Supplementary Material and Methods

Supplementary Material and Methods Synaptosomes preparation and RT-PCR analysis. Synaptoneurosome fractions were prepared as previously described in 1. Briefly, rat total brain was homogenized in ice-cold buffer (50 mm HEPES ph 7.4, 100 mm NaCl, and 3 mm KAc, 1 mm MgSO 4, RNAse inhibitor 40u/ml) by a glass homogenizer and centrifuged at 2000 x g for 2 min. Supernatants were passed through two 100 µm nylon mesh filters, followed by a 5 µm pore filter (Millipore). The filtrate was then centrifuged at 1000 x g for 10 min and then was gently resuspended with same buffer at a protein concentration of 2 mg/ml. The quality of each preparation was checked by western blot analysis for typically enriched synaptic proteins such as PSD-95 and α CaMKII. RNA was isolated from total brain or synaptoneurosome using Trizol reagent (Invitrogen) according to the manufacturer s instructions. We used 1 2 µg of total RNA in reverse transcription reactions. Polysomal analysis and RT-PCR. Cytoplasmic brain extracts were separated on a 5-65% sucrose gardient. Each gradient was collected in 10 fractions, checking the absorbance at 254 nm and the rrna profile. Semi-quantitative radioactive RT-PCR analysis with primers for PSD-95, β Actin, Arc and L22 RNAs (see table below) was performed for each fraction to quantify the percentage of messenger on polysomes (PMP). The PMP value of PSD-95, β Actin and Arc mrnas in WT and FMR1 KO was calculated by comparing the radioactivity of the first 5 fractions of the gradient (polysomes = actively translated mrna) to radioactivity of the total 10 fractions (polysomes + mrnps) after normalization for L22 RNA (a synthetic mrna transcript added in equal amount to each fraction while the gradient was collected). Plasmid construction (fragm 1-5) and Mutagenesis The different plasmids were constructed as described below. Fragment 1 (nt 1-153) of the UTR and was isolated using s-oligo 5'-tgaTTC CTG CCC TGG CTT GG- and as-oligo 5'-GTC CTC CCT CCA ATG GGC TC-. Fragment 2 (nt 154-307) and was isolated using s-oligo 5'-tgA CTC CTC TCT GCA TGT ATC CCT G- and as-oligo 5'-CCC CAG TGG GTC TGT GTG TG-. Fragment 3 (nt 308-460) and was isolated using s-oligo 5'-taG CCT CTG CCC TCC CCA TT- and as-oligo 5'-CGG CGT GGG GAG TTA TGA TGG G-. Fragment 4 (nt 461-613) and was isolated using s- oligo 5'-tga TTT GAG TTC TCC TTT ATT TTC TCC- and as-oligo 5'-CCA CAG TTA GAC CTT CCA CT-. Fragment 5 (nt 593-835) and was isolated using s-oligo 5'-tAG TGG AAG GTC TAA CTG TGG CT- and as-oligo 5'-CAA GTG TCT GTC TCT TCC TTT CAC-. The construct containing the G-quartet was mutagenized as follows. In the first step half of fragment 1

