_194388) was amplified by PCR using the primers tuba2-f (50 -GAA GAT CTA TGC GTG AGT GTA TCT CCA T-30 ) and tuba2-r (50 -CCG CTC GAG CTA ATA CTC CTC ACC TTC CT-30 ) cloned into pCRII-TOPO using a TA cloning kit (Invitrogen). The resulting plasmid containing tuba2 was digested with BglII and XhoI, and the BglII-XhoI fragment contain-

ing tuba2 was introduced into the T2MUASMCSGN digested with BglII and XhoI, generating pT2ZUASGtuba2. In T2ZUASGtuba2, EGFP and tuba2 (BC060904) were fused with a linker sequence (AGA TCT).

Southern Blot Hybridization, Inverse PCR, LinkerMediated PCR Southern blot hybridization, inverse PCR, and linker-mediated PCR were carried out as described previously (Asakawa and Kawakami, 2009). The 32 P-labeled EGFP probe was used to detect T2ZUASGtuba2 insertion.

Microscopic Analysis A fluorescence stereomicroscope (MZ16FA, Leica) equipped with a CCD camera (DFC300FX, Leica) was used to observe and take images of the GFP-expressing embryos. For confocal microscopy, embryos and larvae were embedded in 1% low melting agarose and analyzed with a Zeiss LSM510META laser confocal microscope. Confocal images were acquired as serial sections along the z-axis at 1.0 mm-intervals. For time lapse imaging of single MTs (Fig. 2), a single section was acquired at single time points approximately every 3.9 sec. For live imaging of embryos during the cleavage period, embryos in the chorions were placed at the bottom of Glass Base Dish (IWAKI, 3910035) in the embryonic buffer and embryos in an appropriate orientation for imaging were analyzed. Confocal images were acquired as serial sections along the z-axis at 3.0-mm intervals. The acquired images were processed with Zeiss LSM Image Browser and Adobe Photoshop CS5.

Expression of mRFP1-histone H2a Fusion Protein To create the pT2ZUASH2AR plasmid, The cDNA of h2afz (NP_705930.1) was fused to mRFP1 (Campbell et al., 2002) at the carboxyterminus and the h2afz-mRFP1 fragment was introduced into pT2MUA SMCS. In T2ZUASH2AR, h2afz and mRFP1 were fused with a linker sequence (GGAGGCTCGAATCTCG AG). To visualize the chromosomes,

ZEBRAFISH LINE FOR MONITORING MT IN VIVO 2699

approximately 1 nl of the circular pT2ZUASH2AR (25 ng/ml) was injected into fertilized eggs laid by SAGFF73A;UAS:Gtuba2 female fish. We found that larvae carrying both the SAGFF73A and UAS:H2AR insertions and exhibiting red fluorescence in the whole body were lethal, indicating that a high level of h2afz-mRFP1 expression is toxic.

ACKNOWLEDGMENTS


We thank Drs. M. Hibi, H. Wada, and T. Kotani for technical advices and A. Ito, M. Suzuki, M. Mizushina, N. Mouri, Y. Kanebako, and T. Asada for fish maintenance. This work was supported by the National BioResource Project from the Ministry of education, Culture, Sports, Science and Technology of Japan.

REFERENCES Angers S, Thorpe CJ, Biechele TL, Goldenberg SJ, Zheng N, MacCoss MJ, Moon RT. 2006. The KLHL12-Cullin-3 ubiquitin ligase negatively regulates the Wnt-beta-catenin pathway by targeting Dishevelled for degradation. Nat Cell Biol 8:348–357. Asakawa K, Kawakami K. 2008. Targeted gene expression by the Gal4-UAS system in zebrafish. Dev Growth Differ 50: 391–399. Asakawa K, Kawakami K. 2009. The Tol2-mediated Gal4-UAS method for gene and enhancer trapping in zebrafish. Methods 49:275–281. Asakawa K, Suster ML, Mizusawa K, Nagayoshi S, Kotani T, Urasaki A, Kishimoto Y, Hibi M, Kawakami K. 2008. Genetic dissection of neural circuits by Tol2 transposon-mediated Gal4 gene and enhancer trapping in zebrafish. Proc Natl Acad Sci U S A 105:1255–1260. Brand AH, Perrimon N. 1993. Targeted gene expression as a means of altering

