Mutation of Kinesin-6 Kif20b causes defects in cortical neuron polarization and morphogenesis
- Katrina C. McNeely†1,
- Timothy D. Cupp†1,
- Jessica Neville Little1,
- Kerstin M. Janisch1,
- Ayushma Shrestha1 and
- Noelle D. Dwyer1Email author
© The Author(s). 2017
Received: 29 January 2017
Accepted: 22 March 2017
Published: 31 March 2017
How neurons change their cytoskeleton to adopt their complex polarized morphology is still not understood. Growing evidence suggests that proteins that help build microtubule structures during cell division are also involved in building and remodeling the complex cytoskeletons of neurons. Kif20b (previously called MPP1 or Mphosph1) is the most divergent member of the Kinesin-6 family of “mitotic” kinesins that also includes Kif23/MKLP1 and Kif20a/MKLP2. We previously isolated a loss-of-function mouse mutant of Kif20b and showed that it had a thalamocortical axon guidance defect and microcephaly.
We demonstrate here, using the mouse mutant, that Kif20b is required for neuron morphogenesis in the embryonic neocortex. In vivo and in vitro cortical neurons were labeled and imaged to analyze various aspects of morphogenesis.
Loss of Kif20b disrupts polarization as well as neurite outgrowth, branching and caliber. In vivo, mutant cortical neurons show defects in orientation, and have shorter thinner apical dendrites that branch closer to the cell body. In vitro, without external polarity cues, Kif20b mutant neurons show a strong polarization defect. This may be due in part to loss of the polarity protein Shootin1 from the axonal growth cone. Those mutant neurons that do succeed in polarizing have shorter axons with more branches, and longer minor neurites. These changes in shape are not due to alterations in cell fate or neuron layer type. Surprisingly, both axons and minor neurites of mutant neurons have increased widths and longer growth cone filopodia, which correlate with abnormal microtubule organization. Live analysis of axon extension shows that Kif20b mutant axons display more variable growth with increased retraction.
These results demonstrate that Kif20b is required cell-autonomously for proper morphogenesis of cortical pyramidal neurons. Kif20b regulates neuron polarization, and axon and dendrite branching, outgrowth, and caliber. Kif20b protein may act by bundling microtubules into tight arrays and by localizing effectors such as Shootin1. Thus it may help shape neurites, sustain consistent axon growth, and inhibit branching. This work advances our understanding of how neurons regulate their cytoskeleton to build their elaborate shapes. Finally, it suggests that neuronal connectivity defects may be present in some types of microcephaly.
KeywordsKinesin Kif20b Axon outgrowth Microtubule Mouse Apical dendrite Axon branching Cytoskeleton Filopodia Microcephaly
The function of the nervous system depends on neurons having a receiving end (dendrites) and a transmitting end (a single axon). A cortical pyramidal neuron grows a large apical dendrite at the apex of the cell body, several shorter basal dendrites, and a single axon from the base to connect to distant targets. Early polarization occurs soon after the neuron’s birth, during the migratory phase through the intermediate zone. The newborn neuron first becomes multipolar with several short neurites, and then becomes bipolar with a leading process oriented outward in the direction of migration away from the ventricle. As it migrates, it extends an axon from the trailing end, but does not initiate dendrite growth until it finishes migrating.
Early polarization events can be modeled in dissociated cultures of embryonic hippocampal or cortical neurons . During repolarization in vitro, the neurons first appear round and flat like a fibroblast (Stage 1, unpolarized). Within a day they begin to extend several undifferentiated neurites of about equal length (Stage 2, Multipolar). The growing neurites are full of microtubules, and they are tipped by actin-rich growth cones with lamellae and filopodia. In a stochastic manner, one of the neurites elongates rapidly to become the axon (Stage 3, Polarized). The remaining neurites become dendrites and remain shorter and thicker than the axon. This invaluable system has revealed many factors important for polarization , but the cytoskeletal mechanisms underlying the development of complex neuronal morphology are still not well understood [3, 4].
Various kinesin motor proteins are crucial for establishing or maintaining neuronal polarity and structure through their interactions with microtubules and cargoes . The Kinesin-6 subfamily members are plus-end- directed microtubule motors known as “mitotic” kinesins for their roles in cytokinesis but some may have additional functions . Mammalian cells express three Kinesin-6 family members: Kif23/MKLP1, Kif20a/MKLP2, and Kif20b (formerly called mitotic phospho-protein 1, MPP1 or Mphosph1). RNAi of Kif23 in cultured rat sympathetic neurons disrupted microtubule polarity of dendrites and resulted in longer axons and dendrites [7, 8]. Mutation of the Drosophila Kif23 ortholog, pavarotti (pav), caused excessive microtubule sliding, axon growth, and axon branching [9, 10].
