Activity dependent LoNA regulates translation by coordinating rRNA transcription and methylation

The ribosome is indispensable for precisely controlling the capacity of protein synthesis. However, how translational machinery is coordinated to meet the translational demands remains elusive. Here, we identify a nucleolar-specific lncRNA (LoNA), its 5′ portion binds and sequesters nucleolin to suppress rRNA transcription, and its snoRNA like 3′ end recruits and diminishes fibrillarin activity to reduce rRNA methylation. Activity-dependent decrease of LoNA leads to elevated rRNA and ribosome levels, an increased proportion of polysomes, mRNA polysome loading, and protein translation. In addition, transport of ribosomes to synapses is particularly promoted, resulting in increased levels of AMPA/NMDA receptor, enhanced synaptic plasticity, long-term potentiation and consolidated memory. Strikingly, hippocampal LoNA deficiency not only enhances long-term memory in WT mice, but also restores impaired memory function in APP/PS1 transgenic mice. Together, these findings reveal the multifaceted role of LoNA in modulating ribosome biogenesis to meet the translational demands of long-term memory.


Supplementary
, During the training phase of Morris water maze task, spatial learning was assessed in WT C57BL/6J mice as a function of training day with respect to the following parameters: (a) escape latency, (b) total path length to the target quadrant. Data were presented as the means ± s.m.e. (n=10). WT mice entered the target quadrant with significantly lower travel time and less travel distance, indicating a successful spatial learning. (c), Cellular fractionation. Representative Western blots using antibodies raised against the cell fraction-specific markers FBL (nucleolus), GAPDH (cytoplasm) and lamin A/C (nucleus) to indicate the purity of the nucleolar extracts. (d), lncRNA (validated in Fig1d, but not shown in Fig1f) levels in KCL activated primary neurons at different time point. (e), LoNA expression pattern across different brain regions. (f), LoNA levels were determined in different cell types of mouse brain, including primary cultured neurons and astrocytes, by qPCR. (g-h), Fear conditioning test. (g), the percent freezing for C57BL/6J mice during the tone test following trace fear conditioning. Baseline, trace, and intertribal interval (ITI) were included. (h), the percent freezing for mice during the context test. A 8 min trial of context condition testing was conducted 48 h after trace conditioning was implemented. (i), The object-context discrimination test. A 10 min trial of object and context condition testing was conducted. (j), LoNA expression levels were examined in hippocampal brain of mice subjected to the Morris water maze, Fear conditioning or Object-context discrimination training respectively, as determined by qPCR and normalized against U1 snRNA. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.
were determined in N2a cells treated with three independent antisense oligos (ASO), as measured by qPCR analysis. Data are normalized to U1 snRNA level. (h), LoNA expression was assessed in N2a cells with shRNA knockdown, as determined by Northern blot, U1 snRNA was included as an input control. (i), Levels of pre-rRNA (45S) and mature rRNAs (28S, 18S, 5.8S) in N2a cells treated with three independent LoNA antisense oligo or control sequence were determined by qPCR and normalized against U1 snRNA level. (j), LoNA expression was determined in primary neurons with LoNA knockdown by shRNA, as measured by qPCR. (k), rRNA (45S, 28S, 18S, 5.8S) levels were enhanced in LoNA knockdown neurons, as determined by qPCR. Data were normalized against U1 snRNA level. (l), Plot of LoNA copy numbers in N2a cells and primary cultured neurons. (m), LoNA knock down and control primary neurons were pulsed with 35 S-Met/Cys, cell lysates were resolved on SDS-PAGE. De novo protein synthesis was quantified as the mean ratio of 35 S incorporation relative to total protein (Coomassie). Protein synthesis rate in LoNA knockdown neurons was normalized against control neurons. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.
Supplementary Figure 3| (a), N2a cells were pulled down with biotinylated LoNA antisense DNA probes or control probes with sense DNA sequence, and LoNA level was determined by qPCR. (b), N2a cell lysates were subjected to immunoprecipitation with NCL antibody or control IgG, detected with NCL antibody. GAPDH was measured in total lysates to indicate equal input. (c), LoNA, but not U1, was specifically pulled down by anti-NCL antibody. NCL bound RNA level was determined by qPCR. (d), N2a cells were introduced with various amount of LoNA, cell lysates were subjected to immunoprecipitation with NCL antibody, total RNA was then subjected to qPCR analyses. (e), UV-CLIP assay was conducted with anti NCL antibody, NCL bound LoNA fragment was detected by qPCR with specific primer pair. U1 was included as a negative control. Data were normalized to