FTO-dependent m6A RNA dysregulation underlies memory deficits induced by early-life stress

12 Cognitive functions in adults are mainly attributed to experience -dependent plasticity. 13 Nonetheless, the developmental encoding of memory deficits is still inadequately addressed. 14 Here, we demonstrate that early -life stress (ELS) reprograms the hippocampal 15 epitranscriptome by enhancing N6-methyladenosine (m 6A) deposits during early 16 development leading to memory deficit in adulthood. We observed a shift toward 17 hypermethylation of transcripts including coding and non-coding RNAs (lncRNAs) following 18 maternal separation. We also observed that these transcripts encoding proteins necessary 19 for translational regulation , ribosome biogenesis and mitochondrial function . This 20 epitranscriptomic change is driven by ELS-induced downregulation of the m 6A demethylase 21 FTO (Fat mass and obesity -associated protein). We observe that the overexpression of FTO 22 in young adult mice selectively rescues memory deficits without ameliorating elevated 23 anxiety. Further , the knockdown of FTO in primary hippocampal neuron, mimicking ELS -24 induced reduction of its expression, leads to reduced translation as detected by puromycin 25 labelling. Taken together, our study demonstrated previously uncharacterized mechanism of 26 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 2 ELS-induced epitranscriptomic change linked with memory deficit via the regulation of 27 protein synthesis. 28 29

30 m6A-RNA methylation (N6 -methyladenosine) is the most abundant RNA modification in 31 eukaryotes that fine-tunes an additional layer of RNA regulation (Akhtar et al., 2021; Sendinc 32 & Shi, 2023); dynamically regulated by a set of functionally antagonistic enzymes called 33 "writers" (methyltransferases) and "erasers" (demethylases) (Loedige et al., 2023). The 34 writer complex comprises METTL3 (methyltransferase -like 3), METTL14 35 (Methyltransferase-like 14), and WTAP (Wilms tumor 1 associated protein) that catalyse the 36 barcoding of RNA with m6A, while erasers like FTO (Fat mass and obesity-associated protein) 37 and Alkbh5 (AlkB homolog 5) remove the modification from RNAs (Jia et al., 2011; Ping et 38 al., 2014; P . Wang et al., 2016; Zheng et al., 2013). Another group of proteins called "readers," 39 e.g., YTHDF1-3 (YTH N6 -methyladenosine RNA binding protein 1, 2, 3), recognise the m 6A 40 topography on the transcripts and channel them to specific signalling pathways (H. Shi et al., 41 2018; Zaccara & Jaffrey, 2020). While the role of m 6A in RNA stability and translation has 42 been widely studied (Zhou et al. 2018, Shi et al., 2018) , its contribution to the regulation of 43 early-life stress (ELS) remains elusive. 44 Previous studies have converged on the effects of ELS on long -term cognitive health, 45 particularly in the form of heightened anxiety and memory deficits (Levin & Liu, 2021; Lin et 46 al., 2016; Pattwell & Bath, 2017; A. Wang et al., 2020). The hippocampus, with its dual role in 47 regulating emotion and spatial memory, has also been shown to be affected by ELS (Bertagna 48 et al., 2021; Brosens et al., 2023; Hanson et al., 2015; Lee et al., 2019). Still, the molecular 49 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 3 underpinnings of the hippocampal circuitry modulating such behavioural patterns remain 50 poorly understood. Emerging evidence also suggests that RNA modifications, particularly 51 m6A, play a critical role in regulating RNA stability and expression under stress conditions 52 (Anders et al., 2018; Kisliouk et al., 2020). Here, we investigated how m6A-RNA modification 53 influences hippocampal circuitry in the context of ELS using the long -standing paradigm of 54 maternal separation (MS) (Sood et al., 2018). 55 Our study shows that maternal separation causes an increase in m 6A-RNA deposition along 56 the transcripts in the hippocampus of mice. The impact of change in m6A topography extends 57 beyond protein-coding RNAs to include long non-coding RNAs (lncRNAs). LncRNAs are more 58 than 200 nucleotides long and typically do not contain open reading frames for encoding 59 proteins (Fatica & Bozzoni, 2014; Ma et al., 2013). This subset of RNAs has also been 60 increasingly evidenced to have a role in translation and synaptic plasticity (Banerjee et al., 61 2024; Mattick et al., 2023; Mercer et al., 2008, 2009; C. Shi et al., 2017; A. Wang et al., 2017) 62 and for their regulatory roles in the brain (Briggs et al., 2015; Ernst & Morton, 2013; 63 Samaddar et al., 2023). The hypermethylation of hippocampal transcripts coincides with a 64 significant downregulation of the RNA demethylase FTO, suggesting that the hippocampus is 65 particularly vulnerable to m6A -mediated epigenetic shifts. Reduced FTO expression is 66 directly correlated to the spatial memory deficits observed after ELS, and its overexpression 67 rescues the phenotype. The downregulation of FTO in hippocampal neurons leads to a 68 decrease in global protein synthesis. Collectively, our study reveals a molecular framework 69 involving RNA methylation that may explain the cognitive deficits observed in adults who 70 have been exposed to ELS. 71 72 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 4

73 Ethical clearance for animal maintenance 74 The animals used in this study were housed following the guidelines established by the Institutional 75 Animal Ethics Committee of the National Brain Research Centre (NBRC -IAEC) in India. The study 76 protocols received approval under reference numbers NBRC/IAEC/2015/172 and 77 NBRC/IAEC/2020/178. NBRC-IAEC is accredited by the Committee for the Purpose of Control and 78 Supervision of Experiments on Animals (CPCSEA) under the Ministry of Fisheries, Animal Husbandry, 79 and Dairying, Government of India, with registration number 464/GO/ReBi-S/Re-L/01/CPCSEA. 80 C57BL/6 and CD1 mice strains were housed individually in standard cages under a 12 -hour 81 light/dark cycle (6 a.m. to 6 p.m.). The mice had ad libitum access to food and water . Environmental 82 conditions were maintained at a constant temperature of 26 ± 1°C and a humidity level of 40 ± 5%. 83 Maternal Separation 84 To induce ELS, we explored the maternal separation paradigm as a potent tool (Alves et al., 2022). 85 C57BL/6 pregnant dams were subject to random allocation, dividing them into either a control cohort 86 or a maternally separated group. A litter size of 6-8 pups was maintained. The separation started from 87 postnatal day 2 (P2) and extended until P14, with a daily separation period of 3 hours, from 0900 to 88 1200. Before each separation session, the mother was relocated to a fresh cage, while the pups were 89 transferred to a different room. The pups were gently removed from their original cage individually 90 and placed on soft cotton bedding within a glass beaker positioned on a heating pad, maintaining a 91 consistent temperature range of 36°C to 38°C. The inner wall of the glass beaker was lined with paper 92 to prevent singeing. Upon completion of the separation period, the pups were carefully returned to 93 their original cage, followed by the reintroduction of the mother . Subsequently, the animals were kept 94 under observation to ensure the mother's engagement in tending and nursing the pups. The pregnant 95 dams comprising the control group experienced minimal disruption, with cage changes occurring 96 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 5 once every three days, consistent across all experimental conditions. At P21, male pups were selected 97 for the harvesting of hippocampal tissue, facilitating further investigation and analysis. 