The third member of the mammalian Kinesin-6 subfamily, Kif20b, is less understood and has higher molecular weight and lower abundance than Kif23 or Kif20a. Interestingly, it has a two-fold longer stalk comprised of four coiled-coil domains linked by three hinges [11, 12]. In in vitro assays, KIF20B was sufficient to both slide and bundle microtubules in an ATP-dependent manner . In a previous mouse ENU screen, we found a recessive, neonatal lethal mutant that displayed microcephaly and axon guidance defects in a subset of thalamocortical axons at embryonic day (E) 18.5 . Genetic mapping and a complementation test identified Kif20b as the mutant gene . The point mutation in Kif20b causes an mRNA splicing error, a consequent frameshift and premature termination codons, and reduces Kif20b protein to undetectable levels on immunoblots or cell staining . Mutants show cytokinesis defects in neural stem cells of embryonic cortex. Cortical size and thickness is reduced in the Kif20b mutants due to decreased numbers of neurons and intermediate progenitors. Despite this, laminar organization and most axon tracts appear grossly normal at birth [13, 14].
Here, to explore Kif20b roles in neuronal development during corticogenesis, we took advantage of this genetic loss-of-function mutant. In Kif20b mutant brains, cortical pyramidal neurons have shorter, thinner, apical dendrites which branch closer to the cell body, extra axon branches, and are sometimes misoriented. To separate non- cell-autonomous effects, we pursued further analyses in dissociated neuron cultures. Surprisingly, when isolated from their normal brain environment, Kif20b mutant neurons show a strong polarization defect. This defect may be at least partly explained by a role for Kif20b in localizing the polarity protein Shootin1. The polarization defect is not due to cell fate changes and affected both deep and superficial layer types. Furthermore, the Kif20b mutant neurons that do successfully polarize have a variety of morphological changes including shorter, more branched axons and longer minor neurites. Mutant neurites are wider, and growth cone filopodia are longer with increased microtubule penetration. In live imaging, axons of Kif20b mutant neurons appear to pause less and retract more. These data indicate that Kif20b is important for polarization and maintaining axon growth and preventing branching, and suggest that it acts both by localizing cargo and organizing microtubule bundles.
To prepare for neuron culture and plating, 18 mm round coverslips were washed twice every 10 min with double-distilled, UV-irradiated water and treated in nitric acid overnight. Following three subsequent washes with double-distilled water, coverslips were placed in an oven at 160 °C overnight to dry and sterilize. After allowing to cool the next day, each coverslip was treated with 200 μL poly-L-lysine (PLL) solution (1 μg/mL in borate buffer) and incubated overnight at 37 °C. This was followed by double-distilled water washes (2 h washes were done twice following three quick rinses) and application of neuron plating media (.5 mL-1 mL). Neuron plating media is filter-sterilized and consists of 500 mL Minimum Essential Medium (MEM) with glutamine, 5 mL Penicillin/Streptomycin, 15 mL 20% glucose, 5 mL Sodium Pyruvate, and 10% Fetal Bovine Serum. At E14.5, pregnant females were sacrificed and the embryos placed into cold HBSS mix (500 mL Hank’s Balanced Salt Solution (HBSS) with 5 mL HEPES and 5 mL Penicillin/Streptomycin. Using fine forceps, pieces of cortex were collected and placed into a tube containing a 0.05% trypsin solution for 15 min in a 37 °C water bath. Following trypsin digestion, the resulting neuron pellets were rinsed 3 times every 5 min with HBSS mix and then treated with neuron plating medium during trituration. Appropriate volumes of the resulting solution of neuron plating medium and dissociated neurons were pipetted onto the PLL-coated coverslips to achieve a density of 50,000 cells/mL. Medium was switched from Neuron Plating Medium to Neurobasal and B27 (NB27) after three h. 48 h after being plated initially, the neurons were fixed in 2% Paraformaldehyde (PFA) for ten minutes and then in 2% PFA with 30% sucrose for 10 min. Finally, coverslips were rinsed in one time in Phosphate Buffer Saline (PBS) for 10 min three times and kept at 4 °C until ready for immunofluorescent staining.
Following dissociation, neuron plating, and fixation, coverslips were incubated for an hour at room temperature in blocking buffer (0.1% Triton-X, 2% Normal Goat Serum in PBS) and then overnight at 4 °C or for 3 h at room temperature in appropriate primary antibody solution (antibody diluted in blocking solution). After primary incubation, coverslips were rinsed in PBS (3 times every 10 min) and then incubated at room temperature with appropriate secondary antibody solution (1:200 dilution) for 30 min in the dark. Following final washes in PBS, coverslips were mounted onto glass slides with Fluoromount.