5 was PCR isolated from the pt/a-fragment 5 plasmid using the s-oligo 5'-TAG TGG AAG GTC TAA CTG TGG CT- and the as-oligo 5'-CAT gtc Tag ACC CCC gct TCC CAA AAA AAT AAA ATC- containing the 5' half of the randomized G-quartet (underlined). The as-oligo also encoded a unique XbaI site. This PCR fragment was T/A cloned into the Promega Easy T/A cloning plasmid (pt/a-gqsub). A second PCR reaction was performed on the pt/a-fragment 5 cdna using the s-oligo 5'-CAT gtc Tag AgA TgA ggg gag Tgg gga ATg Tgg ga- encoding the half of the randomized G-quartet (underlined) and the as-oligo 5'-ATT CAC Tag TgA TAC AAg TgT CTg TCT CTT CC- encompassing the SpeI site within the T/A vector sequence. This PCR product was cloned into pt/a-gqsub from XbaI to SpeI to create pgqmut. pt/a-fragment 1-Gquartet was created using oligos to the putative G-quartet of PSD-95 (5'-cat ggg GGA AAA GGG GAG GGA TGG GTC TGG GGA GTG Gcc gc- and 5'-ggC CAC TCC CCA GAC CCA TCC CTC CCC TTT TCC Cc-). The oligonucleotides were annealed, kinase treated and then ligated into pt/a-fragment 1 between NcoI and SacII sites. The constructs containing the mutagenized U-rich regions were obtained using the QuickChange sitedirected mutagenesis kit (Stratagene). The first (I) U-rich mutant was produced using the s-oligo 5 - GGA AGG TCT AAC TGT GGG CGT GGT CCT GCC CTG GGA TT- and the as-oligo 5 - AA TCC CAG GGC AGG ACC ACG CCC ACA GTT AGA CCT TCC-. The second (II) U- rich mutagenesis was performed using the s-oligo 5 - CCT GCC CTG GGA TTT GCC GGG ATC CTG TGC CAC GGC ATG CGT TTG GGA AAA GGG GAG GGA TGG G- and the as-oligo 5 -C CCA TCC CTC CCC TTT TCC CAA ACG CAT GCC GTG GCA CAG* GAT CCC GGC AAA TCC CAG GGC AGG-. The third (III) U-rich region was been mutagenized using the s- oligo 5 -GCC GCC TTC TGC AAT CCG GCG ATG GAA GCT TTT GAG AGA GTG AAA GG- and the as-oligo 5 -CC TTT CAC TCT CTC AAA AGC TTC CAT CGC CGG ATT GCA GAA GGC GGC-. Oligonucleotides used in this study: 5 -GGC TTC ATT CCC AGC AAA CG- PSD-95-UTR 5 -CAT CAA GGA TGC AGT GCT TC- PSD-95 CDS 5'-CAA AAC TCC AAT GAA GTC AG- 5'-GGT CTT CAA TGA CAC GTT TCA- 2

5 -GCG GCC CGG AAA AGC GCG CCC TGA G- Histone H3 5 -GTA TCA CCC ATC CCT TCT GCA TAT TA- 5'-CGT GGG CAC GTC AGT CAC G - L22 5'-CTC GAG GTC GAC GGT ATC GA- 5'-AGC AAG AGA GGC ATC CTG ACCβ-Actin 5'-CTC ATT GCC GAT AGT GAT GAC C- 5'-AAG CTG GAG AAC AAC TTG GAC GG- Mouse Arc 5'- CCC CCA AGA CTG ATA TTG CTG AG- 5'-AAG CTG GAG AAC AAC TTG GAC GG- Rat Arc 5'-TGG TAT TGC TGA GCC CCA ACT GAC- 5'-GGC AAG ATG GGG TAT AGA GA- MAP1B 5'-CCC ACC TGC TTT GGT ATT TG- 5'-GGC AAT GAC CGC TAT GAG GGC- GluR1 5'-TCC ACA GTC AGG AAG GCA GCC- GlyRα 5'-AGT TTC GGT TCC ATC GCT GAG- 5'-TGT AGA GGA CAT TGC CAT TCC- α-tubulin 5 -TTT TCC ACA GCT TTG GTG GGG G- 5 -TCT TGA TGG TGG CAA TGG CAG- 5 -ACC CTG GCC TGG TCC TTC AAT G- α-camkii 5 -AGC CAT CCT CAC CAC TAT GCTG G- GFAP 5 -CTC CTT GTC TCG AAT GAC TCC- 5 -CCT TAA GAA CTG GAT CTC CTC- 5 -TAA TAC GAC TCA CTA TAG GGA AAA PSD-95 I G-rich GGG GAG GGA TGG 5 -CCA ACC CCT GAC CCC TTG- 5 -TAA TAC GAC TCA CTA TAG GGT GTC PSD-95 II G-rich CGG GAG CCA GGG- 5 -GCA GAA GGC GGC AGC ATT T- 3