98 2.3. Immunoprecipitation and dot-blot 99 To perform m 6A RNA immunoprecipitation (MeRIP) assay, hippocampi were isolated from P21 100 C57BL/6 male mice and collected in an immunoprecipitation buffer containing 20 mM Tris-HCl (pH 101 7.5), 100 mM KCl, 5 mM MgCl2, 0.5% NP-40, 1 mM DTT , 0.5 µg/µl heparin, protease inhibitor (Sigma; 102 1 µl/ml), phosphatase inhibitor (Sigma; 1 µl/ml), and RNase inhibitor (SuperaseIn; 100 U/ml). The 103 hippocampi were homogenised into a single -cell suspension. Cell debris was removed by 104 centrifugation at 10,000g for 5 minutes at 4°C, and the supernatant was collected. 10% of the total 105 crude lysate was kept aside as total input. Simultaneously, agarose beads were prepared by washing 106 three times with wash buffer 1 (20 mM Tris -HCl, pH 7.5, 100 mM KCl, 5 mM MgCl2, 0.5% NP -40, 0.5 107 µg/µl heparin, and RNase inhibitor) at 10,000g for 5 minutes at 4°C. The beads were then incubated 108 in 2.5% bovine serum albumin (BSA), prepared in wash buffer 1, for 1 hour at 4°C, 50 rpm. After 109 incubation, the beads were precipitated by centrifugation at 10,000g for 5 minutes at 4°C, and the 110 BSA solution was decanted. The collected supernatant was added to the BSA-incubated beads for pre-111 cleaning, incubating for 1 hour at 4°C, 50 rpm. The beads were then precipitated again at 10,000g for 112 5 minutes at 4°C, and the supernatant was collected and divided into two equal parts. Rabbit m6A 113 antibody was added to one half, and rabbit IgG was added to the other half, followed by incubation 114 for 4 hours at 4°C with gentle shaking. The RNA -antibody mixture was then added to agarose beads 115 pre-equilibrated in wash buffer 1 (by washing three times at 10,000g for 5 minutes at 4°C) and 116 incubated for 2 hours at 4°C, 50 rpm to allow the antibody -RNA complexes to bind to the beads. 117 Following incubation, the beads were washed twice with a high-salt wash buffer (20 mM Tris-HCl, pH 118 7.5, 150 mM KCl, 5 mM MgCl2, 0.5% NP-40, and RNase inhibitor) and then twice with a normal wash 119 buffer (20 mM Tris -HCl, pH 7.5, 100 mM KCl, 5 mM MgCl2, 0.5% NP -40, and RNase inhibitor) to 120 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 6 remove non-specifically bound RNA. 1 µl of bead was used for dot-blot and the rest of the beads were 121 stored at -80 ̊C. 122 1 µl of the sample from the total input, m6A-IP , and IgG-IP was dotted onto a nitrocellulose membrane. 123 The membrane was incubated in a blocking reagent (5% BSA) for 2 hours, followed by overnight 124 incubation at 4°C in a 1:1000 dilution of rabbit -m6A primary antibody. Afterwards, the membrane 125 was washed four times for 10 minutes each in 1X TBST , then incubated with a 1:1000 dilution of goat 126 anti-rabbit secondary antibody for 2 hours at room temperature. The membrane was washed again 127 four times for 10 minutes each in 1X TBST . Finally, the membrane was developed using ECL and 128 imaged with a Uvitech imager . 129 Protein Estimation and Immunoblot 130 Hippocampi were isolated from P21 C57BL/6 mice and lysed in RIPA buffer (25 mM Tris-HCl, pH 7.4, 131 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with protease (Sigma; 132 1 µl/ml) and phosphatase (Sigma; 1 µl/ml) inhibitors. The lysate was centrifuged at 12,000 x g for 10 133 minutes at 4°C to pellet the debris. The supernatant was collected and the protein concentration was 134 determined using BCA assay. First, a series of protein standards with known concentrations was 135 prepared by diluting a stock solution of bovine serum albumin (BSA). The BCA working reagent was 136 prepared by mixing reagents A and B in a 50:1 ratio, as per the manufacturer's instructions. 137 Subsequently, 200 µL of the BCA working reagent was added to each well containing the standards 138 and samples. The plate was then incubated at 37°C for 30 minutes to allow for the colourimetric 139 reaction to develop. After incubation, the absorbance of each well was measured at 562 nm using a 140 microplate reader . The protein concentrations of the samples were determined by comparing their 141 absorbance values to the standard curve generated from the BSA standards. 142 The protein samples with 4X Laemmli buffer (250 mM Tris-HCl, pH 6.8, 8% SDS, 40% glycerol, 0.02% 143 bromophenol blue, 20% β-mercaptoethanol) were heated at 95 ̊ C for 5 minutes. Equal amounts of 144 protein were loaded onto SDS-PAGE gel and were run in running buffer (25 mM Tris, 192 mM glycine, 145 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 7 0.1% SDS). 5 µg protein was loaded for Tubulin β1, 30 µg of protein was loaded for FTO, and 30 µg of 146 protein was loaded for METTL3. The proteins from the gel were transferred to a PVDF membrane 147 using a transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol) at 65 V for 90 minutes. After 148 transfer , the membrane was blocked in 5% BSA (Bovine Serum Albumin) in TBST (20 mM Tris -HCl, 149 pH 7.6, 150 mM NaCl, 0.1% Tween-20) for 1 hour at room temperature. The membrane was incubated 150 with primary antibody against Tubulin β1 (Sigma; Anti-mouse; 1:5000), FTO (Abcam; Anti -mouse; 151 1:1000), METTL3 (Abcam; Anti -rabbit; 1:1000), diluted in blocking buffer , overnight at 4°C. The 152 membrane was washed three times, for 10 minutes each, with TBST , then incubated with HRP -153 conjugated goat anti-mouse secondary antibody against Tubulin β1 (1:10000) and FTO (1:5000) and 154 with HRP-conjugated goat anti -rabbit secondary antibody against METTL3 (1:5000), diluted in the 155 blocking buffer for 2 hours at room temperature. The membrane was washed again for three times, 156 for 10 minutes each, with TBST . For FTO overexpression and knockdown conformations, GAPDH was 157 used as an internal control with a concentration of 1:10000 for primary (Sigma; Anti -Rabbit) and 158 1:10000 for secondary (HRP -conjugated goat anti -rabbit secondary antibody). Finally, the protein 159 bands were detected using an enhanced chemiluminescence (ECL) substrate, and the signal was 160 captured using an imaging system (Fig. S8-S13). 161 qRT-PCR and cDNA preparation 162 P21 C57BL/6 male mice were sacrificed, and their hippocampi were harvested in 600 µl of ice -cold 163 TRIzol reagent. The tissue was triturated with an 18 mm -gauge syringe to create a single -cell 164 suspension. The mixture was centrifuged at 1000 x g for 5 minutes at 4°C to pellet debris. The 165 supernatant was collected, and 300 µl of room -temperature chloroform was added and mixed 166 thoroughly. After incubating on ice for 5 minutes, the mixture was centrifuged at 18,000 x g for 15 167 minutes at 4°C to separate the aqueous phase. The aqueous layer was carefully collected, and 150 µl 168 of room-temperature isopropanol was added and mixed thoroughly. The mixture was incubated at -169 80°C overnight and then centrifuged at 18,000 x g for 3 hours at 4°C to precipitate the RNA. The 170 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 8 supernatant was decanted, and the pellet was washed with 600 µl of 70% ethanol at room 171 temperature and centrifuged at 18,000 x g for 1 hour at 4°C. The RNA pellet was air-dried in a laminar 172 flow hood and dissolved in 20 µl of nuclease -free water warmed to 70°C. The RNA was treated with 173 DNase to ensure purity. A cDNA library was prepared using the SuperScript III cDNA preparation kit. 174 All qRT-PCR primers were made using the ENCODE database, and PCR was performed using SYBR 175 Green. 