Primary antibodies used were a rabbit or mouse monoclonal antibody against neuron-specific beta-III tubulin (Tuj1) at a dilution of 1:500 in blocking buffer (Covance clone 1-15-79 D71G9 and MMS-435P, Abcam ab52623), a rat polyclonal antibody against Ctip2 at a dilution of 1:500 (Abcam ab18465), mouse tubulin-alpha clone (DM1alpha) at a dilution of 1:500 (Thermo Scientific MS-581-PO), a mouse monoclonal antibody against Tau at a dilution of 1:200 (Millipore mab2370), a rabbit polyclonal antibody against DCX at a dilution of 1:1200 (Abcam, ab18723) and (green)-Phalloidin (Molecular Probes O7466) at a dilution of 1:50. Species-specific secondary antibodies were conjugated to Alexa fluorophores (Invitrogen) at a dilution of 1:200 in blocking buffer. Shootin1 antibody (1: 100, B627, raised against the first 456aa) was a gift from Tamar Sapir and Orly Reiner, Weizmann Institute .
Multiple Kif20b antibodies have been verified for specificity by detecting signal in the midbodies of control dividing cells (HeLa and MEFs) and not in depleted or mutant cells. All of these antibodies were subsequently tried on mouse neurons. Published antibodies were kind gifts of Fabienne Pirollet (made to full length human KIF20B, ) and Orly Reiner (made to 1002 to 1442 of Kif20b, ). A commercially available monoclonal from Santa Cruz (raised against amino acids 1557–1675 of human KIF20B, catalog sc-515194) was tried. Both peptide anti-sera and purified antibodies to the N-terminal domain and C-terminal were also tried . Multiple fixation conditions including TCA, PFA/MeOH, PFA only, permeabilized before fixing, and BRB80 then PFA were tried in combination with the previous mentioned antibodies. Due to large background to signal ratio no noticeable difference between control and Kif20b mutant neurons was detected with any of the antibodies or different fixation conditions.
Image acquisition and analysis
Fluorescent images were obtained on a Carl Zeiss widefield epi-fluorescence microscope via AxioVision camera and software. Low-magnification images for stage analyses were taken at 20× while high-magnification images for neuron measurements were taken at 40× or 100×. Image analysis was completed through ImageJ software. We used the NeuronJ software plugin to trace and measure axons and minor neurites. Axon branches were only counted if they were a minimum of 5 μm. Neurite lengths were measured from the base of the process at the soma to the tip of the Tuj1 stain. Stage 3 neurons’ minor neurites were measured in thickness at 0 μm, 5 μm, and 10 μm from the edge of the soma. Stage 3 neurons’ axons were measured in thickness at 0 μm, 5 μm, 10 μm, and 25 μm from the edge of the soma. For Fig. 7D, images of tubulin in axons were captured using a Deltavision widefield microscope and then deconvolved. The intensity of tubulin staining was measured across the width of the axon at 5 μm, 10 μm, and 25 μm using Zeiss Zen 2 lite imaging software profile tool.
Polarity stage analyses
Neurons were considered to be in Stage 1 if they extended broad lamellipodia with no clear, coalesced neurites. The neurites of Stage 2 neurons are all of similar length. A neuron was placed into the Stage 3 category when one neurite was at least twice as long as the next longest neurite. Neuronal protrusions were considered to be neurites or neurite branches if they had significant microtubule invasion.
Measuring growth cones and filopodia
The growth cone area was measured from the base of intense phalloidin-stained actin at the axon neck, around the growth cone tracing lamellipodial edges. The lamellipodial edge was considered to be where the filopodium becomes uniform in width (the filopodial base). We considered filopodia to be actin protrusions extending out of lamellipodia at growth cones. Filopodial extensions generally have brighter phalloidin staining than neighboring lamellipodial edges. Measurements were taken from the filopodia tip to its base, where the edges begin to splay apart and become lamellae.
Shootin1, DCX, and Tau localization measurements
Images of Shootin1 immunostaining were captured using the Zeiss AxioImagerZ1 Microscope with a 100× oil objective, and a constant 600 milliseconds exposure time. Tau and DCX images were captured using the 40× objective on the Zeiss Observer Z1 microscope with Axiocam 506 mono camera. Exposure time was kept constant (DCX 300 ms and Tau 200 ms). Zeiss Zen 2 lite imaging software profile tool was used to measure intensity. Excel was used to compare individual cells line scans as well as create averages across genotypes for comparison.
Mouse embryonic fibroblasts (from E14 mice) and mouse cortical neurons (4 DIV) were grown at a density of ~750000 cells/dish. RNA extraction of the harvested cells was done using the PureLinkTM RNA Mini Kit from life technologies according to the manufacturer’s handbook. RNA content in the samples was determined with a nano-drop spectrometer at 260 nm. 100 ng RNA were used in the reactions with the Invitrogen superscript III one-step RT-PCR kit. The PCR products were run on 3% ultrapure agarose with ethidium bromide at 95 V for 1 h. Gels were visualized on a UV light box.
