Bicaudal-D regulates fragile X mental retardation protein levels, motility, and function during neuronal morphogenesis

Abstract

The expression of the RNA-binding factor Fragile X mental retardation protein (FMRP) is disrupted in the most common inherited form of cognitive deficiency in humans. FMRP controls neuronal morphogenesis by mediating the translational regulation and localization of a large number of mRNA targets, and these functions are closely associated with transport of FMRP complexes within neurites by microtubule-based motors. However, the mechanisms that link FMRP to motors and regulate its transport are poorly understood. Here we show that FMRP is complexed with Bicaudal-D (BicD) through a domain in the latter protein that mediates linkage of cargoes with the minus-end-directed motor dynein. We demonstrate in Drosophila that the motility and, surprisingly, levels of FMRP protein are dramatically reduced in BicD mutant neurons, leading to a paucity of FMRP within processes. We also provide functional evidence that BicD and FMRP cooperate to control dendritic morphogenesis in the larval nervous system. Our findings open new perspectives for understanding localized mRNA functions in neurons.

Graphical abstract

Figure 1

FMRP Is a Novel BicD-Associated Protein and Colocalizes with BicD in Moving Particles within Neurons (A and B) Pull-down from embryo extracts via the wild-type or K730M BicD CTD fused to GST. (A) Coomassie stain; arrow indicates the FMRP band enriched on the wild-type CTD and asterisks mark GST fusion proteins. (B) The identity of FMRP was confirmed by western blotting. (C) Endogenous FMRP specifically coprecipitates with endogenous BicD from Drosophila embryonic extracts in an RNase-sensitive manner. BicD:Egl complexes are insensitive to RNase. IP, immunoprecipating antibody; IB, immunoblotting antibody. (D) Unlike the known Egl-interacting proteins BicD and Dynein light chain (Dlc) [28], FMRP is not immunoprecipitated with Egl::GFP. Asterisk marks endogenous Egl that is immunoprecipitated with Egl::GFP in an RNase-independent manner. (E) The pull-down of FMRP from extracts by BicD-CTD is RNase sensitive. Load is 1% of input in (A) to (E). (F) FMRP::GFP and BicD::mCherry colocalize in puncta in live, primary Drosophila neurons. See Movie S1A for time-lapse. Puncta have maximum instantaneous velocities of ∼0.7 μm/s, consistent with the involvement of molecular motors. The mean velocity of these motile particles was ∼0.2 μm/s, which is similar to that reported for transported mRNPs in other neuronal cell types [15, 16]. (G) Stills of Movie S1B showing a bidirectional cargo containing FMRP::GFP and BicD::mCherry (arrow). (H) Kymographs of the particle in (G). Left to right in the kymographs represents movement away from the cell body. Bars represent 2 μm. A red and green image set was captured every 2 s.

Figure 2

BicD Is Required for FMRP Accumulation in Neuronal Processes and to Maintain FMRP Protein, but Not RNA, Levels (A) Appearance of FMRP::GFP, expressed with the panneuronal driver C155-GAL4, within the chordotonal organ neuron cluster of wild-type and BicD mutant third instar larvae (de, dendrites; ax, axons; cb, cell bodies). Images are projections of four z-sections of ∼1 μm each. See Supplemental Experimental Procedures for details of BicD null genotype. (B) Western blot showing strongly reduced levels of endogenous FMRP and FMRP::GFP (asterisk) proteins in extracts from BicD mutant third instar brains, lacking detectable BicD protein. β-actin acts as a loading control. A similar reduction in endogenous FMRP levels is also observed in the absence of FMRP::GFP (Figure S3A). The intensity of the signal of FMRP in BicD mutant extracts was 36% ± 11.5% of the wild-type (mean ± SEM, n = 3; measured from the major endogenous isoform). (C) FMRP levels are unchanged by partial loss-of-function mutations in dynein and kinesin-1 heavy chains, when BicD is strongly overexpressed (o.e.) with C155-GAL4 or in Fmr1 null mutants. (D) Fmr1 mRNA levels, normalized to a β-actin mRNA control, are not significantly different in wild-type and BicD mutant third instar larval brains. For each genotype, n = 3 independent quantitative RT-PCR experiments (each done in duplicate). Error bars represent standard error of the mean (SEM). WT represents wild-type in this and all other figures.

Figure 3

BicD Regulates FMRP Motility within Neurons in Culture and In Situ (A) The proportion of FMRP::GFP particles within neurites of primary cultured neurons is reduced in BicD mutants and increased by 2-fold overexpression (o.e.) of BicD (via tub-BicD::mCherry). Panneuronal overexpression (via C155-GAL4) of a dominant-negative dynactin subunit, ΔGlued (ΔGl), also decreases the proportion of FMRP::GFP particles in neurites. (B) The accumulation of FMRP::GFP in distal regions of processes of cultured neurons is similarly sensitive to BicD dosage and ΔGl. y axis is percentage of total FMRP::GFP particles in the neuron (i.e., including the cell body). n = 200–393 in (A) and (B). (C–F) Mean values of run length and net displacement (disp.) for only the motile subset of FMRP::GFP particles in primary cultured neurons (C, D; n = 125–182) and chordotonal organs in situ (E, F; n = 25 and 27 in WT and BicD mutant, respectively). Run lengths are defined as the distance of travel between reversals or pauses. The frame rate was between 1 and 1.4 s⁻¹ for neurons in culture and 2 s⁻¹ for larvae. Error bars represent SEM; ^∗∗∗p < 0.001; ^∗∗p < 0.01. t tests were used for statistical evaluations (compared to WT), except in (A) (Fisher's exact test using raw numbers).

Figure 4

BicD Is Required for Correct Dendritic Morphogenesis and Requires FMRP to Induce Dendritic Branching (A–G) Representative confocal projections of dorsal ddaC neurons within segment A2 of third instar larvae, visualized with the class IV da-specific driver ppk-GAL4 and UAS-CD8::GFP. Red arrows show axons; other processes are dendritic. BicD o.e., UAS-BicD overexpressed specifically in class IV da neurons with ppk-GAL4; Df, Df(3R)BSC526. Insets show higher magnification views of the typical density of dendritic termini. (H) Quantification of number of branch termini. (I) Quantification of the number of branches of each order (defined as in the schematic cut-away of a neuron [right]). In (H) and (I), the values for Fmr1^Δ50M and Fmr1^Δ50M, BicD o.e. are not significantly different (t test). Note that the terminal processes of neurons overexpressing BicD are frequently shorter than wild-type (e.g., compare regions near asterisks in A and F; terminal branches frequently extend back to this region in WT, but not in BicD o.e.). n = 6 or 7 neurons (from 4–7 larvae) for each genotype in (H) and (I). Error bars in (H) and (I) represent SEM; ^∗∗∗p < 0.001 (t tests, compared to WT).