176 177 178 Table 1. qPCR primers for m6A regulators. 179 180 Cloning of FTO overexpression and knockdown vectors 181 FTO cDNA was obtained through PCR amplification from the pRK5-myc-FTO vector . The PCR product 182 included an N-terminal T2A sequence and specific overhangs. The FTO-T2A PCR product was run on 183 an agarose gel, after which the desired band was excised and purified. The purified FTO-T2A fragment 184 was then cloned into the pAAV .hSynapsin.EGFP .WPRE.bGH (Osten-Frank) vector (Penn Core Vectors: 185 p1696), which had been linearised using the AgeI restriction enzyme, through infusion cloning. To 186 propagate the recombinant vector , the Stbl3 strain of E. coli was used (Fig. S2, S12). 187 No LncRNA Forward Primer (5’ – 3’) Reverse Primer (5’ – 3’) 1 Tubulinβ-1 GCGCCTTTGGACACCTATTCA CCGCGCCCTCCGTATAGTGC 2 GAPDH CGTATTGGGCGCCTGGTCAC CGGCCTCACCCCATTTGATG 3 Hprt ACCTCTCGAAGTGTTGGATACAGG CTTGCGCTCATCTTAGGCTTTG 4 Mettl14 AGACGCCTTCATCTCTTTGG AGCCTCTCGATTTCCTCTGT 5 Mettl3 ATTGAGAGACTGTCCCCTGG AGCTTTGTAAGGAAGTGCGT 6 Alkbh5 ACAAGATTAGATGCACCGCG TGTCCATTTCCAGGATCCGG 7 Fto CTGAGGAAGGAGTGGCATG TCTCCACCTAAGACTTGTGC 8 Ythdf1 CATTATGAGAAGCGCCAGGA AGATGCAACAATCAACCCCG 9 Ythdf2 ACCAACTCTAGGGACACTCA GGATAAGGAGATGCAACCGT 10 Ythdf3 TGCACATTATGAAAAGCGTCA AGATGCGCTGATGAAAACCA (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 9 For FTO RNAi, two independent shRNA constructs targeting FTO (shRNA1 and shRNA2) were 188 individually cloned into the AAV-U6-sgRNA-hSyn-mCherry vector (Addgene plasmid #87916) using 189 SacI and ApaI restriction enzyme sites. The use of two distinct shRNAs minimised the likelihood of 190 off-target effects. A scrambled shRNA control was similarly cloned into the same vector to serve as a 191 negative control (Fig. S6, S13). 192 193 No. Primers Oligo sequence 5’-3’ 1. Forward TGCTAGCAAGGATCCACCGGATGAAGCGCGTCCAGACCG 2. Reverse CTCACCATGGTGGCGACCGGTGGGCCGGGATTTTCCTCCACGTCCCCGCATGTTAGTA GACTTCCCCTGCCCTCGGATCTTGCTTCCAGCAGCTG 194 Table 2. Infusion primers for insertion of FTO-t2a sequence in AAV-Synapsin-EGFP backbone. 195 No. Primers Oligo sequence 5’-3’ 1. Control shRNA ATCTCGCTTGGGCGAGAGTAAG 2. FTO shRNA 1 (KD1 RNAi) AGATCGCCGCTGCATGTCA 3. FTO shRNA 2 (KD2 RNAi) GCAGCTGAAATACCCTAAACT 196 Table 3. FTO shRNA primers inserted in AAV-U6-sgRNA-hSyn-mCherry backbone. 197 198 HEK-293t cell culture and adeno-associated virus preparation 199 HEK-293t cells were grown and propagated in low -glucose DMEM solution (Gibco) with 10% fetal 200 bovine serum. Cells were maintained in incubators at 37 °C with 5% CO2. All expression vectors were 201 propagated utilising the Stbl3 strain of E. coli, while the packaging plasmids, pAAV2/9n and pDeltaF6, 202 were propagated using the DH5 α strain. Purification of the plasmid was achieved through an 203 Endotoxin-free Maxiprep kit (Qiagen). HEK293t cells underwent transfection using the purified 204 plasmid. For transfection of 2 × 10^6 cells (one 100mm plate), a mixture comprising 10 µg of the 205 chosen expression vector , 10µg pAAV2/9n, and 20µg of pAdDeltaF6 was prepared in a 206 microcentrifuge tube. To this mixture, 500 µl of 2X HBS was added, and the total volume was adjusted 207 to 1ml with nuclease -free water . 50 µl of 2.5 M calcium chloride (CaCl2) was added dropwise, 208 followed by vortexing and incubation at room temperature for 25 minutes. Post -incubation, the 209 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 10 mixture was added dropwise to the culture in the 100mm plate and thoroughly mixed. The culture 210 was then maintained at 37°C with 5% CO2 in low -glucose DMEM (Gibco) supplemented with 10% 211 fetal bovine serum, with a single media change after 8 hours. The culture was maintained for 96 hours 212 post-media change. The addition of NaCl (final concentration 500 mM) to the plate was followed by 213 incubation at 37°C for 3 hours to lyse the cells and release the virus into the supernatant. The 214 supernatant was collected in a centrifuge tube and stored at 4°C. The remaining cells in the dish were 215 scraped and collected, undergoing a free -thaw cycle: -80°C for 30 minutes followed by 37°C for 30 216 minutes, repeated three times to ensure complete dissociation of viral particles from the cell 217 membrane. The scrapped cells and the previously collected supernatant were filtered through a 100 218 µ nylon mesh filter initially to separate the membrane debris. The resulting filtrate was further 219 filtered through a 45 µ filter , and the final filtrate was concentrated using Amicron -Ultra-15 (Merk 220 Millipore #UFC910024) filters. Virus titre determination and calculation of the multiplicity of 221 infection (MOI) were carried out by infecting 80K HEK293T cells with serially diluted AAV solutions. 222 An MOI of 1-2 was employed for subsequent experiments. 223 Primary Neuronal Culture 224 Following a previous protocol (Kaech & Banker , 2006), primary hippocampal culture was performed. 225 Pregnant CD1 dams were selected and sacrificed in a CO 2 chamber when their embryos were at the 226 E16 stage (embryonic day 16). The E16 pups were dissected out from the uterus and kept in ice-cold 227 Dissection Media for proper anaesthesia. The pups were then decapitated to extract the brain. The 228 hippocampi were dissected and collected in cold dissection media (Gibco). It was then digested with 229 the addition of 2.5% trypsin to the dissection media at 37 ̊C for 10 mins. Trypsin -treated tissue was 230 washed with dissection media thrice at room temperature and collected in Glial Minimum Essential 231 Media (GMEM; Gibco). Further , trituration was done first with an unpolished Pasteur pipette and then 232 with a fire -polished Pasteur pipette to obtain a single -cell suspension, which was plated on poly -L-233 lysine-coated 60mm dishes (1 mg/mL, Sigma) at a concentration of 200 -250 cells/mm2. The culture 234 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 11 was maintained at 37 ̊C, 5% CO 2 in Neurobasal Media (Gibco) with B27 supplement (Gibco) 235 throughout the experiment. Cells were infected with AAV at DIV-7 and harvested at DIV-21 for further 236 experiments. 