Kif20b exon 3–7: FW:TGCTGAAAGACCCTCAAAGCATCCT, RV:ACTGGACTGGTCACAACTGTTCACG
Kif20b exon 17–19: FW: GGTTCAGGCACTCAAGACATCAAGT, RV: CGATACTTCTTGCAGCAGTCTCCAT
Beta-actin (Actb gene): FW: GATGACCCAGATCATGTTTGAGACC,
The GFP-KIF20B construct was electroporated into mouse cortical neurons. Shortly, neurons were dissociated from E14 mice cortices freed from meninges. After a cell count, the appropriate volume for one million neurons density was pipetted in eppendorf tubes and the neurons were spun down at 1.8 rpm for 10 min. The supernatant was carefully removed and the cells were re-suspended in 100 μl electroporation solution and 5 μg total plasmids were added (4 μg of GFP-KIF20B plasmid, 1 μg mCherry plasmid) and the suspension was mixed gently. The neurons were transferred carefully into an electroporation cuvette and electroporated with an Amaxa Electroporator following the instructions of the manufacturer for mouse cortical neurons (setting: O-005). Immediately after electroporation, 500 μl medium with serum was added to the cells and the suspension is carefully transferred into an eppendorf tube. 200 μl of cell suspension was then plated on either glass bottom dishes or coverslips and plating medium is added to make up 1 ml. Medium was replaced after 3 h with neuron growth medium with B27. Cells were grown at 37 °C, with 5% CO2 until usage.
Lipophilic DiI (1,1’-Dioctadecyl-3,3,3’,3’-Tetramethylindocarbocyanine Perchlorate) was used to stain neurons in E15.5 control and Kif20b mutant brains. Embryos were collected from pregnant heterozygote mothers at E15.5 and brains were dissected and fixed in 4% PFA for 2 days. A single small DiI crystal (Invitrogen, D-282) was then placed in the mid-lateral cortex of each hemisphere of each fixed brain using a needle pin. Brains were incubated at 37 °C in the dark for 3 days. Next, brains were cut in coronal sections at 100 μm on a Leica VT1000S vibratome and sections stained with DAPI (Invitrogen, D1306). Sections were mounted on slides with VectaShield (Vector Laboratories, H-1000) mounting medium. Z-stack images were taken of individual pyramidal neurons in the cortex at 40× on a Zeiss AxioImager.Z1 fluorescent microscope. Analysis of individual neurons was completed with ImageJ and NeuronJ by creating a maximum intensity projection of stacked images.
The Golgi method was used to stain neurons in E18.5 control and Kif20b mutant brains. Embryos were collected from pregnant heterozygote mothers at E18.5 and brains were dissected and rinsed with distilled water. The FD Rapid GolgiStain™ Kit (FD NeuroTechnologies, Inc., PK401) was used to stain whole brains, with the modification of placing brains in Solution A/B for 3 weeks instead of 2 at room temperature. Brains were next placed in Solution C for 1 week at 4 °C. Afterwards, brains were flash frozen on dry ice and stored at -80 °C until they were mounted in distilled water and cut on a Leica CM 3050S cryostat. Coronal sections were cut at 100 μm and mounted on slides with Solution C. The sections were allowed to dry for 3 days before completing the staining procedure with Solutions D and E as per instructions in the FD Rapid GolgiStain™ Kit. Slides were coverslipped with CytoSeal 60 (Thermo Scientific, 8310-4). Z-stack images were taken of individual pyramidal neurons in the cortex at 40× on a Zeiss AxioImager.Z1 microscope with Brightfield illumination. Neurons were also traced with Neurolucida Neuron Tracing Software (MBF Bioscience) for analysis. Analysis was completed by combining results from Neurolucida with results obtained from minimum intensity projection stacked images in ImageJ and NeuronJ.
Neuron live imaging
The neurons were allowed to grow at 5% CO2 and 37 °C for 45 to 54 h in the incubator before imaging, and kept at 37 °C and 5% CO2 in the microscope controlled environmental chamber between time 0 and 6 h. The microscope was a Zeiss AxioObserver with an inverted 20× objective, a motorized stage, and Definite Focus that allowed for multiple scenes within each chamber to be imaged. The camera was AxioCam Mrm and image analysis was done with Zeiss Zen software.
