Fig 2: Kif20b expression and localization in cultured cortical neurons. A RT-PCR detects Kif20b weakly in neuron culture lysates (E14.5 plus 4DIV) after 30 cycles. Bands are stronger in MEFs. Bands are weaker in mutant samples due to the splice mutation [14]. B Immunocytochemistry for Kif20b does not detect signal in control neurons (+/+) above background seen in mutant (-/-) neurons. Images shown are of a polyclonal antibody to the full length protein [11], but are representative of staining appearance of five independent antibodies (see Methods). Scale bars, 20 µm. C Plasmids encoding GFP-tagged human KIF20B and mCherry were electroporated into mouse cortical neurons and imaged live. GFP puncta were distributed throughout the neurites and axon (arrows, inset). C’ and C”: Some GFP puncta moved anterogradely in the axon (arrows point to the same puncta at different time points). 10 µm scale bar and 5 µm in insets

Fig 3: Kif20b mutant neurons have wider axons and minor neurites than controls. A Representative images of polarized control (+/-) and mutant (-/-) neurons cultured for two days and stained for actin (green), beta-III-tubulin (red) and nuclei (DAPI, blue). Insets show magnified axon segments for comparison. B Average width (±s.e.m.) of mutant axons (-/-, gray line) is greater than controls (+/-, black line) at each given distance from the edge of the soma. n = 79 +/- and 65 -/- axons. C Average width (±s.e.m.) of mutant minor neurites (gray line) is increased over controls (black line) at each given distance from the cell body. n = 323 +/- and 262 -/- minor neurites. D Representative images of and linescans across control (+/-) and mutant (+/-) axons stained for beta-III tubulin show that tubulin intensity was lower, more irregular, and more spread out in mutants than controls. Dotted lines represent where tubulin intensity linescans were taken. E Inset bar graph shows average total tubulin intensity summed across the width 25 µm from the soma was significantly decreased in mutant axons compared to controls. n = 15 each +/- and -/- from 1 experiment. *, p <.05; **, p <.01; ***, p <.001; ****, p <10-9. 10 µm scale bars

Fig 4: Kif20b mutant neurons’ growth cones have longer filopodia. A Representative images of control and Kif20b mutant growth cones of dissociated E14.5 cortical neurons cultured 2 DIV and immunostained to label neuronal tubulin (TuJ1, red), actin (phalloidin, green), and nuclei (DAPI, blue). Scale bars 10 µm. B The growth cones of axons are significantly larger in area than those of minor neurites, but similar in control (+/-, gray bars) and mutant (-/-, white bars) cells. C The average number of filopodia per growth cone did not differ significantly in Kif20b mutant neurons on either axons or minor neurites. For B-C, n = 38 control (+/-) axonal and 54 minor neurite growth cones; mutant (-/-) n = 37 axonal and 49 nascent dendritic growth cones. D Kif20b mutant neurons have longer filopodia on both axonal and nascent dendritic growth cones. Axonal filopodia are slightly but significantly longer than dendritic filopodia regardless of genotype. n = 217 control and 308 mutant filopodia from 39 and 37 axonal growth cones, and n = 155 +/- and 134 -/- filopodia from 25 and 22 minor neurite growth cones respectively. For (A-D), error bars are ± s.e.m. *, p < 0.01; **, p <10-3; ***p <10-5; ****p <10-6; *****, p <10-8; n.s., not significant, t-test. E In Kif20b mutant neurons, a higher percentage of growth cone filopodia contain detectable microtubules than in controls, in both axons and minor neurites. n = 217 for +/- and 308 -/- filopodia on axonal growth cones and n =155 +/- and 134 -/- filopodia on minor neurite growth cones. *, p <0.05 Fisher’s exact test. F Axonal growth cone labeled for actin (green) and tubulin (red). Arrowheads mark filopodia containing microtubules

Fig 5: Summary of phenotypes of Kif20b mutant neurons seen in culture. A Normal cortical neurons transition from multipolar (Stage 2) to polarized with a single axon (Stage 3) properly when Kif20b is present. Kif20b may organize or stabilize microtubule arrays in the axon and growth cone, keeping the axon thin and limiting its branching, and preventing microtubule invasion of filopodia. Kif20b may also tether Shootin1 at the microtubule plus ends in the axonal growth cone, which reinforces axonal growth. Kif20b may thereby enhance stability and inhibit axon retraction. B When Kif20b is absent, some neurons fail to transition from stage 2 to stage 3. Those that do polarize show a variety of phenotypes: longer minor neurites, shorter axons with extra branches, wider axons and minor neurites, and longer filopodia. Microtubules appear more loosely organized and extend into more filopodia. Shootin1 enrichment in the growth cone is severely reduced