237 Stereotaxic Surgery 238 All surgical procedures were conducted on male P30 C57BL/6 animals. Anaesthesia was induced by 239 a combination dose of Ketamine (80mg per 1kg of body weight) and Xylazine (20mg per 1kg of body 240 weight), administered via subcutaneous injections. Following anaesthesia, the fur covering the head 241 was shaved, and the animal was positioned onto the stereotaxic stage. An incision was made on the 242 skin, exposing the underlying skull. The coordinates of the Bregma were precisely determined as (X, 243 Y , Z) = (0, 0, 0), serving as a reference point. The skull was drilled to create a small aperture in the 244 cranium at the predetermined coordinates for CA1 (Anterior -0.18cm; Lateral +/ -0.19cm; Ventral -245 0.12cm relative to Bregma (0, 0, 0)). The viruses were bilaterally injected into the CA1 region at a 246 controlled rate of 0.1 µl/min, with a total volume of 0.5 µl per hemisphere. Following the injection of 247 0.25 µl, a brief 5 -minute pause was given to prevent excessive cranial pressure buildup. The syringe 248 was carefully withdrawn 10 minutes after the 0.5 µl injection and an interval of 5 minutes was 249 maintained before injection into the second hemisphere. Upon completion, the surgical site was 250 sealed, and the incision was sutured using a nylon thread. The animal was transferred to a clean cage, 251 where it was allowed to recover for 1 week. Experimental procedures were conducted 2 weeks post-252 surgery, ensuring ample recovery time. 253 Cryosectioning 254 For cryosectioning, the mouse brain was harvested and fixed using a 4% paraformaldehyde solution 255 to preserve tissue integrity. The brain was further incubated in a 4% paraformaldehyde solution for 256 complete fixation of tissue. Post -fixation, the brain is incubated in a 30% sucrose solution until it 257 becomes fully saturated, which takes 5-6 days. The brain is then flash -frozen using dry ice to ensure 258 rapid and uniform freezing. Using the cryostat, the frozen brain is sectioned into 50 µm thick slices, 259 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 12 at a maintained temperature of around -20°C. These sections are collected onto slides pre-coated with 260 gelatine to enhance tissue adherence. The slides are stored at -80°C or processed immediately for 261 further staining and microscopic examination. 262 Spatial Object Recognition Assay 263 Male C57BL/6 mice at postnatal day 50 (P50) were housed individually under standard housing 264 conditions. The mice underwent a one-week habituation period where they were gently handled for 265 2 minutes daily. The spatial object recognition assay involved five sessions, each lasting 10 minutes 266 (Leger et al., 2013; Wimmer et al., 2012). The first four sessions were training sessions conducted on 267 the same day, followed by test session 3 hours or 24 hours later . In Session 1, mice were introduced 268 to an empty arena (12 cm x 15 cm) with spatial cues such as triangular and circular markings on one 269 wall and alternating black and white stripes on another wall. This allowed the mice to explore and 270 habituate for 10 minutes. During Sessions 2 -4, two objects were placed at different locations in the 271 arena, with a 2 cm gap between each object and the arena wall, allowing the mice to explore freely 272 for 10 minutes per session. A 5 -minute interval between sessions was used to clean the arena with 273 70% ethanol to remove any odours. The training involved pairs of mice in two separate arenas 274 simultaneously. For the test session, conducted 3 hours or 24 hours after the training, one object (a 275 cylindrical glass object) in one arena was moved to a new location, while no changes were made in 276 the other arena, which served as a control. The mice were then reintroduced to their respective 277 arenas and allowed to explore for 10 minutes. All sessions were recorded using an overhead-mounted 278 Logitech HD webcam C270. The recordings were analysed using MATLAB-based software to measure 279 the time spent with each object. Movement patterns and bout counts were determined, with a bout 280 defined as an exploration of an object within a 2 cm radius, excluding climbing on the object. 281 Additionally, all analyses were manually blind-tested. 282 The discrimination index (DI) was calculated to assess the animals' ability to recognise object 283 displacement, using the formula: 284 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 13 DI = [time spent with object displaced - time spent with object not displaced] / [time spent with 285 displaced object + time spent with undisplaced object] 286 A positive DI indicated a preference for the displaced object, reflecting the retention of spatial 287 memory. 288 Light/Dark Box Assay 289 Male C57BL/6 mice at P42 underwent a one-week habituation period in a low-lux environment (2-5 290 Lux) (Bourin & Hascoe t, 2003; Takao & Miyakawa, 2006). Each mouse was gently habituated for two 291 minutes daily in the experimental room during this period. Throughout the study, the mice were 292 individually housed in standard cages under standard housing conditions. The experimental setup 293 included a light/dark apparatus consisting of a dark box (occupying one-third of the total exploration 294 area) and a bright box (occupying two -thirds of the total exploration area), separated by a sliding 295 door . The dark box was kept at a low light intensity of 2 Lux, while the bright box had a higher 296 intensity of 390 Lux. At the beginning of each test session, the mouse was placed in the dark box. After 297 10 seconds, the door separating the two boxes was opened, allowing the mouse to explore both 298 compartments freely for 10 minutes. To track and analyse the mouse's behaviour , a Logitech HD 299 webcam C270 was positioned overhead to record its movements. The recorded data was analysed 300 using Toxtrac, a free online software, which provided measurements such as the time spent in the 301 light box, distance travelled within the light box, transitions between compartments, etc. (Rodriguez 302 et al., 2018). 303 Puromycin incorporation assay 304 To assess global protein synthesis in primary hippocampal neurons, the Click -iT™ Plus OPP Protein 305 Synthesis Assay Kit, Alexa Fluor ™ 488 (Thermo Fisher Scientific), was employed following the 306 manufacturer’s protocol. Primary hippocampal neuronal cultures were prepared and maintained 307 under standard conditions. At DIV 7, neurons were transfected with 2µg FTO knockdown constructs 308 (shRNA1 and shRNA2) and Control RNAi using Lipofectamine 2000 (Invitrogen). To halt 309 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 14 transcriptional activity, Actinomycin D (5 µg/mL) was applied to the culture media for 30 minutes at 310 37°C before OPP treatment to ensure that any newly synthesized proteins result from translation of 311 pre-existing mRNA pools. Following actinomycin treatment, neurons were incubated with O -312 propargyl-puromycin (OPP) at a final concentration of 20 µM for 30 minutes at 37°C. OPP is 313 incorporated into nascent polypeptide chains, acting as a puromycin analogue to label newly 314 synthesised proteins. After incubation, cells were gently washed with PBS and fixed with 4% 315 paraformaldehyde for 15 minutes at room temperature. Cells were permeabilized using 0.25% Triton 316 X-100 in PBS for 10 minutes, followed by Click -iT reaction according to the manufacturer’s 317 instructions to conjugate Alexa Fluor ™ 488 azide to incorporated OPP . Nuclei were counterstained 318 and were mounted using Vectashield-DAPI. 319 Imaging 320 Fluorescence imaging was performed using a point -scanning confocal microscope (Nikon AXR). The 321 integrated OPP signal per cell was quantified using ImageJ. Images were acquired with a 60X oil 322 immersion objective (NA = 1.4) at a resolution of 1024 × 1024 pixels, using a step size of 0.75 µm 323 across 10 –12 z -sections per image. Fast sequential acquisition mode was used with minimum 324 spectral crosstalk. Fluorophores were excited with 405 nm (DAPI), 488 nm (OPP incorporation/GFP), 325 561 nm (mCherry), and 640 nm (MAP2/far -red) lasers. For higher -magnification imaging of 326 dendrites, a 2X optical zoom was applied with a step size