Loss of Kif20b disrupts morphogenesis of pyramidal neurons in embryonic cortex
To test whether the neurite and orientation defects were detectable from early growth stages, we examined cortical neurons of E15.5 brains. Golgi-Cox staining does not label cortical neurons at this age, so retrograde dye-tracing was employed. In both control and mutant cortices, lateral diI crystal placements could retrogradely label pyramidal neurons in the cortical plate, confirming that at least some mutant neurons had extended long axons (Fig. 1 F, a-d). However, the mutant cortical plate appeared disorganized, with fewer neurons labeled than in controls, and less than half of those oriented properly with apical dendrites perpendicular to the pial surface (12/31 mutant cells vs. 20/31 control cells), (Fig. 1Fa,c, 1G). Mutant apical dendrites were only two thirds the normal height (Fig. 1H). Strikingly, the mutant apical dendrites branched much closer to the cell body, at about one third the normal distance (Fig. 1Fb and d, arrowheads, and 1H). Axons are not expected to branch much at this early age, but branching on proximal axons appeared increased among mutant neurons, with 5/15 mutant neurons having at least one detectable branch, versus only 3/22 control neurons detected with one branch each. Again, the average soma size was indistinguishable in mutants and controls (Fig. 1I). Together, these analyses demonstrate that Kif20b is required for multiple aspects of normal neuron morphogenesis in the developing cortex, including polarity, branching, and dendrite width. The requirements for Kif20b could be cell-autonomous or non-autonomous. For example, the shortening of apical dendrites was proportional to the decrease in cortical thickness at both E15.5 and E18.5. By contrast, the shortened distance to the first branch point of the apical dendrites remained significant even after normalization to cortical thickness. Therefore, to sort out cell autonomous requirements for Kif20b in neuron growth from non-autonomous effects due to abnormal brain size and shape, we pursued further analyses on dissociated cortical neurons in vitro. This also enabled us to analyze more features of neuron shape.
Kif20b is required for normal polarization in dissociated cortical neurons
Shootin1 is less enriched in growth cones of Kif20b mutant axons
Polarization and structure differences are not due to a shift in cell or layer fates
Kif20b mutant neurons that do polarize have structural changes in both axons and minor neurites
To confirm that the changes in branching and neurite width were not due to a change in neuron types in the Kif20b mutant cultures, we again controlled for this. Indeed, the foregoing findings held true when we controlled for neuron layer type: Kif20b mutant Ctip2+ neurons had increased axon branching, minor neurite length, and neurite width, compared to Ctip2+ control neurons (Additional file 1: Figure S1). Worth noting, Ctip2+ neurons did not differ significantly from Ctip2- neurons by our measurements, and were affected similarly by Kif20b loss. This suggests that at least at early days in vitro, deep and upper layer pyramidal neurons develop similarly and both require Kif20b.
Kif20b mutant neurons have longer filopodia with increased microtubule invasion
All filopodia contain bundled actin, but when microtubules penetrate filopodia this may lead to stabilization of the filopodium, inducing a branch or growth cone extension [25, 26]. We hypothesized that longer filopodia could be due to increased microtubule invasion. We tested this by comparing the percentage of filopodia with detectable tubulin in control and mutant polarized neurons. Consistent with our hypothesis, the filopodia on Kif20b mutant growth cones were more likely to contain tubulin than control filopodia, suggesting increased microtubule invasion of the growth cone periphery (Fig. 8E, F). Together with the above analyses, these data suggest that loss of Kif20b only partly blocks polarization, and the cells that do polarize have structural changes in both axons and minor neurites - length, branching, width, and filopodia length – suggestive of dysregulated microtubule packing or microtubule-actin interactions.
Kif20b mutant axons retract more and pause less than control axons
The dramatic reorganization of the neuronal cytoskeleton during polarization and axon growth is an area of intense research. Recent work points to microtubule associated proteins being important downstream effectors of signaling cues that mediate neuronal polarization, axon guidance and branching . While many kinesins have been shown to be important for intracellular transport in mature neurons, less is known about which kinesins are important for these early processes. Recent work in Drosophila or by mammalian knockdown approaches has suggested that mitotic kinesins are re-used in postmitotic neurons to organize microtubule arrays [7, 9].
Our analyses highlight the importance of studying neuronal morphogenesis both in vivo and in vitro, because phenotypes vary and provide different clues to gene function. While the Kif20b mutant neurons had a robust polarization defect in vitro, clearly most neurons in vivo polarized and made an axon. A neuron polarizing in vitro has perhaps a more difficult task than one polarizing in vivo because there are no directional cues, and the substrate is stiff and two-dimensional. However, the orientation defect in the Kif20b mutant brains may reflect polarization defects in a subset of cells. Axon branching defects were seen both in vivo and in vitro, as well as effects on neurite width. In vitro we were able to analyze microstructures such as filopodia number and length. Previously, we reported that in the Kif20b mutant mouse brains, a subset of thalamocortical axons showed an axon guidance defect . Though the exact cause of the misrouting is still unclear, our new data analyzing individual neurons suggest that axon guidance defects could be secondary to axon outgrowth timing or abnormal branching.
The Kinesin-6 gene family appears to have expanded by duplication during evolution [6, 27]. C. elegans has one Kinesin-6 gene, zen-4, orthologous to Kif23/MKLP1. Drosophila has two Kinesin-6 genes, pavarotti and subito. It may be that three members of this family are needed to build a more complex nervous system or larger neurons. While gene redundancy is often observed in neural development to ensure robustness , this appears not to be the case for the mammalian Kinesin-6 family, since loss of Kif20b causes deleterious phenotypes in both cytokinesis and neuron development that are not compensated for by Kif23 or Kif20a (; this work). Kif20b must have some molecular function that the other two cannot substitute for. The family members are highly homologous in the motor domain, but the stalks and tails are completely divergent, enabling interactions with distinct binding partners and cargoes. The significance of the extra- long stalk of Kif20b is not known, but it may increase flexibility, or allow the motor to bind microtubules or other proteins at a greater distance apart.
