Bicaudal-D Regulates Fragile X Mental Retardation Protein Levels, Motility, and Function during Neuronal Morphogenesis

Summary 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 [1–3], and these functions are closely associated with transport of FMRP complexes within neurites by microtubule-based motors [2–4]. 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.


Primary neuronal cell culture
Third instar larval brains were isolated, disaggregated and cells cultured as previously described in [11]. Cells appear round immediately after plating and differentiate during the following three to four days. Consistent with previous observations [11], we were unable to observe a clearly differentiated axon in the vast majority of cells. Cells were analyzed after 5 to 7 days of culture.

In vivo imaging
Third instar larvae were dissected and mounted in Ca 2+ -free saline (130 mM NaCl, 36 mM sucrose, 5 mM KCl, 4 mM MgCl 2 , 2 mM CaCl 2 , 5 mM HEPES (pH 7.3) and 0.5 mM EGTA) to prevent muscle contraction. Primary neurons were plated in Lab-Tek chambered cover glasses (Thermo Scientific). Larvae and cells were imaged at 22-23°C using a spinning disk UltraVIEW ERS confocal microscope (PerkinElmer) using a 63/1.4NA oil-immersion PlanApo Olympus objective and an Orca ER CCD camera (Hamamatsu). Image series, from a single focal plane, lasted from 2 to 6 min both for cells and larvae. Larval fillets were imaged only during the first 20 min following dissection, although robust cargo transport in these and other neurons was maintained for at least 30 min after dissection. Chordotonal organ neurons were imaged because, unlike in other larval neuronal cell types in situ, it was possible to image and analyze FMRP::GFP transport due to their relatively simple shape, which permits filming in a single focal plane.

Analysis of FMRP::GFP motility
Motile FMRP::GFP particles were tracked manually using ImageJ and run length, net movement and velocity calculated using Microsoft Excel. Run length was defined as the distance a particle moves before a reversal or a pause. Net displacement was defined as the total direct distance a particle moves during the whole time of observation.
Mean run length was calculated by averaging the values for all runs for a given genotype (i.e. considering the different movements of the same particle as independent events). The mean velocity of ~ 0.2 μm/s, stated in the legend to figure 1, was also calculated by averaging the values for all runs for a given genotype. Mean net displacement was calculated by averaging the net displacement of each individual particle. The data in figure 3 quantifying FMRP::GFP motility in neuronal culture were obtained from 40 to 60 movies per genotype. 6 to 10 independent primary cultures were prepared per genotype. Quantification of motility in chordotonal organs was based on imaging 12 and 18 BicD mutant and wild-type larval fillets, respectively.

GST pull downs, immunoprecipitation from embryo extracts and yeast two hybrid assays
The K730M substitution is encoded by the BicD r11 allele [1]. In heteroallelic combinations with BicD null mutant alleles, BicD r11 behaves as a null allele based on phenotypic analysis [1]. As described above, only a very small proportion of BicD null mutants eclose from the pupal case, with those that do dying in the next few hours [1, 2].
Thus, K730M mutant material cannot be readily used for biochemistry. We therefore previously assayed the effect of this substitution on Egl and Rab6 binding by expressing epitope-tagged versions of wild-type or K730M BicD in egg chambers in the presence of endogenous BicD and/or with binding assays with recombinant proteins or the yeast two- Yeast two-hybrid assays used the vectors pGAD424 and pGBDU-C1 [13].
pGAD424 carries the LEU2 marker gene, and pGBDU-C1 carries the URA3 marker gene. Fragments of Drosophila BicD or FMRP were PCR-amplified and cloned in frame with the Gal4p transactivation domain in pGAD424 or the Gal4p DNA binding domain in pGBDU-C1, respectively. The integrity of all inserts was confirmed by sequencing.
Plasmids were transformed into the yeast two-hybrid strains PJ69-4a or PJ69-4α [13], and pairs of BicD and FMRP fragments introduced into the same cells by mating.
Successful mating was confirmed by plating on -LEU/-URA medium. Interactions between the bait and prey, which result in expression of HIS3 and ADE2 were assayed by plating on -HIS/-LEU/-URA medium for 72-96 hours. Due to the relatively weak interaction of FMRP and the BicD CTD, growth conditions were made less stringent by including adenine in the medium.

Liquid chromatography mass spectrometry/mass spectrometry
Gel bands were excised, washed, alkylated, and in-gel digested with trypsin. The extracted peptides were separated by nanoscale liquid chromatography (LC Packings, Armsterdam, Netherlands) on a reverse phase C18 column (150 X 0.075 mm ID, flow rate 0.25 μl/min). The eluate was introduced directly into a Q-STAR hybrid tandem mass spectrometer (MDS Sciex, Ontario, Canada). Spectra were searched against a NCBI nonredundant database with MASCOT MS/MS Ions search (www.matrixscience.com).