of 0.5 µm. Post -acquisition, images were 327 processed using Fiji, and maximum intensity projections were generated. mCherry-positive neurons 328 were masked using MAP2 staining, and OPP intensity was measured within selected dendritic regions 329 of interest (ROIs) to assess de novo protein synthesis. The same approach was used to quantify GFP 330 intensity within the soma. 331 Analysis of RNA-IP data 332 Raw data were obtained from the independent datasets. Adapter trimming and quality filtering were 333 performed using Cutadapt (v3.7), where low -quality bases (Phred score <20) and Illumina adapter 334 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 15 sequences were removed. The resulting high -quality reads were aligned to the mouse reference 335 genome (mm10) using HISAT2 (v2.2.1), and alignment statistics were compiled. Peaks of enriched 336 methylation were identified using exomePeak (v2.16.0), which also detected differential methylation 337 between experimental conditions. Statistically significant and replicate -consistent peaks were 338 classified as hypermethylated (log2FC > 0) or hypomethylated (log2FC < 0), with further filtering 339 based on a ≥2-fold change. Peaks were annotated as coding or noncoding based on genomic context. 340 Volcano plots were generated to visualise p-values and fold changes, highlighting robust differentially 341 methylated sites across conditions. Violin plots were generated to visualise the global shift of m6A 342 deposition along the transcripts. 343 Primary neuronal culture recordings 344 Primary hippocampal neurons were prepared and maintained under standard conditions. Neurons 345 were transfected at DIV7 with 2 µg of FTO -targeting shRNA constructs (shRNA1) or control RNAi 346 using Lipofectamine 2000 (Invitrogen). Recordings were performed at DIV18–DIV25. 347 Whole-cell patch -clamp recordings were performed using borosilicate glass electrodes (3 –8 M Ω). 348 Neurons were voltage-clamped at −70 mV . mEPSCs were recorded for 300 s per cell in extracellular 349 solution consisting of (in mM: NaCl 119, KCl 5, CaCl₂ 2, MgCl₂ 2, glucose 30, HEPES 10), with pH 350 adjusted to 7.4 and osmolarity of 310 –320 mOsm, supplemented with tetrodotoxin (1 µM) and 351 bicuculline (10 µM). The internal solution consisted of (in mM: cesium gluconate 100, EGTA 0.2, MgCl₂ 352 5, ATP 2, GTP 0.3, HEPES 40), with pH adjusted to 7.2 and osmolarity of 285 –290 mOsm. 353 Cells with series resistance ≤30 MΩ and stable recordings were analysed. Cells with holding currents 354 < −100 pA or unstable recordings were excluded. 355 Statistical analysis 356 Statistical analyses were performed for all experiments, with data presented as mean ± SEM. The 357 designation 'N' refers to the number of pregnant dams or litters used, while 'n' indicates the total 358 number of pups obtained from these dams. For qRT-PCR data, immunoblots on m6A regulators, and 359 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 16 puromycin incorporation assay, we employed unpaired Student’s t-tests with Welch’s correction for 360 unequal variance. Paired Student’s t -tests were used for analysing FTO overexpression and 361 knockdown efficiency data in HEK293t cells. Two -way ANOVA with Fisher’s LSD test and no 362 correction was applied for analysing light/dark and SOR data. Data for litter effect was assessed using 363 2-way ANOVA and corrected with Bonferroni’s multiple comparison test. Post -hoc power analysis 364 was conducted using GPower 3.1 for all experiments. Outliers were identified using the ROUT method 365 (0.1% aggressive). Immunoblot band intensity analysis and image stitching were performed using 366 ImageJ software. The volcano plot was obtained with FDR > 0.05 Gene Ontology plots were generated 367 using RStudio and SRplot (Tang et al. 2023). Statistical significance was defined as p < 0.05. 368 369

370 Maternal separation induces deposition of m6A mark in hippocampal transcriptome during 371 early development. 372 Previous studies established the role of m 6A in developing neural cells and also identified its 373 dysregulation as an emerging hallmark of neurodegenerative disorders, e.g. Huntington's Disease (Du 374 et al., 2019; Pupak et al., 2022). To understand the effect of ELS on global m 6A modification, we 375 performed a bulk RNA -seq from m 6A-modified precipitated hippocampal RNAs (Fig. 1A -B). This 376 approach revealed 291 significantly hypermethylated transcripts among 853 hypermethylated 377 candidates (both mRNAs and lncRNAs) and one significantly hypomethylated transcript out of 71 378 hypomethylated candidates (Fig. 1C). Moreover , analysis of the distribution of methylation peaks 379 indicated a broader shift in m 6A deposition along lncRNA transcripts and within the 3’ -UTR and 5’ -380 UTR of mRNAs, suggesting a mechanism by which early-life stress may reprogram hippocampal gene 381 expression through RNA methylation (Fig. 1D-E). In addition, we examined the m6A peak enrichment 382 around canonical DRACH motifs within transcripts encoding key RNA -binding proteins (RBPs) 383 relevant to neuronal function. Among the candidates analysed exhibited reduced enrichment at the 384 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 17 adenosine of the DRACH motif, where the m 6A peak is typically localised (Fig. 1H). A comparable 385 pattern of altered distribution was observed for Pum1 and, to a lesser extent, G3bp2 (Fig 1M, 1I). In 386 contrast, Srsf2, Srsf9 and Fxr1 did not display detectable changes in peak profiles (Fig. 1L, 1K, 1J). 387 Whereas Srsf1 and Pum2 showed a shift in the m6A peak (Fig. 1F , 1G), thereby suggesting the 388 existence of distinct modes of stress-induced remodelling of the RBP-associated methylation sites. 389 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 18 390 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 19 Figure 1. Maternal Separation enhances m6A modifications of hippocampal transcripts . (A) 391 Experimental workflow, N = 3. (B) Representative dot blot showing m6A immunoprecipitation (IP) 392 efficiency, N = 3. (C) Volcano plot for RNA expression of home -caged (HC) versus maternally 393 separated (MS) P21 pups. Red dots = significantly upregulated RNAs; blue dots = significantly 394 downregulated RNAs in maternally separated pups, total variables = 23,211, N = 3, FDR > 0.05. (D) 395 Density plot for distribution of m6A peaks along different regions of mRNA transcripts (5' UTR, coding 396 sequence (CDS), and 3' UTR), between HC and MS groups, N=3. (E) Density plot for distribution of 397 m6A peaks along lncRNAs in HC vs MS groups, N=3. (F-M) The relative change position of the binding 398 site of different RBPs: (F) SRSF1, (G) PUM2, (H) MBNL1, (I) G3BP2, (J) FRX1, (K) SRSF9, (L) SRSF2, 399 and (M) PUM1 in relation to the m6A (DRACH) motif; the “ A” position is considered as 0. 400 Also see Supplementary Figure 1. 401 —--------------------------------------------------------------------------------------------------------------------------- 402 403 Maternal separation selectively downregulates FTO expression in a sex-specific manner 404 Given the dynamic regulation of m 6A by readers, writers and erasers, we examined their expression 405 at the transcript as well as protein level, to identify the factors underlying the altered methylation 406 landscape induced by maternal separation. 