control IgG. (f), N2a cells were transfected with WT or mutant LoNA respectively, total LoNA levels were determined by qPCR. (g), Pre-rRNA 45S level was determined by qPCR in N2a cells transfected with NCL or LoNA shRNA, 45S level was normalized to U1. (h), Nascent 45S rRNA levels weres measured by nuclear run-on analysis in N2a cells transfected with NCL or LoNA shRNA, 45S levels were normalized to U1. (i), NCL knockdown by shRNAs in N2a cells, shown by Western blot and densitometric analyses. (j), Representative immunofluorescence images of NCL (green) and DAPI (blue), in control or N2a cells with LoNA knockdown. Scale bar: 10 µm. (k), NCL mRNA levels were examined in N2a cells with LoNA overexpression or knockdown, by qPCR. Data were normalized to GAPDH and plotted against to their respective controls. (l), NCL levels were not altered by an increased amount of LoNA. N2a cells were introduced with various amount of LoNA, cell lysates were then subjected to immunoblotting with NCL antibody. (m), NCL levels were not altered by WT nor mutant LoNA. N2a cells were introduced with same amount of WT or mutant LoNA, cell lysates were then subjected to immunoblotting with NCL or UBF antibody. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.
Supplementary Figure 4| (a), N2a cell lysates were subjected to immunoprecipitation with FBL antibody or control IgG, followed by SDS-PAGE and immunoblotting detection with FBL antibody. A small amount of lysates were used as input, and GAPDH was included to indicate equal input. (b), UV-CLIP assay was conducted with anti FBL antibody, FBL bound LoNA fragment was detected by qPCR with specific primer pair. Data were normalized to control IgG. U1 snRNA was included as a negative control. (c), Binding affinities of FBL to M1 or M2 or M1+M2 LoNA were determined by pulldown of biotinylated LoNA. GAPDH was included as an input control. (d), Binding affinities of NCL to Mut or Del LoNA were determined by pulldown of biotinylated LoNA. GAPDH was included as an input control. (e), FBL levels were not altered by increased amounts of LoNA. N2a cells were introduced with various amount of LoNA, a plasmid without target sequence was included as a negative control. Cell lysates were then subjected to immunoblotting with FBL antibody. GAPDH was included to indicate equal input. (f), FBL levels in N2a were not affected by introduction of WT, Mut, or Del LoNA, as measured by Western blot. (g), FBL mRNA levels were examined in N2a cells with LoNA overexpression or knockdown, by qPCR. Data were normalized to GAPDH and plotted against to their respective controls. (h), Representative immunofluorescence images of FBL (red) and DAPI (blue) in N2a cells with LoNA deficiency. Scale bar: 10 µm. (i), N2a cells were transfected with WT, box C/D del or box C/D mut LoNA respectively, total LoNA levels were determined by qPCR. (j), FBL knockdown by shRNA in N2a cells, shown by Western blot and densitometric analyses, GAPDH was used to indicate equal input. FBL levels in knockdown cells were normalized against control shRNA. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.
Supplementary Figure 5| A schematic of RTL-P approach for detection of the presence of 2'-O-methylation in RNA. (a), RT reaction was conducted with RT primer at low (1M) or high (1mM) concentration of dNTP respectively. At high level of dNTP, the RT reaction produce long RT products (denoted in green color), at low level of dNTP, the RT reaction is impeded by the 2'-O-methyl groups, leading to a shorter RT products (denoted in red color). (b), PCR with different PCR primer pairs targeting upstream (Um) or downstream (Dm) of a specific methylation site. For RNAs with high level of methylation sites, the quantity of the RT primer/Um primer PCR products is less than those RNAs with low level of methylation sites in presence of low concentration of dNTP. This quantity is similar in presence of high concentration of dNTP. (c), Electrophoresis to indicate the size and intensity of PCR products. Figure 6| (a-c), The detection of 12 methylation sites distributed throughout the 18S and 28S rRNA by RTL-P. Six sites were determined for the 18S, and other six for the 28S. Total RNAs from N2a cells containing LoNA box C/D mut (a), or LoNA box C/D del (b), or FBL shRNA (c) were subjected to RT with low (1M) or high (1mM) concentration of dNTP respectively. cDNA was then amplified with primer pair corresponding to upstream (Um) or downstream (Dm) of a specific methylation site. (d-f), In RTL-P assay, data were presented as signal intensity ratio of amplification products at low dNTP (1M) over high dNTP (1mM) level. Methylation ratio in LoNA box C/D mut N2a cells were normalized to its control N2a cells (d), in LoNA box C/D del cells were normalized to its control N2a cells (e), and in FBL knockdown cells were normalized to cells containing control shRNA (f). Position of the nucleotide that is analyzed was indicated in x-axis. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.