This work substantially deepens the evidence that Kinesin-6 family members function in neuronal development. Previous RNAi data supported roles in mammalian neurons for Kif23/MKLP1 in microtubule organization and neurite growth . Knockdown or mutation of the Drosophila Kif23 ortholog Pav caused increased microtubule sliding and excess axon branching . Kif20b knockdown was shown to decrease axon length and Shootin1 distribution . Here we provide genetic evidence for a requirement for Kif20b in multiple aspects of mammalian neuron morphogenesis. Kif20b loss has morphological consequences throughout the cell, consistent with the appearance of exogenous GFP-KIF20B throughout the neurites. As the cell matures, Kif20b may accumulate at the axon tip due to axonal preference of the motor head as shown in . The appearance of looser microtubules in neurites and more microtubules in growth cone filopodia suggests Kif20b functions in microtubule crosslinking or packing organization in neurites. Indeed, the wider neurites are reminiscent of the wider midbodies we previously observed in dividing neural progenitors . Both these phenotypes could be explained by loss of microtubule bundling activity of Kif20b that was demonstrated in biochemical assays . A key process driving neuron polarization is likely to be local stabilization and bundling of parallel microtubules [4, 28]. As the growth cone advances, microtubule bundling at the neck may be a crucial requirement for converting the growth cone into axonal structure . Thus, defective microtubule bundling could explain the mutant cells’ difficulty forming the axon, regulating neurite width, and preventing axon retraction.
Interestingly, filopodia length and axon branching are linked mechanistically by microtubule invasion. This work provides one of the few known manipulations to increase the length of growth cone filopodia or increase microtubule penetration into them [25, 26]. Axonal branches are initiated as filopodia  and are dependent on filopodia formation . Formation of axon branches correlates with localized splaying and breaking of microtubules [2, 26]. Thus, the increased axon branching and longer filopodia in Kif20b mutants could both be related to a loss of microtubule crosslinking or tight bundling.
Kif20b is likely to have functions beyond microtubule organization. For example, as a translocating motor protein, it can transport proteins or link them to microtubules. We have provided evidence that at least one endogenous polarity protein, Shootin1, requires Kif20b for enriched localization at the axon growth cone. The role of Shootin1 is proposed to reinforce the identity of the axon and keep it growing [18, 19]. Kif20b may transport Shootin1 from the cell body to the growth cone, or it may help retain it at the growth cone after transport by other mechanisms . The high abundance of Shootin1 relative to Kif20b, as well as the remaining weak Shootin1 signal in Kif20b mutant axons support the notion of another transport mechanism. Shootin1 has also been shown to bind the actin cytoskeleton and mediate traction of the growth cone . Thus through Shootin1, Kif20b might link microtubules and actin. Importantly, there may not be one Kif20b function that explains all the phenotypes. There may well be as yet unidentified binding partners that Kif20b transports or tethers at different locations or times in development.
These analyses considerably advance our knowledge of post-mitotic roles of Kif20b in neuron polarization, branch inhibition, and axon growth. Kif20b may affect these processes by its activities in organizing microtubules and localizing effector proteins. Loss of Kif20b appears to have different consequences for mammalian neuron morphology than loss of the other Kinesin-6 family member Kif23. Understanding how different motor proteins and other microtubule associated proteins enable neurons to reorganize their cytoskeletons during early polarization, axon growth, and arborization is fundamental to knowing how the brain is wired and changes during learning. In addition, there are important implications for human health. Since the Kif20b mouse mutant is a novel model for human microcephaly, it is important to know that there could be defects in connectivity in addition to the small brain size in some cases of microcephaly. Finally, KIF20B is elevated in several cancers [6, 32, 33], and while that could be due to its role in cell division, our data suggest it could also be due to effects on cell motility and morphology that could enhance metastasis.
Days in vitro
We thank Luke Arnell for assistance with polarization and width quantification, Caitlin Cook for assistance with neuron tracing. Many thanks to Bettina Winckler, Kevin Pfister, Christopher Bott, Xiaowei Lu, Jing Yu, and Jung-Bum Shin for advice and discussions. We are grateful to Fabienne Pirollet, Naoyuki Inagaki, Tamar Sapir and Orly Reiner for antibodies and plasmids.
This work was supported by funding from the National Institutes of Health (R01 NS076640 to NDD) and the University of Virginia.