407 We assessed the expression of core m6A regulatory components in hippocampal tissue from P21 male 408 mice using quantitative real -time PCR. No significant alterations were detected in the methylases 409 METTL3 (52.93±41.65, p=0.24) and METTL14 (65.63±51.07, p=0.24) (Fig. 2A -B), the demethylase 410 Alkb5 (29.07±18.19, p=0.15) (Fig. 2D), or the readers YTHDF2 (14.91 ± 19.46, p=0.46) and YTHDF3 411 (17.57± 14.15, p=0.24) (Fig. 2F -G). In contrast, maternal separation significantly reduced the 412 expression of the demethylase FTO (38.58 ± 14.64, p=0.02) (Fig. 2C) and the reader YTHDF1(36.15 ± 413 14.35, p=0.03) (Fig. 2E), suggesting that early -life stress selectively targets components of the m 6A 414 machinery. 415 To observe the expression at the protein level, we performed immunoblotting for METTL3 and FTO 416 in hippocampus lysates. In P21 male mice, METTL3 protein abundance was not significantly altered 417 (0.6140 ± 0.4141, p=0.15) (Fig. 2J), whereas FTO expression was significantly reduced following MS 418 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 20 (0.4549 ± 0.1757, p=0.02) (Fig. 2I). To determine whether this effect persisted into young adulthood, 419 we re-examined the expression profile of METTL3 and FTO at P28. We found a similar pattern where 420 FTO was significantly downregulated (0.6170±0.2238, p=0.023) and there was no effect on METTL3 421 protein expression (0.2554±0.2805, p=0.37) (Fig 2L, M). 422 To understand the sexual dimorphism in this regulation, we also examined the levels of FTO and 423 METTL3 in the hippocampus of maternally separated female mice at P21 and P28. At P21, expression 424 of FTO as well as METTL3 in female hippocampus were similar to that of P21 males; with a 425 significantly downregulated FTO (0.6842±0.178, p=0.004) and insignificant change in METTL3 426 expression (0.098±0.188, p=0.612) (Fig. S1A -C). But at P28, there was no significant difference in 427 expression of FTO (0.049±0.316, p=0.87) or METTL3 (0.0005±0.33, p=0.99) between the MS and the 428 control mice (Fig. S1D-F). The sustained reduction of FTO, observed in males, appears to be transient 429 in females. This highlights a sexually dimorphic regulation in epigenetic response to ELS. 430 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 21 431 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 22 Figure 2. Regulation of readers, writers and erasers upon maternal separation. Differential 432 expression of methylases (A) METTL3 and (B) METTL14, demethylases (C) FTO and (D) Alkb5 and 433 reader proteins (E) YTHDF1, (F) YTHDF2 and (G) YTHDF3 transcripts as obtained by qPCR. *p < 0.05, 434 **p < 0.01, ns = not significant. N=7, Data shown as Mean ± SEM. Unpaired Student’s t -test with 435 Welch's correction. Immunoblot showing expression of FTO and METTL3 in males (H) at P21, and 436 (K) at P28. Quantification of FTO expression in (I) P21; N=3-4, n=7-11 and in (L) P28; N=3-4, n=8, 437 males. Quantification of METTL3 expression in (J) P21 and in (M) P28 males. *p < 0.05, **p < 0.01, ns 438 = not significant. Mean ± SEM. Unpaired Student’s t-test with Welch's correction. 439 ---------------------------------------------------------------------------------------------------------------------------- 440 441 FTO overexpression does not rescue MS-induced anxiety-like behaviour 442 MS markedly increased anxiety in P45 C57BL/6 male mice, as shown by reduced distanced traveled 443 in the light compartment (5.908±2.796, p=0.04) and fewer transitions made between the two 444 compartments (13.86±3.683, p=0.0009) as compared to the HC (homecaged) animals (Fig. 3A -3D). 445 The change in time spent in the light compartment showed no significant difference (19±72.26, 446 p=0.79) between the MS and the HC cohort. This was probably because, due to heightened anxiety, 447 the MS animals froze upon entering the light box (Fig. 3E). 448 FTO overexpression in the hippocampus did not alleviate anxiety -like behaviour in the MS group. 449 FTO-OE (FTO overexpression) MS mice exhibited comparable anxiety levels to EGFP MS controls, 450 with no significant differences in time spent in the light box (54.47±66.62, p=0.42), distance travelled 451 (4.107±2.578, p=0.12), or transitions made (6.686±3.395, p=0.059). FTO overexpression alone did 452 not affect baseline anxiety, as no differences were observed between FTO -OE HC and EGFP HC mice 453 across time (46.57±75.21, p=0.54), distance (3.796±2.91, p=0.2) or transitions (3.643±3.833, 454 p=0.35) (Fig. 3C-E). FTO-OE MS mice displayed elevated anxiety relative to FTO-OE HC controls (time: 455 11.1±69.81, p=0.87; distance: 13.81±2.701, p< 0.0001; transitions: 16.9±3.558, p< 0.0001), which 456 again mirrored the behavioural pattern of EGFP MS mice (time: 11.1±69.81, p=0.87; distance: 457 13.81±2.701, p< 0.0001; transitions: 10.21±3.833, p=0.01) (Fig. 3C-E). To rule out litter-specific bias, 458 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 23 we analysed behavioural variations within litters under the same conditions and detected no 459 significant differences (Fig. S5). Collectively, FTO overexpression in the hippocampus did not rescue 460 the anxiety phenotype induced by MS. 461 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 24 462 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 25 Figure 3. FTO overexpression does not reverse elevated anxiety following maternal 463 separation. (A) Paradigm depicting light/dark box assay. (B) Representative trace in the bright -lit 464 arena in the light/dark box. HC = Homecaged, MS = Maternally Separated, EGFP = Empty plasmid with 465 GFP fluorescence, FTO -OE= Plasmid with FTO overexpression and GFP fluorescence. (C) Distance 466 covered in the light box (in meters). (D) Total number of transitions made between light and dark 467 box. (E) Time spent in the light box (in seconds). Data shown as mean±SEM, N=3, n=7 -10, *p<0.05 468 **p< 0.01, ***p<0.005, ****p<0.0001, ns not significant. Two-way ANOVA with Fisher’s LSD test. 469 Also see supplementary figure S5 470 —---------------------------------------------------------------------------------------------------------------------------- 471 472 FTO overexpression rescues spatial memory deficits of young adult mice caused by maternal 473 separation. 474 FTO regulates memory consolidation via an m 6A-dependent BDNF-TrkB signalling pathway 475 (Chang et al., 2023; Song et al., 2024). Previous studies have demonstrated that de novo 476 protein translation is differentially engaged across early (1–3 hrs) and late (24 hrs) phases 477 following neuronal activity, corresponding to short-term and long-term memory formation 478 (Santini et al., 2014, Schafe et al., 2000). We therefore investigated the role of FTO in 479 rescuing spatial memory deficits induced by ELS by testing spatial memory at two time 480 points post-training: 3 hours and 24 hours. 481 Test at 3 hours post-training: The SOR assay measures preference for a displaced object over a non -482 displaced object. But at 3 hours post-training, the percentage time spent with displaced object in test 483 session of EGFP-MS animals was comparable to that of the EGFP -HC group (3.739±6.03, p>0.0009). 484 Additionally, MS mice receiving the FTO-OE (FTO-OE MS) construct also did not display a significant 485 difference in percentage time spent with displaced object in test session compared to their EGFP-MS 486 counterparts (5.028±6.03, p>0.0009) (Fig. 4C). Briefly, no significant change in spatial memory was 487 observed in MS animals upon FTO overexpression at the hippocampus 3 hours post training. 