Supplementary
Supplementary Figure 7| (a), Readings for each polysomal fractions from 3 independent ribosome profiling were averaged and plotted. (b), Rps6 levels were determined in each polysome fraction of LoNA deficient or control N2a cells to indicate successful fractionations, as measured by western blot. (c-d), LoNA was knocked down in vivo using an AAV vector (containing GFP element) in the hippocampus of C57BL/6J WT mice. (c), a representative image of LoNA knockdown in hippocampal brain. Scale bar: 100 µm. (d), LoNA level in injected hippocampal brain was determined by qPCR. (e), Representative images of TUNEL staining on AAV injected brain sections, GFP (green), apoptotic signal (red) and DAPI (blue). Dnase I treatment was included as a positive control. Scale bar: 20 µm. (f-i), Synaptosome fractions were purified from hippocampal brain of control, LoNA or mutant LoNA administered mice. (f-g), Levels of synaptic protein PSD95, snap25 and synaptophysin were determined by western blot and subsequent densitometric analysis. Equal amounts of brain tissues were used and GAPDH was measured as an input control. (h-i), Levels of AMPA receptor GluA2, NMDA receptor NR1, NR2A and NR2B were measured by western blot and densitometric analysis. Equal amounts of brain tissue were used and GAPDH was determined as an input control. (j), Levels of synaptophysin were determined by western blot in synaptosome isolated from WT hippocampal brain to indicate a successful isolation. Mitochondria marker prohibitin was included as a negative control. (k), Ribosomal protein levels were determined in total hippocampal brain lysates, by western blot. GAPDH level was determined as an input control. (l), Ratio of ribosomal proteins in synaptosomes fractions and total lysates, by densitometric analysis. (m-n), Levels of ribosomal protein Rps6, Rps3 and Rpl23 were determined by western blot (m) and densitometric analysis (n) in synaptosome fractions of LoNA or mutant LoNA administered mice. Equal amounts of brain tissue were used and GAPDH was measured as an input control. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.
Supplementary Figure 8| (a), LoNA or mutant LoNA neuronal overexpression in vivo using an AAV vector in the hippocampus of C57BL/6J WT mice. (b-e), Morris water maze behavioral test of LoNA or mutant LoNA hippocampal overexpression and control mice. Target quadrant entering time (b) and travel distance (c) were assessed during the training phase (n=10 for each group). Time to locate the hidden platform (d) and frequency crossing the target quadrant (e) were determined during the probe trial (n=10 for each group). (f), pre-and mature rRNA levels of LoNA hippocampal injected and control mice were analyzed by qPCR, data were normalized against U1 snRNA. (g-i), Pattern separation memory was assessed in an objectcontext discrimination test of WT mice administered with LoNA or mutant LoNA (g), LoNA shRNA (h), of APP/PS1 transgenic mice administered with LoNA shRNA (i). (j), Levels of RP11-517C16.2 in variety of human cells, as determined by qPCR. Data were normalized to GAPDH. Error bars, s.e.m.; *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA or two-tailed Student's t test.

Nucleoli isolation and High-Throughput Sequencing
N2a cell nucleoli were isolated as previously described The low quality sequenced reads (Q≤10) were removed and clean reads were aligned on Enseml (version mm10) and GENCODE (version 19) of the mouse genome, using TopHat.