cell fates and generating dominant phenotypes. Development 118:401–415. Campbell RE, Tour O, Palmer AE, Steinbach PA, Baird GS, Zacharias DA, Tsien RY. 2002. A monomeric red fluorescent protein. Proc Natl Acad Sci U S A 99:7877–7882. Carminati JL, Stearns T. 1997. Microtubules orient the mitotic spindle in yeast through dynein-dependent interactions with the cell cortex. J Cell Biol 138: 629–641. Davison JM, Akitake CM, Goll MG, Rhee JM, Gosse N, Baier H, Halpern ME, Leach SD, Parsons MJ. 2007. Transactivation from Gal4-VP16 transgenic insertions for tissue-specific cell labeling and ablation in zebrafish. Dev Biol 304:811–824. Grieder NC, de Cuevas M, Spradling AC. 2000. The fusome organizes the microtubule network during oocyte differentiation in Drosophila. Development 127: 4253–4264. Hong E, Jayachandran P, Brewster R. 2010. The polarity protein Pard3 is required for centrosome positioning during neurulation. Dev Biol 341: 335–345. Jaglin XH, Poirier K, Saillour Y, Buhler E, Tian G, Bahi-Buisson N, FalletBianco C, Phan-Dinh-Tuy F, Kong XP, Bomont P, Castelnau-Ptakhine L, Odent S, Loget P, Kossorotoff M, Snoeck I, Plessis G, Parent P, Beldjord C, Cardoso C, Represa A, Flint J, Keays DA, Cowan NJ, Chelly J. 2009. Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria. Nat Genet 41:746–752. Keays DA, Tian G, Poirier K, Huang GJ, Siebold C, Cleak J, Oliver PL, Fray M, Harvey RJ, Molnar Z, Pinon MC, Dear N, Valdar W, Brown SD, Davies KE, Rawlins JN, Cowan NJ, Nolan P, Chelly J, Flint J. 2007. Mutations in alphatubulin cause abnormal neuronal migration in mice and lissencephaly in humans. Cell 128:45–57. Pauls S, Geldmacher-Voss B, CamposOrtega JA. 2001. A zebrafish histone variant H2A.F/Z and a transgenic H2A.F/Z:GFP fusion protein for in vivo

studies of embryonic development. Dev Genes Evol 211:603–610. Poirier K, Keays DA, Francis F, Saillour Y, Bahi N, Manouvrier S, Fallet-Bianco C, Pasquier L, Toutain A, Tuy FP, Bienvenu T, Joriot S, Odent S, Ville D, Desguerre I, Goldenberg A, Moutard ML, Fryns JP, van Esch H, Harvey RJ, Siebold C, Flint J, Beldjord C, Chelly J. 2007. Large spectrum of lissencephaly and pachygyria phenotypes resulting from de novo missense mutations in tubulin alpha 1A (TUBA1A). Hum Mutat 28:1055–1064. Rusan NM, Fagerstrom CJ, Yvon AM, Wadsworth P. 2001. Cell cycle-dependent changes in microtubule dynamics in living cells expressing green fluorescent proteinalpha tubulin. Mol Biol Cell 12:971–980. Scott EK, Mason L, Arrenberg AB, Ziv L, Gosse NJ, Xiao T, Chi NC, Asakawa K, Kawakami K, Baier H. 2007. Targeting neural circuitry in zebrafish using GAL4 enhancer trapping. Nat Methods 4:323–326. Strome S, Powers J, Dunn M, Reese K, Malone CJ, White J, Seydoux G, Saxton W. 2001. Spindle dynamics and the role of gamma-tubulin in early Caenorhabditis elegans embryos. Mol Biol Cell 12: 1751–1764. Takeda S, Funakoshi T, Hirokawa N. 1995. Tubulin dynamics in neuronal axons of living zebrafish embryos. Neuron 14:1257–1264. Tischfield MA, Baris HN, Wu C, Rudolph G, Van Maldergem L, He W, Chan WM, Andrews C, Demer JL, Robertson RL, Mackey DA, Ruddle JB, Bird TD, Gottlob I, Pieh C, Traboulsi EI, Pomeroy SL, Hunter DG, Soul JS, Newlin A, Sabol LJ, Doherty EJ, de Uzcategui CE, de Uzcategui N, Collins ML, Sener EC, Wabbels B, Hellebrand H, Meitinger T, de Berardinis T, Magli A, Schiavi C, Pastore-Trossello M, Koc F, Wong AM, Levin AV, Geraghty MT, Descartes M, Flaherty M, Jamieson RV, Moller HU, Meuthen I, Callen DF, Kerwin J, Lindsay S, Meindl A, Gupta ML Jr, Pellman D, Engle EC. 2010. Human TUBB3 mutations perturb microtubule dynamics, kinesin interactions, and axon guidance. Cell 140:74–87.