488 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 26 Test at 24 hours post -training: The test sessions conducted 24 hours post -training showed 489 significantly low exploration of the displaced object in EGFP -MS animals as compared to the EGFP -490 HC animals (14.67±4.508, p=0.051). When FTO was overexpressed in the hippocampus of MS animals 491 (FTO-OE MS), this loss of spatial memory was rescued, as shown by a higher percentage of time spent 492 with the displaced object in the test session as compared to EGFP -MS group (12.64±4.41, p=0.04). 493 The exploration of FTO -OE MS animals was comparable to FTO -OE HC animals (4.692±4.41, 494 p>0.0009) and to EGFP -HC animals (2.03±4.287, p>0.0009). Thus, FTO overexpression conferred a 495 measurable improvement in spatial memory under ELS. Importantly, FTO-OE HC group did not alter 496 baseline exploration with percentage time spent with displaced object in test session values similar 497 to EGFP-HC mice (2.662±4.508, p>0.0009) (Fig. 4E). 498 The behavioural variation within litter was ruled out by studying the litter effect within the same 499 experimental condition (Fig. S4). Further , no object preference was seen in any experimental group, 500 ruling out any object bias (Fig. S3). Therefore, we observed that FTO overexpression in the 501 hippocampus of MS-induced animals was sufficient to rescue spatial memory deficits 24 hours post-502 training. 503 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 27 504 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 28 Figure 4. FTO overexpression reverses the effect of MS -induced deficits in memory . (A) 505 Paradigm for SOR. (B, D) Representative heat maps in the arena with the displaced object during the 506 test session, white ‘D’= object displaced; (B) 3 hours and (D) 24 hours post -training. HC = 507 Homecaged, MS = Maternally Separated, EGFP = Empty plasmid with GFP fluorescence, FTO -OE = 508 Plasmid with FTO overexpression and GFP fluorescence. (C, E) Percentage time spent with displaced 509 object during the test session, (C) 3 hours, N=3, n=7-9, and (E) 24 hours, N=3, n=9-11, post-training. 510 Data shows Mean ± SEM., *p< 0.05, ns = not significant. Two-way ANOVA with Bonferroni’s correction. 511 Also see supplementary figures S3 and S4. 512 —---------------------------------------------------------------------------------------------------------------------------- 513 FTO downregulation reduces protein synthesis in primary hippocampal neurons 514 FTO knockdown was used to mimic the stress-induced reduction observed in vivo. This resulted in a 515 significant decrease in puromycin -labelled nascent proteins, reflecting diminished translation 516 activity in the absence of FTO (Fig. 5A). To minimise off -target effects, two shRNA constructs were 517 used: FTO-RNAi_1 and FTO_RNAi_2. The knockdown efficiency of these constructs was verified by 518 transfecting each construct and a control RNAi in HEK293T cells, followed by an immunoblot (FTO -519 RNAi_1: 55.12±20.87, p=0.02; FTO-RNAi_2: 64.84±22.78, p=0.15) (Fig. S6). FTO knockdown resulted 520 in lower protein synthesis both in soma (FTO -RNAi_1: 54.89 ± 12.50, p<0.0001; FTO -RNAi_2: 50.17 521 ± 13.08, p=0.0003) and the dendrites (FTO -RNAi_1: 38.65 ± 7.579, p<0.0001; FTO -RNAi_2: 42.77 ± 522 7.259, p<0.0001) (Fig. 5B, C). Together , these data indicate that FTO is required to maintain basal 523 protein synthesis in hippocampal neurons, and its loss disrupts translation in both somatic and 524 dendritic domains of mature neurons. 525 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 29 526 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 30 Figure 5. FTO RNAi reduces translation in primary hippocampal neurons. (A) Representative 527 images of cultured primary hippocampal neurons transduced with control or FTO -targeting shRNA 528 constructs (FTO_RNAi_1 and FTO_RNAi_2), co -expressing mCherry (shRNA marker , red), O -529 propargyl-puromycin (OPP) (GFP channel, green), and MAP2 (fRed, magenta). Scale bars: 50 μ. (B–530 C) Quantification of puromycin incorporation as % fold change in signal intensity in (B) dendrites 531 and (C) soma. Data are shown as Mean ± SEM. Statistical significance was determined by Unpaired t-532 test with Welch’s correction, N = 24-36, n = 94-172, ***p < 0.001, ****p < 0.0001. 533 —---------------------------------------------------------------------------------------------------------------------------- 534 m6A-modified RNA -interacting RBPs regulate RNA metabolism including protein synthesis 535 Gene Ontology (GO) enrichment of the hypermethylated transcripts revealed broad associations with 536 RNA metabolic processes, including RNA 5’ -end processing, ribosome biogenesis, translation and 537 ribonucleoprotein complex assembly (Fig. 6). To gain deeper insights into the functional pathways 538 potentially regulated by these transcripts, we performed GO analysis on their interacting RBPs. The 539 biological process showed a significant regulation for pathways including mRNA metabolic processes, 540 RNA localisation, cytoplasmic translation and post -transcriptional regulation of gene expression. 541 Other terms, like dendritic transport of ribonucleoprotein, highlighted the potential regulation of 542 translation in neuronal processes. Additionally, enrichment of terms like stress granule assembly and 543 3’UTR-mediated RNA stabilisation and destabilisation suggest a possible MS -induced translational 544 control of these RBPs (Fig. S7A). 545 Further , cellular component analysis was enriched with terms including dendritic spines, distal 546 axons, synapses, polysomes and cytoplasmic stress granules. These terms are associated with the 547 localisation of these RBPs to neuronal compartments specialised for RNA storage and translation (Fig. 548 S7B). Molecular function enrichment revealed a predominance of RNA -binding activities, including 549 mRNA 3’UTR binding, translation initiation factor binding and N6 -methyldenosine (m6A)-modified 550 RNA binding, as well as interactions with regulatory RNA elements such as G -quadruplexes and 551 miRNAs (Fig S7C). Together , these findings suggest that the m6A-modified RNA-interacting RBPs are 552 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 31 well positioned to orchestrate spatial events at the synapse, potentially contributing to experience -553 dependent synaptic plasticity. 554 555 556 Figure 6. Gene Ontology enrichment of m⁶A -modified transcripts upon maternal separation. 557 (A) Biological process, (B) Cellular components, FDR>0.05. 558 Also see supplementary figure S7. 559 560

