Co-regulation of mRNA translation by TDP-43 and Fragile X Syndrome protein FMRP

For proper mammalian brain development and functioning, the translation of many neuronal mRNAs needs to be repressed without neuronal activity stimulations. We have discovered that the expression of a subclass of neuronal proteins essential for neurodevelopment and neuron plasticity is co-regulated at the translational level by TDP-43 and the Fragile X Syndrome protein FMRP. Using molecular, cellular and imaging approaches, we show that these two RNA-binding proteins (RBP) co-repress the translation initiation of Rac1, Map1b and GluR1 mRNAs, and consequently the hippocampal spinogenesis. The co-repression occurs through binding of TDP-43 to mRNA(s) at specific UG/GU sequences and recruitment of the inhibitory CYFIP1-FMRP complex by its glycine-rich domain. This novel regulatory scenario could be utilized to silence a significant portion of around 160 common target mRNAs of the two RBPs. The study establishes a functional/physical partnership between FMRP and TDP-43 that mechanistically links several neurodevelopmental disorders and neurodegenerative diseases. Electronic supplementary material The online version of this article (doi:10.1007/s00401-016-1603-8) contains supplementary material, which is available to authorized users.

The wild type and mutated forms of 3'UTR and CDS of Rac1 were cloned into NotI/AflIII sites of phSYN-myr-dEGFP-BDNF 3'UTR plasmid DNA, a generous gift from Dr. Baoji Xu's lab, Georgetown University, Washington, DC 20057, USA, to replace BDNF-3'UTR in the sites of phSYN-myr-dEGFP-BDNF 3'UTR plasmid and to generate pmyr-dEGFP-3'UTR. The pmyr-dEGFP-based plasmid was used to track the dendritic local translation of the corresponding reporter constructs. All the primers used for cloning are listed in Table S4.

Transfection of plasmid DNAs and RNAi oligos
For RNAi knockdown, the cultured hippocampal neurons at DIV 4 or DIV 12 were transfected with different RNAi oligos: TDP-43 RNAi oligo (TDP-si), FMRP RNAi oligo (FMRP-si) or control oligo (Sc) with the use of lipofectamine 2000 (Invitrogen) following the standard protocol [12]. In one set of experiments, the hippocampal neurons grown on poly-D-lysine coated cover slips. At DIV 12 they were co-transfected with pGFP-FMRP (2µg) or pGFP (2µg) and TDP-si or Sc oligo. Sc and TDP-si oligos are same as we used in Majumder et al. (2012).
Transfection of HEK293T cells with different plasmid DNAs and FMRP-si oligo (Sigma Aldrich) was performed using the lipofectamine 2000 (Invitrogen) following standard protocol.
For quantitative RT-PCR analysis, the total RNA was isolated using the Trizol method. The first strand cDNA synthesis was done using the Superscript RT (Invitrogen). Real time PCR was performed using the Light Cycler machine (Roche biochemical sciences) and appropriate primers corresponding to mouse Rac1 (NM_009007.2), Map1b (NM_008634.2), GluR1 (Gria1, NM_001113325.2) and Gapdh (NM_001289726.1) mRNAs. For semiquantitative RT-PCR analysis, the total RNA was also isolated using the Trizol method. The first strand cDNA was synthesized and PCR was performed using appropriate primers. The full list of RT-PCR primers are given in Table   S5.

Immunofluorescence staining
Primary hippocampal neurons grown on poly-L-lysine coated cover slips were washed with PBS and fixed using 4% PFA. After permeabilization with 0.5% Triton X 100 in PBS, the neurons were blocked with 10% FBS

Fluorescence in situ hybridization (FISH) and combined immunofluorescence (IF) staining
The FISH probes were designed using the program on the website of Biosearch Technology. The reliability of selected probes against 3'UTR of the mouse Rac1 mRNA were tested by OLIGOWALK and mFold software.
The sequence specificities of the selected 3'UTR regions for probe design were checked by BLAST on the NCBI website [10]. FISH was performed using Alexa 488-conjugated 5'-TTGACTGGTTCATTGGTTCA-3'(anti-sense probe 1, green, see Figs 5a, 6 and S5) and Alexa 647-conjugated 5'-ACAAAGGTTCCAGGCAGGAC-3' (antisense probe 2, red, see Fig S5b) targeting the 3'UTR of Rac1 mRNA, and an Alexa 488-conjugated 5'-TGAACCAATGAACCAGTCAA-3'oligonucleotide sense probe was used as the negative control (Life Technologies, Japan). Cells on poly-D-lysine coated glass coverslips were fixed in 3.7% formaldehyde for 30 min, permeabilized in 70% ethanol at 4 °C for 1 hr, incubated with 10% formamide/2xSSC for 10 min at room temperature, and hybridized overnight at 37 °C with the oligonucleotide probe(s) (12.5nM) in hybridization buffer containing 10% formamide, 2xSSC, and 100mg/mL dextran sulphate. Simultaneous imaging of RNA by FISH and proteins by IF staining was carried out using standard protocols [17]. The FISH probes and antibodies were mixed and incubated with the neurons at 37 °C overnight. The cells were then further incubated with secondary antibodies (1:5000 for each, Invitrogen) in 10% formamide/2xSSC for 30-60 min at 37 °C in the dark. The GLOX imaging buffer was used to decrease the phtobleaching of the fluorescent reporters, and the FISH-IF images were acquired by LSM780 (Zeiss, Germany) and analyzed by Zen2010 (Zeiss, Germany) and Metamorph (Molecular devices) software.