SAPAP4 5 - GGA TAC CTA CCT GGA CAG TC- 5 - GGA GAC TCC TCT CAG ATG AC - 5 -GGC AGA GAT ATG GCA AAC AGG- RhoA 5 -CCA GAG TTC TTG CAG TTG ACA- 5 -CCA ATG GAA GCA GCT GGC TT- EF-1A 5 - CCG TTC TTC CAC CAC TGA TTA- 5 - GGT GGC TGA GGA GAT TCA AG- APP 5 -TCA CGG TTG CTA TGA CAA CGC- 5 - GGC TAT AAT GGG TAT GGA GG- HnRNPA2 5 -CCA CTT CTT CAT TAG TTA CCT- 5 -GGG AGT TTG TGC GAC AGT AC- G3BP 5 -CCA GAA TCA TCA GGC ACC AC- MBP 5 -GGG TTC CTG TCA CTG AAT CG- 5 -GGT GCT TCT GTC CAG CCA TA- 5 -CCT GCT CGA TGG CAA GAA CG- VDAC1 5 - GCA CGA TGG GAA CAG GGT GT- Shank 1 5'- GGA TGT GGA CGA CGG CGA GT- 5'- GCT ACT CCG ACT GTC CAC CT- Homer 1a 5'- GGA GAA GAT GGA GCT GAC CA- 5'- GGC CTG TGG TAA AGC TTT CC- Luciferase 5 -GTG AAG TTC GTC GTC CAA CAT T- 5'-AAC GTC AGG TTT ACC ACC TTT TAC T- TaqMan probe FAM - CCT CGT GAA ATC CCG 4

Figures Supplementary Figure 1. G-quartet consensus and similarities to PSD-95 mrna. Upper sequence: putative PSD-95 G-quartet sequence on the mouse PSD-95 mrna (UTR) extending from nt 668 to nt 689. Lower sequence: Consensus of the G-quartet structure according to 4. Nucleotides that do not fit the consensus sequence are listed in gray. Supplementary Figure 2. PSD-95 mrna localizes in synaptoneurosomes from total brain. To verify that PSD-95 mrna was localized at synapses we assessed its presence in synaptoneurosomes (according to 1 ). We find that the PSD-95 mrna is localized at synapses with a remarkable dendrite/soma enrichment ratio, similar to α-camkii and Arc mrnas, two well-known synaptically localized mrnas. RT-PCR for GFAP, β Actin, Histone, α CaMKII, Arc and PSD-95 mrnas was performed on total brain RNAs (T) or synaptoneurosomes RNAs (S). The relative percentage of each mrna in total brain (gray) and in synaptoneurosomes (white) was reported in the histogram. GFAP, used as a glial specific negative control and two mrnas present purely in neuronal cell bodies (β Actin and Histone H3) are almost completely absent in the synaptoneurosome preparation. Supplementary Figure 3. PSD-95 mrna is specifically detected by in situ hybridization. (a) In situ hybridization performed using a sense riboprobe specific for PSD-95, α-tubulin and α CaMKII mrnas (red panels) combined with an immunofluorescence for FMRP (green) on hippocampal cultures (DIV 10). (b) Radioactive in situ hybridization performed using a sense riboprobe (coding region) specific for PSD-95 mrna on brain sections from wild type and FMR1 KO mice. Supplementary Figure 4. PSD-95 mrna is dendritically localized in vivo. Inverted images of PSD- 95, α Tubulin and α CaMKII mrnas in the hippocampal formation. White arrows indicate the apical dendrites of the stratum lacunosum-moleculare. Supplementary Figure 5. FMRP regulates the stability of PSD-95 mrna in hippocampal cells. RNA was isolated at the indicated times after Actinomycin D (10 µg/ml) application to hippocampal neurons from WT or FMR1 KO mice. The stability of PSD-95 mrna in WT or FMR1 KO hippocampal 5

