(A) Schematic diagram of rat ZBP1 domain structure and amino acid sequence of the linker region between the KH2 and KH3 domains. Phosphorylation site at Tyr396 is indicated. (B) Sequence alignment of ZBP1/IMP1 in four species and homologs (IMP2, IMP3, and VgRBP). Although the phosphorylated tyrosine position is conserved among all members of the IMP family, the sequence surrounding the tyrosine is conserved only in ZBP1. Red residues, phosphorylatable tyrosine; Blue residues, conserved residues between species or within ZBP1 homologs. Xen., Xenopus. (C) Specificity of anti-phospho-ZBP1 antibody. Indicated plasmids were transfected into 293T cells and analyzed with immunoblotting. GFP-ZBP1: GFP-fused rat ZBP1. Overexpression of constitutively-active Src tyrosine kinase increased the immunoreactive band corresponding to GFP-ZBP1, whereas no signal against the Y396F mutant was detected. (D) The effect of Src family tyrosine kinase (SFK) inhibitor. GFP-ZBP1 and Fyn, a neuronal SFK, were transfected into 293T cells and the cells were treated with a SFK inhibitor. The SFK-specific inhibitor PP2, but not PP3 (an inactive derivative of PP2), inhibited the increase in ZBP1 phosphorylation induced by Fyn. (E) Time course of Tyr396 phosphorylation of endogenous ZBP1 in rat cortical neurons after stimulation by BDNF. Stimulation of cortical neurons with BDNF resulted in an increase in Tyr396 phosphorylation of endogenous ZBP1, which reached its maximum at 15 minutes and decreased thereafter.

(A) Immunofluorescence (IF) of axons and growth cones of rat cortical neurons stimulated with BDNF for 10 minutes and analyzed with phospho- and total ZBP1 antibodies. In merged images (bottom panel), phospho-ZBP1 (Pi-ZBP1), total ZBP1, and F-actin were represented by red, green, and blue, respectively. Scale bar = 10 µm. (B) Quantification of fluorescent intensities of phospho-ZBP1 and total ZBP1 signals in growth cones under the same treatments as in (A). Data are shown as mean ± SEM (n = 10). Note that phospho-ZBP1 signals increased more than total ZBP1 signals after BDNF stimulation. *p < 0.05, **p <0.001, significantly different from control (no stimulation) using a student t-test.

(A) Localization of wild type (wt) and Y396F mutant forms of flag-tagged mCherry-chick ZBP1 (fZBP1) in cell bodies and growth cones. Both wt and Y396F forms of fZBP1 were co-transfected into cortical neurons with GFP as a morphologic marker. Only growth cones over 100 µm distal from the cell body were selected and analyzed. Cell body and growth cone images were acquired with the same exposure time. The fluorescence intensity of growth cone images from cells transfected with wt or mutant ZBP1 were both enhanced 10 times from the original images for improved visualization. Scale bar = 10 µm. (B) Quantification of fluorescent intensities of the flag-tagged ZBP1 signals in growth cones using IF with anti-flag antibody. Relative intensity is represented as a ratio of fluorescent intensity in a growth cone (GC) to that in a cell body (CB). Data are shown as mean ± SEM (n = 10). n.s., not significantly different using student t-test. (C) Fluorescence in situ hybridization (FISH) of β-actin mRNA using antisense and scrambled probes (right panels). Note that scrambled probes did not detect a significant signal. GFP expression in the same growth cone is shown (left panels). (D) In situ hybridization for β-actin mRNA in growth cones of cortical neurons transfected with wild type or Y396F mutant of fZBP1 (bottom panels). GFP was co-transfected as a morphologic marker (top panels). β-actin mRNA was detected using three oligonucleotide probes labeled with DIG, and was visualized with anti-DIG antibody. Scale bar = 10 µm. (E) Quantification of fluorescent intensities of β-actin mRNA signals in growth cones under the same treatments as in (D). Data are indicated as mean ± SEM (n = 10). No significant difference in β-actin mRNA levels between neurons that express wild type or Y396F mutant was apparent using a student’s t-test.

(A) Immunostaining of axons and growth cones of cortical neurons transfected with wild type or Y396F mutant form of flag-tagged mCherry-chick ZBP1 (fZBP1). GFP was also co-transfected as a morphologic marker. The transfected neurons were stimulated with BDNF for 15 minutes, and then were processed for IF using anti-β-actin and anti-flag antibodies. In merged images, β-actin protein, fZBP1 (wild type or Y396F mutant), and GFP are displayed in red, blue, and green respectively. Scale bar = 10 µm. (B) Quantification of (A). BDNF stimulation results in an increase in β-actin protein localization in neurons expressing the wild type, but not the mutant (Y396F). Data are shown as mean ± SEM (n = 10). *p < 0.05, significantly different from 0 min using one-way ANOVA and Bonferroni post hoc test. (C) The effect of anisomycin on β-actin protein localization in growth cones. Anisomycin (Ani) was incubated for 30 min, and then BDNF was applied for 15 min. Data are indicated as mean ± SEM (n = 10). *p < 0.05, significantly different from control using student t-test.

Neurons were transfected with designated reporters and cultured as neuron balls extending axons over 1 mm in length (see methods, and Fig. S6). (A) Vector structures of fluorescent reporters. Both GFP and mCherry reporters were fused with a membrane-targeting sequence (dual palmitoylation from GAP43) and a sequence for shorter half-life (d2EGFP and d2mCherry). The GFP reporter (gd2G) contains the 3’UTR sequence of rat β-actin, but the mCherry reporter (gd2R) does not contain this sequence. (B) Representative images of growth cones of cortical neurons in neuron ball cultures co-transfected with gd2G + 3’UTR and gd2R. Pseudocolored images show that the absence of fluorescent signal is indicated by black, with increasing fluorescence indicated by transitions to blue, green, yellow, and red. Note that BDNF stimulation increased the fluorescence of gd2G + 3’UTR, but not gd2R, after 60 min. Scale bar = 10 µm. (C) Effect of the β-actin 3’UTR on expression of the GFP reporter in growth cones. The GFP reporter (gd2G or gd2G + 3’UTR) was co-transfected with gd2R into cortical neurons. Neurons were stimulated with BDNF, and fluorescent ratios of the GFP reporters to gd2R were calculated at indicated times, and then normalized to the initial fluorescence ratio. The fluorescent ratio of gd2G+3’UTR to gd2R increased from 30 min after BDNF stimulation, but the fluorescent ratio of gd2G did not. (D) The SFK inhibitor PP2, but not its inactive derivative, PP3, inhibited the increase of the fluorescent ratio of gd2G + 3’UTR to gd2R after BDNF stimulation.

(A) Representative images of Xenopus laevis growth cones prior to (00’) and 30 minutes following (30’) pipette application of BDNF onto wild type GFP-ZBP1 (wt) and GFP-ZBP1 Y396F (Y396F) transfected neurons. Insets: Selected cells expressing GFP. Traces of individual growth cones are shown in the right panels. (B) Average turning angles of growth cones in response to BDNF. (C) Average lengths of net neurite extension. Note that the mutant suppressed turning angle, while neurite extension was not affected.