Immunostaining and detection of mitochondria
Dissection of third instar larval neuromuscular system fillets was performed as previously described [15]. Larvae were dissected in Ca 2+ -free solution and fillets fixed immediately afterwards (15 min in 4% paraformaldehyde in PBS). Samples were then blocked for 1 h at room temperature in 5% bovine serum albumin (Sigma) in PBS/0.1% NP-40. Primary antibodies (mouse anti-GFP (1:500; Roche) and rabbit anti-HRP (1:300; Jackson Immunoresearch)) were applied overnight at 4°C and secondary, fluorescentlyconjugated antibodies were incubated for 1 h at room temperature. 6 washes of 15 min each with PBS/0.1% NP-40 were conducted after antibody incubations.
Dorsal ddaC neurons were imaged with a Biorad Radiance or Zeiss 710 confocal microscope using a 25x/0.8 NA oil immersion or a 20x/0.8NA air objective. Images shown in figures 4 and S2 are a projection of 30 to 40 z-sections. ddaC neurons are an established model to study neuronal morphology as they present the following advantages: 1) they are class IV neurons, i.e. they harbour the highest level of morphological complexity that can be found in the larval epithelium, thus allowing the scoring of detailed aspects of neuronal shape and 2) ddaC neurons are the only class IV neurons present in the dorsal side of a larva; it is therefore possible to correctly attribute dendritic arbors to the cell body they originated from, allowing a single neuron to be scored.
For immunofluorescence of primary neurons, cells were fixed in 4% formaldehyde for 10 min. BicD was detected using antibodies raised in rabbit against amino acids 1-543 of Drosophila BicD, expressed in bacteria and purified using an Nterminal His tag. FMRP was detected using a mixture of mouse anti-FMRP 5A11 and 5B6 (1:500). Samples were imaged using a Zeiss 710 confocal microscope. . Note that BicD also has a much wider distribution than its cargoes on other cell types [16,17] (MD and SLB, unpublished observations). This presumably reflects a sizeable pool of an auto-inhibited form of the protein: in the absence of binding of a cargo, the CTD engages in an interaction with an N-terminal region, thus preventing recruitment of the dynein/dynactin complex to the N-terminal region [17]. It is also possible that BicD transports other cargoes in these cells and that this contributes to the broad distribution of BicD relative to FMRP. There are significantly more branches in heterozygous than in homozygous mutant larvae (P < 0.0004 (t-test)). Branching is also reduced, compared to wild-type, following overexpression of Fmr1::GFP under control of ppk-GAL4 in the presence of wild-type levels of endogenous FMRP (i.e. Fmr1 +/+ ) (***, P < 0.001; t-test). Similar effects on neuronal morphology of Fmr1 loss-of-function and overexpression have previously been reported in other neurons; Morales et al. [18] found that both Fmr1 null mutations and overexpression caused reduced extension of the axons of DC neurons in the adult Drosophila brain. The authors concluded that dose of FMRP may be under stringent regulation and that precise levels are required for proper function. Our results in dorsal ddaC neurons support this conclusion. ppk-GAL4 driven expression of Fmr1::GFP fully rescues terminal branch number defects in Fmr1 +/-ddaC neurons. Thus, the FMR1::GFP fusion protein is functional. Note that is has previously been reported that terminal branch number in larval da neurons is increased in Fmr1 mutants [4]. However, these authors analyzed the entire cluster of da neurons on the ventral side, whereas we analyzed the phenotype in a single neuron (ddaC; the only class IV da neuron in the dorsal cluster). The differential requirements for Fmr1 in controlling the morphology of different neuronal cells is consistent with previous findings in Drosophila brains. Morales et al. [18] showed that  shown is representative of two independent experiments. Note that the lower abundance FMRP isoform present in larval brain extracts was not detectable in these D-mel cell extracts. In larval brain extracts, the abundance of both isoforms is dependent on BicD (A). Quantification of microtubule polarity in wild-type primary cultured neurons, assayed by filming the plus-end-tracking marker EB1::GFP (under C155-GAL4 control) [10]. A similar number of microtubules grow away (anterograde) and toward (retrograde) the cell body. n = 306 comets in total, from 27 cells. Table S1. Yeast two hybrid interactions tested between BicD and FMRP c, the K730M mutation does not inhibit binding of the BicD CTD to other copies of BicD in the yeast two hybrid assay [3], indicating that is does not have a general toxic effect on protein folding or stability.