cells was measured by real-time fluorescent quantitative RT-PCR, normalizing the signal for the β Actin mrna. Supplementary Figure 6. FMRP doesn t regulate the stability of PSD-95 mrna in cortical cells. RNA was isolated at the indicated times after Actinomycin D (10 µg/ml) application to cortical neurons (DIV 10) from WT or FMR1 KO mice and the relative levels of PSD-95 and β Actin mrnas were examined by semi-quantitative RT-PCR. Supplementary Figure 7. Viability of WT and FMR1 KO hippocampal neurons during Actinomycin D treatments. Photographs were taken from hippocampal neurons (WT and FMR1 KO mice) during Actinomycin D treatments before RNA extraction. Scale bar: 10 µm. Supplementary Figure 8. HuD and HuR protein levels in mouse hippocampus, cerebellum and cortex. Western blot was performed with 20 g of hippocampus, cerebellum or cortex cytoplasmic extracts using a specific antibody for the HuD protein (a) or HuR protein (b) purchased from Santa Cruz and used 1:200. The histograms represent the quantification of three independent experiments (s.e.m.). *, p< 0,05 by Student s t test. References 1. Shin, C. Y., Kundel, M. & Wells, D. G. Rapid, activity-induced increase in tissue plasminogen activator is mediated by metabotropic glutamate receptor-dependent mrna translation. J Neurosci 24, 9425-33 (2004). 2. Zalfa, F. et al. The fragile X syndrome protein FMRP associates with BC1 RNA and regulates the translation of specific mrnas at synapses. Cell 112, 317-27 (2003). 3. Denman, R. B. Deja vu all over again: FMRP binds U-rich target mrnas. Biochem Biophys Res Commun 310, 1-7 (2003). 4. Zhang, Y. Q. et al. Drosophila fragile X-related gene regulates the MAP1B homolog Futsch to control synaptic structure and function. Cell 107, 591-603 (2001). 5. Brown, V. et al. Microarray identification of FMRP-associated brain mrnas and altered mrna translational profiles in fragile X syndrome. Cell 107, 477-87 (2001). 6. Darnell, J. C. et al. Fragile X mental retardation protein targets G quartet mrnas important for neuronal function. Cell 107, 489-99 (2001). 7. Lu, R. et al. The fragile X protein controls microtubule-associated protein 1B translation and microtubule stability in brain neuron development. Proc. Natl. Acad. Sci. USA 101, 15201-6 (2004). 8. Zhong, N., Ju, W., Nelson, D., Dobkin, C. & Brown, W. T. Reduced mrna for G3BP in fragile X cells: evidence of FMR1 gene regulation. Am J Med Genet 84, 268-71 (1999). 9. Westmark, C. J. & Malter, J. S. FMRP Mediates mglur5-dependent Translation of Amyloid Precursor Protein. PLoS Biol 5, e52 (2007). 10. Chen, L., Yun, S. W., Seto, J., Liu, W. & Toth, M. The fragile X mental retardation protein binds and regulates a novel class of mrnas containing U rich target sequences. Neuroscience 120, 1005-17 (2003). 6

11. Sung, Y. J. et al. The fragile X mental retardation protein FMRP binds elongation factor 1A mrna and negatively regulates its translation in vivo. J Biol Chem 278, 15669-78 (2003). 12. Brown, V. et al. Purified recombinant Fmrp exhibits selective RNA binding as an intrinsic property of the fragile X mental retardation protein. J Biol Chem 273, 15521-7 (1998). 13. Wang, H. et al. Developmentally-programmed FMRP expression in oligodendrocytes: a potential role of FMRP in regulating translation in oligodendroglia progenitors. Hum Mol Genet 13, 79-89 (2004). 7