Polysome profile analysis
Hippocampal neurons at DIV 6 or HEK293T cells were harvested in PBS containing 100 g/ml cycloheximide at 4°C, and then re-suspended in RSB-150 (10 mM Tris-HCl, pH 7.4, 3mM MgCl2, and 150 mM NaCl) containing 100 g/ml cycloheximide, 40 g/ml digitonin (Calbiochem), 20 U/ml RNasin (Promega), and protease inhibitors (Complete; Roche Diagnsotics). After incubation on ice for 30 minutes, the lysates were centrifuged at 3,000 g for 2 min at 4°C. The cytoplasmic extracts were further cleared by 30 min centrifugation at 11,000 g and then loaded onto a linear gradient of 15-40% (wt/wt) sucrose in RSB-150 and centrifuged at 40,000 rpm for 2.5 hr at 4°C in a Beckman SW41 rotor. After centrifugation, the gradients were monitored at 254 nm using  Table S5. Proteins were isolated and analyzed by Western blotting.

Immunoprecipitation
To study the TDP-43/FMRP interaction, lysates of HEK293T cells co-expressing GFP tagged FMRP and Flag-tagged full-length TDP-43 or different mutant forms of TDP-43 were pre-cleaned with 50% protein G-agarose beads (GE Healthcare) for 30 min at 4 °C, and then incubated with anti-Flag (Sigma Aldrich) overnight. Protein Gagarose beads were then added again and the incubation continued for another 4 hr. The bound precipitates were washed 3-4 times and analyzed by Western blotting using anti-GFP (Sigma Aldrich), anti-Flag antibodies.
In Vitro RNA-Protein binding assay using biotinylated oligoneucleotides DNA oligonucleotides corresponding to different regions of Rac1 mRNA were synthesized and annealed with a T7 promoter DNA oligonucleotide (the primers to generate the different oligonucleotides are listed in Table   S6). In vitro transcription was then performed at 37 °C for 4 hr using the biotinylated rNTP mix (Roche Applied Science) and T7 RNA polymerase (Promega) following the method described before [3]. After DNase digestion of the reaction mixture, the biotin labeled RNA transcripts were gel purified, conjugated to streptavidin beads (Sigma), and then incubated with total protein extracts prepared from transfected HEK293T cells in 10 mM Tris-HCl, pH 7.

Local dendritic translation assay
Hippocampal neurons of DIV 12 were transfected with different siRNA oligos to deplete the endogenous TDP-43 or FMRP. 36-42 hr later the neurons were again transfected with phSYNmyr-dEGFP carrying Rac1 3'UTR sequence. 6-12 hr after the second transfection, the neurons were either subjected to live cell imaging to track the GFP reporter expression in dendrites for another 6 hr followed by cell fixation and immunofluorescence staining analysis, or directly fixed and the GFP fluorescence was analyzed along with the immunofluorescence staining. In some cases, the neurons were treated with 5 µM TTX or vehicle for 6 hr; fixed and the GFP fluoresce was monitored. Quantification of the GFP fluorescent intensities along the longer dendrites of the transfected neurons were analyzed by the Metamorf software (Molecular Devices, USA).

Pulse-labeling assay of translation
Hippocampal neurons at DIV 6 with or without depletion of TDP-43 or FMRP by RNAi were washed and preincubated with methionine-and cysteine-free DMEM (Invitrogen) for 30 min, and then labeled with 100 Ci/ml [ 35 S]methionine/cysteine ( 35 S-ProMix; GE Healthcare) for 1 hr. Subsequently, cells were washed and lysed in a lysis buffer used in immunoprecipitation described before. Total proteins from different samples were subjected to IP against anti-Rac1 (Milipore, Germany). Bead-bound proteins were extracted and resolved by 10% SDSpolyacrylamide gel electrophoresis followed by autoradiography. 10% of the total protein (Input) from each sample was also loaded on gel as a control.

Bioinformatics analysis
To identify the putative common mRNA targets of FMRP and TDP-43, multiple target datasets were  Table S2 for more details of these datasets). A unique gene list consisting of, 2717 mRNAs was derived from analysis of these 5 datasets by an in-  Table S2 for more details). Functional annotation of the mRNAs (Table S3) was carried out with the PANTHER classification system.    determined by RT-PCR analysis. Error bars represent SD from three technical repeats using the same biological samples. Percentage of Rac1 mRNA level was significantly different between control 4EGI-1 treated and TDP-si transfected/ 4EGI-1 treated neurons at 2 nd and 7 th -9 th fractions and between control 4EGI-1 treated and FMRP-si transfected/4EGI-1 treated neurons at 6 th -9 th fractions (p<0.01, pairwise t test).  control IgG, and analyzed by semi-quantitative (left) and quantitative (right) RT-PCR using primers specific for GluR1 mRNA and Map1b mRNA. GAPDH mRNA signals served as the negative control (shown in Fig 4c). RNA inputs were similar, as shown in Fig 7a(i). b. RNA-protein complexes isolated from the total cell extracts of primary hippocampal neurons transfected with different RNAi oligos were analyzed by quantitative RT-PCR using primers specific for Rac1 mRNA. Significant differences are indicated by ***p < 0.0001, **p<0.001 and *p<0.01 (Student's t test).