Availability of data and material
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
KM collected and analyzed axon length, Shootin1 localization, live imaging of axons, helped with the DCX localization, Tau localization and Tubulin intensity measurements, prepared figures, and helped write the manuscript. TC collected and analyzed polarization, minor neurite and axon morphology, and filopodia data. JNL collected and analyzed in vivo data, performed RT-PCR, helped with the DCX localization, Tau localization and Tubulin intensity measurements. KJ performed RT-PCR. KJ and AS immunostained Kif20b on neurons and electroporated GFP-KIF20B in neurons. ND conceived the project, designed and supervised experiments, analyzed data, and wrote the manuscript. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
All animals were used humanely and in accordance with University of Virginia’s Institutional Animal Care and Committee (IACUC) approved protocols.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Dotti CG, Sullivan CA, Banker GA. The establishment of polarity by hippocampal neurons in culture. J Neurosci. 1988;8(4):1454–68.PubMedGoogle Scholar
- Lewis Jr TL, Courchet J, Polleux F. Cell biology in neuroscience: cellular and molecular mechanisms underlying axon formation, growth, and branching. J Cell Biol. 2013;202(6):837–48.View ArticlePubMedPubMed CentralGoogle Scholar
- Menon S, Gupton SL. Building blocks of functioning brain: cytoskeletal dynamics in neuronal development. Int Rev Cell Mol Biol. 2016;322:183–245.View ArticlePubMedPubMed CentralGoogle Scholar
- van Beuningen SF, Hoogenraad CC. Neuronal polarity: remodeling microtubule organization. Curr Opin Neurobiol. 2016;39:1–7.View ArticlePubMedGoogle Scholar
- Hirokawa N, Niwa S, Tanaka Y. Molecular motors in neurons: transport mechanisms and roles in brain function, development, and disease. Neuron. 2010;68(4):610–38.View ArticlePubMedGoogle Scholar
- Baron RD, Barr FA. The Kinesin-6 Members MKLP1, MKLP2 and MPP1. In: Kozielski FSBF, editor. Kinesins and cancer. Dordrecht: Springer Netherlands; 2015. p. 193–222.Google Scholar
- Lin S, Liu M, Mozgova OI, Yu W, Baas PW. Mitotic motors coregulate microtubule patterns in axons and dendrites. J Neurosci. 2012;32(40):14033–49.View ArticlePubMedPubMed CentralGoogle Scholar
- Yu W, Cook C, Sauter C, Kuriyama R, Kaplan PL, Baas PW. Depletion of a microtubule-associated motor protein induces the loss of dendritic identity. J Neurosci. 2000;20(15):5782–91.PubMedGoogle Scholar
- del Castillo U, Lu W, Winding M, Lakonishok M, Gelfand VI. Pavarotti/MKLP1 regulates microtubule sliding and neurite outgrowth in Drosophila neurons. Curr Biol. 2015;25(2):200–5.View ArticlePubMedGoogle Scholar
- Goldstein AY, Jan YN, Luo L. Function and regulation of Tumbleweed (RacGAP50C) in neuroblast proliferation and neuronal morphogenesis. Proc Natl Acad Sci U S A. 2005;102(10):3834–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Abaza A, Soleilhac JM, Westendorf J, Piel M, Crevel I, Roux A, Pirollet F. M phase phosphoprotein 1 is a human plus-end-directed kinesin-related protein required for cytokinesis. J Biol Chem. 2003;278(30):27844–52.View ArticlePubMedPubMed CentralGoogle Scholar
- Kamimoto T, Zama T, Aoki R, Muro Y, Hagiwara M. Identification of a novel kinesin-related protein, KRMP1, as a target for mitotic peptidyl-prolyl isomerase Pin1. J Biol Chem. 2001;276(40):37520–8.View ArticlePubMedGoogle Scholar
- Dwyer ND, Manning DK, Moran JL, Mudbhary R, Fleming MS, Favero CB, Vock VM, O'Leary DD, Walsh CA, Beier DR. A forward genetic screen with a thalamocortical axon reporter mouse yields novel neurodevelopment mutants and a distinct Emx2 mutant phenotype. Neural Dev. 2011;6(1):3.View ArticlePubMedPubMed CentralGoogle Scholar
- Janisch KM, Vock VM, Fleming MS, Shrestha A, Grimsley-Myers CM, Rasoul BA, Neale SA, Cupp TD, Kinchen JM, Liem Jr KF, Dwyer ND. The vertebrate-specific Kinesin-6, Kif20b, is required for normal cytokinesis of polarized cortical stem cells and cerebral cortex size. Development. 2013;140(23):4672–82.