Plasmids construction, ASO and transfection
All plasmids were constructed with restriction-enzyme digestion (NEB) and ligation methods (Vazyme). LoNA overexpression vector was generated by inserting its

Generation of lentivirus
Lentivirus plasmid was transfected into 293T cells together with packaging plasmid (pHR'8.2deltaR) and envelope plasmid (pCMV-VSV-G), virus containing medium was harvested 24 h and subjected to ultracentrifuge for virus precipitations. The virus was resuspended in PBS and viral titer was determined by qPCR. pre-rRNA promoter luciferase vector was introduced into these cells. Cells were lysed 24 h later and lysates were incubated with luciferase substrate. Data were presented as luciferase activities.

Quantification of RNA copy number per cell
DNA fragments corresponding to LoNA were amplified from mouse cDNA, and purified fragments were used to plot standard curves by qPCR. Total RNA was extracted from 1.0x10 5 cells, followed by cDNA synthesis. The copy numbers per cell in each cell line were calculated on the basis of cell numbers and the Ct value by using the standard curve.

Silver staining and Mass-Spectrometry
Silver staining was performed according to a previous published protocol 4 . RNA pulldown protein samples were resolved on electrophoresis gels, then were placed in fixing solution (50% ethanol, 12% acetic acid and 0.05% formalin) for 1 h and washed with 20% ethanol for 20 min. Gels were subsequently sensitized with 0.02% sodium thiosulfate and stained with cold 0.2% silver nitrate for 20 min. Silver stained gels were developed using a solution containing 6% sodium carbonate, 0.0004% sodium thiosulfate and 0.05% formalin. The reaction was terminated with 12% acetic acid. Bands were cut, digested, extracted from gels and subjected to massspectrometry analyses (PTM Biolabs).

Puromycin Incorporation Assay (SUnSET assay)
SUnSET was performed according to previously published protocol 6  and protein concentration was determined by BCA protein assay kit (Pierce). Equal amounts of lysate were subjected to immunoprecipitation with anti-synaptophysin or anti-PSD95 antibodies, followed by immunoblotting with anti-puromycin antibody.

S Incorporation Assays
Cells were washed twice with methionine/cysteine-free media (Life technologies) and then incubated with Met/Cys-free media containing 10% fetal bovine serum, and 100 Ci ml -1 EasyTag EXPRESS 35 S Protein Labeling Mix for 30 min. Cells were washed twice with ice cold PBS containing 100 g mL -1 cycloheximide, then lysed with RIPA buffer. Extracted proteins were subjected to protein concentration determination by BCA kit (Pierce) and separation on 10% SDS-PAGE. Gels were dried for two hours at 80C and exposed 24 h on a phosphorimager screen at -80°C freezer. Signal was quantified by ImageJ software and normalized to total protein determined by Coomassie staining.

Polysomal profiling assays
N2a cells were pre-treated with cycloheximide (Sigma) for 15 min at 37°C, and

Chromatin Immunoprecipitation (ChIP)
ChIP assays were performed as described previously 7    PCR signal intensities were analyzed using ImageJ software. The methylation ratio of each site was determined from density of PCR bands obtained using high and low dNTP concentrations.

Generation of Adeno-associated virus and hippocampal injection
Adeno-associated virus (AAV) was produced according the protocol described by

Behavioral tests
The Morris water maze was performed according to a previously published protocol 10 .
Briefly, mice (n=10 each group) were trained in the Morris water task to navigate a direct path to the hidden escape platform when started from semi-random locations around the perimeter of the tank. Four training trials were given to each mouse per day for 5 consecutive days during the learning phase. After total 20 training trials, each mouse was given a 30 second probe trial with the platform removed. Time spent in the target quadrant when compared with other quadrants and how often the mice entering the target quadrant was determined. Mice were sacrificed 24 h after training and tissues were saved for RNA analysis. This time window was kept consistent between experiments.
The object-context discrimination test was performed as described 11 . Briefly, mice were placed in an open chamber with a specific floor pattern and two identical objects, followed by 10 min exploration and an inter-trail interval. Mice then were placed in a second chamber with different floor pattern and the objects unique from the objects in the first trial. Mice finally were tested for 10 min in a chamber consisting of a floor pattern from either trial one or trial two, one object from trial one, and one object from trial two. The time mice spent in exploring the object in novel context was compared to the different object in the old context. Mice were sacrificed 24 h after training and tissues were saved for RNA analysis. This time window was kept consistent between experiments.
The fear conditioning test was carried out as described previously 12 . Briefly, on day 1, mice were allowed to explore the chamber for 3 min and then exposed to 20 sec tone