- Sapir T, Levy T, Sakakibara A, Rabinkov A, Miyata T, Reiner O. Shootin1 acts in concert with KIF20B to promote polarization of migrating neurons. J Neurosci. 2013;33(29):11932–48.View ArticlePubMedGoogle Scholar
- Salzberg A, D'Evelyn D, Schulze KL, Lee JK, Strumpf D, Tsai L, Bellen HJ. Mutations affecting the pattern of the PNS in Drosophila reveal novel aspects of neuronal development. Neuron. 1994;13(2):269–87.View ArticlePubMedGoogle Scholar
- Lipka J, Kapitein LC, Jaworski J, Hoogenraad CC. Microtubule-binding protein doublecortin-like kinase 1 (DCLK1) guides kinesin-3-mediated cargo transport to dendrites. EMBO J. 2016;35(3):302–18.View ArticlePubMedPubMed CentralGoogle Scholar
- Toriyama M, Shimada T, Kim KB, Mitsuba M, Nomura E, Katsuta K, Sakumura Y, Roepstorff P, Inagaki N. Shootin1: a protein involved in the organization of an asymmetric signal for neuronal polarization. J Cell Biol. 2006;175(1):147–57.View ArticlePubMedPubMed CentralGoogle Scholar
- Shimada T, Toriyama M, Uemura K, Kamiguchi H, Sugiura T, Watanabe N, Inagaki N. Shootin1 interacts with actin retrograde flow and L1-CAM to promote axon outgrowth. J Cell Biol. 2008;181(5):817–29.View ArticlePubMedPubMed CentralGoogle Scholar
- Maliga Z, Junqueira M, Toyoda Y, Ettinger A, Mora-Bermudez F, Klemm RW, Vasilj A, Guhr E, Ibarlucea-Benitez I, Poser I, et al. A genomic toolkit to investigate kinesin and myosin motor function in cells. Nat Cell Biol. 2013;15(3):325–34.View ArticlePubMedGoogle Scholar
- Fu X, Brown KJ, Yap CC, Winckler B, Jaiswal JK, Liu JS. Doublecortin (Dcx) family proteins regulate filamentous actin structure in developing neurons. J Neurosci. 2013;33(2):709–21.View ArticlePubMedPubMed CentralGoogle Scholar
- Ettinger AW, Wilsch-Brauninger M, Marzesco AM, Bickle M, Lohmann A, Maliga Z, Karbanova J, Corbeil D, Hyman AA, Huttner WB. Proliferating versus differentiating stem and cancer cells exhibit distinct midbody-release behaviour. Nat Commun. 2011;2:503.View ArticlePubMedPubMed CentralGoogle Scholar
- Kuo TC, Chen CT, Baron D, Onder TT, Loewer S, Almeida S, Weismann CM, Xu P, Houghton JM, Gao FB, et al. Midbody accumulation through evasion of autophagy contributes to cellular reprogramming and tumorigenicity. Nat Cell Biol. 2011;13(10):1214–23.View ArticlePubMedPubMed CentralGoogle Scholar
- Singh D, Pohl C. A function for the midbody remnant in embryonic patterning. Commun Integr Biol. 2014;7:e28533.View ArticlePubMedPubMed CentralGoogle Scholar
- Gallo G. Mechanisms underlying the initiation and dynamics of neuronal filopodia: from neurite formation to synaptogenesis. Int Rev Cell Mol Biol. 2013;301:95–156.View ArticlePubMedGoogle Scholar
- Pacheco A, Gallo G. Actin filament-microtubule interactions in axon initiation and branching. Brain Res Bull. 2016;126(Pt 3):300–10.View ArticlePubMedGoogle Scholar
- Dagenbach EM, Endow SA. A new kinesin tree. J Cell Sci. 2004;117(Pt 1):3–7.View ArticlePubMedGoogle Scholar
- Witte H, Neukirchen D, Bradke F. Microtubule stabilization specifies initial neuronal polarization. J Cell Biol. 2008;180(3):619–32.View ArticlePubMedPubMed CentralGoogle Scholar
- Tanaka EM, Kirschner MW. Microtubule behavior in the growth cones of living neurons during axon elongation. J Cell Biol. 1991;115(2):345–63.View ArticlePubMedGoogle Scholar
- Bastmeyer M, O'Leary DD. Dynamics of target recognition by interstitial axon branching along developing cortical axons. J Neurosci. 1996;16(4):1450–9.PubMedGoogle Scholar
- Dwivedy A, Gertler FB, Miller J, Holt CE, Lebrand C. Ena/VASP function in retinal axons is required for terminal arborization but not pathway navigation. Development. 2007;134(11):2137–46.View ArticlePubMedPubMed CentralGoogle Scholar
- Kanehira M, Katagiri T, Shimo A, Takata R, Shuin T, Miki T, Fujioka T, Nakamura Y. Oncogenic role of MPHOSPH1, a cancer-testis antigen specific to human bladder cancer. Cancer Res. 2007;67(7):3276–85.View ArticlePubMedGoogle Scholar
- Liu X, Zhou Y, Liu X, Peng A, Gong H, Huang L, Ji K, Petersen RB, Zheng L, Huang K. MPHOSPH1: a potential therapeutic target for hepatocellular carcinoma. Cancer Res. 2014;74(22):6623–34.View ArticlePubMedGoogle Scholar