Synaptosomes purification
Synaptosomes (n=7 each group) were isolated using discontinuous Percoll (GE Healthcare) gradients as described previously 13  and subjected to centrifugation at 31,000 g. Fractions 3-5 were collected, diluted with sucrose buffer and centrifuged at 29,000 g. Pellets were synaptosomes and saved for further analyses.

Electrophysiology
C57BL/6J mice were administered with LoNA or LoNA shRNA in hippocampal brain through adeno-associated virus delivering system (n=10 for each group). Three weeks after the injection, mice were anaesthetized and placed in the stereotaxic apparatus. The recording electrode was placed at the cell body layer of CA1 and the stimulating electrode at the CA3 according to stereotaxic parameters. The long-term potentiation (LTP) was then evoked and recorded. The parameters for theta burst stimulus to induce LTP were five trains of 20 pulses at 200 Hz stimulation for 1 s with each train separated by a 1 min interval.

Golgi Staining and Dendritic Spine Analysis
Golgi staining was performed as described previously 14 . Briefly, freshly dissected brains were immersed in Golgi-Cox solution (5% w/v K2Cr2O7, 5% w/v HgCl2, 5% K2CrO4) for 2 weeks at room temperature and then transferred to Tissue Protection solution (0.3% w/v NaH2PO4, 1% w/v Na2HPO4, 1.8% w/v NaCl in 100 mM PBS) for 24 h at RT. Brain sections were cut at 150 µm using Leica vt1200 microtome, then mounted onto slides for visualizations. The pyramidal neurons in the CA1 region of the hippocampus were visualized and analyzed. Five neurons were randomly selected per brain region from each mouse, and at least two segments were chosen randomly per neuron. Dendritic spines density was measured in a blinded manner.

Western Blot and Densitometry Analysis
Cells or brain tissues were lysed on ice in RIPA buffer (50 mM Tris pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS). Protein concentrations were determined using a BCA Protein Assay kit (Pierce). Equal amounts of protein were used for SDS-PAGE. The immunoreactive bands were visualized by enhanced chemiluminescence (Thermo Scientific) and detected by ChemiScope (CLiNX). For densitometric analyses, immunoreactive bands were quantified using Image J software.

RNA extraction, Reverse transcription and Quantitative PCR
Total RNA was isolated from cells or tissues using Trizol (Invitrogen) and subjected to DNase I digestion to remove genomic DNA contamination. Total RNA was dissolved in nuclease-free water and stored at -80 °C for further use. Reverse transcription was performed using SuperScript II RNase H-reverse transcriptase (Invitrogen) following the manufacture's instruction. Quantitative PCR was performed with SYBR Green qPCR Master Mix (Roche) on LightCycler 96 system (Roche) according to standard procedures. The real-time value for each sample was averaged and compared using the CT method, where the amount of target RNA (2−ΔΔCT) was normalized to an endogenous reference (ΔCT) and related to the amount of target gene in tissues or cells, which was set as the calibrator at 1.0. QPCR detection primers are listed in Supplementary Table1.

Primary Neuronal Culture
Mouse primary neurons were prepared from embryonic day 17 of C57BL/6J mice as previously described 15 . Neurons were seeded at a density of 500,000 cells per 6cm dish or 350,000 per 6 well plate. Cultures were grown in Neurobasal medium supplemented with B27 (Invitrogen).

Confocal Microscopy
Fluorescence signals were captured with HRm CCD camera mounted on LSM710 confocal microscope system (Zeiss), with ZEN software.

Statistical Analysis
All quantified data represent an average of at least triplicate samples. Statistical significance was determined by Student's t-test or two-way ANOVA in GraphPad Prism 5.0. P < 0.05 was considered significant (indicated by an asterisk in the figures), P < 0.01 (indicated by two asterisks in the figures), P < 0.001 (indicated by three asterisks in the figures).