561 The differential m 6A methylation patterns observed in the hippocampus of male mice 562 following maternal separation (MS) indicate a substantial reorganisation of the m 6A 563 epitranscriptomic landscape. Transcriptomic analysis of RNA -seq data revealed 291 564 transcripts that were significantly hypermethylated and only one that was hypomethylated 565 after MS, suggesting a global shift toward hypermethylation (Fig. 1C). Previous studies have 566 reported that m 6A modifications within the 3′UTR promote transcript localisation to 567 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 32 dendrites and axons in cultured hippocampal neurons (Livneh et al., 2020; Widagdo et. al., 568 2018). This process is essential for synaptic function and plasticity (Malovic & Pandey, 2023). 569 Consistent with these observations, analysis showed a shift in m 6A deposition toward the 570 3′UTR regions of coding transcripts and along the distal ends of lncRNAs, suggesting ELS -571 induced remodelling of the m6A -topology (Fig. 1D -E). We also observed a few candidates 572 that are known to have a role in synaptic functions and translation, like Srsf1, Srsf2, Srsf9, 573 Pum1, Pum2, Frx1, G3bp2 and Mbnl1. This shows a shift in m 6A density along the conserved 574 DRACH motif (D= A, G, U; R= A, G; H= A, C, U) that specifically recruits methyltransferase 575 complexes to catalyse m6A modifications (Bayoumi & Munir, 2021; Ji et al., 2024; Santos -576 Rodriguez et al., 2024) (Fig. 1F-M). 577 Concurrently, the m6A demethylase FTO was significantly downregulated after MS, whereas 578 the expression of methyltransferases remained unchanged in both sexes (Fig. 2, Fig. S1). The 579 reduction in FTO levels following maternal separation aligns with findings from previous 580 studies, which also report decreased FTO expression in the hippocampus in response to 581 stress (Jiang et al., 2023; Kisliouk et al., 2020). However, FTO downregulation persisted 582 longer in males than in females, hinting towards a sexually dimorphic regulation of m 6A-583 decoration in the hippocampal post -ELS (Fig. 2, Fig. S1). The reduction of FTO may lead to 584 the accumulation of m 6A marks on stress -sensitive transcripts and potentially alter their 585 stability, localisation, or translation. Thereby, it can impair neuronal and behavioural 586 functions. 587 588 The reduction in FTO levels following MS is supported by previous findings, which report 589 decreased FTO expression in the hippocampus and impaired working memory (Spychala & 590 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 33 Ru ther, 2019). Similarly, in individuals with major depressive disorder (MDD), FTO was 591 found to be downregulated in the hippocampus (Mitsuhashi & Nagy, 2023). Recent studies 592 have also shown that FTO expression at the synapse can affect coding as well as non -coding 593 RNAs involved in synaptic functions and behaviour (Liau et al., 2023; Walters et al., 2017). 594 Behaviourally, overexpression of FTO in the hippocampus did not reverse the anxiety -like 595 phenotype induced by MS. The movement patterns were characterised by thigmotaxis and 596 freezing, which are indicative of anxiety rather than exploration (Simon et al., 1994) (Fig. 3). 597 These effects were not attributable to litter variability (Fig. S5). Thus, FTO overexpression 598 appears insufficient to mitigate MS -induced anxiety, implying that additional stress -599 responsive molecular pathways contribute to anxiety -related behaviours. In contrast, FTO 600 overexpression partially restored spatial memory deficits induced by MS in the SOR assay. 601 While MS animals failed to recognise the displaced object at 24 hours post -training, FTO-602 overexpressing MS animals successfully identified it (Fig. 4D -E). The rescue effect was 603 independent of litter differences (Fig. S4) and highlights that m 6A-dependent regulation of 604 memory-related transcripts is partially reversible by elevating FTO levels. Interestingly, there 605 was no memory deficit in any condition 3 hours post -training (Fig. 4B -C). The absence of 606 memory deficits 3 hours post-training indicates that short-term memory formation remains 607 unaffected by MS and is independent of FTO expression. However , the impairment observed 608 at 24 hours and its partial rescue by FTO overexpression imply that FTO influences the 609 stabilisation of transcripts required for long-term memory persistence. 610 In primary hippocampal neurons, knockdown of FTO was done to mimic post -stress FTO 611 levels in the hippocampus. We observed that loss of FTO led to a decrease in global protein 612 synthesis both in dendrites and soma (Fig. 5). This implies that FTO modulates post -613 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 34 transcriptional translation efficiency rather than local translation. Therefore, FTO affects 614 transcripts necessary for sustaining long -term synaptic plasticity. Stress -induced 615 downregulation of FTO may impair memory consolidation by disrupting m 6A-dependent 616 regulation of protein synthesis in hippocampal neurons. 617 Further, to identify biological pathways affected by altered methylation, GO analysis of 618 hypermethylated transcripts revealed enrichment in translation, ribosome biogenesis, and 619 mitochondrial function, implying widespread effects on neuronal energetics and protein 620 synthesis (Fig. 6). Further, GO analysis of RNA -binding proteins (RBPs) interacting with 621 hypermethylated transcripts identified key terms related to RNA transport, localization, and 622 stability, as well as dendritic localization, ncRNA processing, and 3′UTR -mediated RNA 623 stabilization. Cellular component terms included synapse, dendrite, polysome, and P -body, 624 while molecular function categories included translation activation, RNA and lncRNA 625 binding, miRNA binding, and ribosome association (Fig. S7). These results collectively 626 suggest that stress -induced hypermethylation perturbs RNA -RBP interactions critical for 627 synaptic plasticity. 628 Together, these findings propose a model wherein early-life stress downregulates FTO in the 629 hippocampus, leading to persistent hypermethylation of m6A-marked transcripts involved in 630 protein synthesis and synaptic regulation. This epitranscriptomic shift alters RNA -RBP 631 interactions and translation dynamics. Therefore, it can have long-lasting effects on neuronal 632 plasticity and cognitive function. While anxiety-like behaviours appear independent of FTO-633 mediated demethylation, spatial memory deficits are, at least partially, reversible through 634 restoration of FTO activity. This finding also highlights a dissociation between molecular 635 pathways underlying stress-induced anxiety and memory impairments. 636 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 35

637 The study is supported by NBRC core fund . Plasmids used for designing viral vectors were obtained 638 from Addgene and Penn Vector Core. All art has been created with BioRender .com. 639

640 1. Akhtar , J., Lugoboni, M., & Junion , G. (2021). m6A RNA modification in transcription 641 regulation. Transcription, 12(5), 266–276. 642 https://doi.org/10.1080/21541264.2022.2057177 643 2. Alves, J., de Sa Couto-Pereira, N., de Lima, R. M. S., Quillfeldt, J. A., & Dalmaz, C. (2022). Effects 644 of Early Life Adversities upon Memory Processes and Cognition in Rodent Models. 645 Neuroscience, 497, 282–307. https://doi.org/10.1016/j.neuroscience.2022.04.023 646 3. Anders, M., Chelysheva, I., Goebel, I., Trenkner , T ., Zhou, J., Mao, Y ., Verzini, S., Qian, S. B., & 647 Ignatova, Z. (2018). Dynamic m6a methylation facilitates mRNA triaging to stress granules. 648 Life Science Alliance, 1(4), 1–12. https://doi.org/10.26508/lsa.201800113 649 4. Banerjee, D., Sultana, S., & Banerjee, S. (2024). Gas5 regulates early-life stress-induced anxiety 650 and spatial memory. Journal of Neurochemistry , 168(9), 2999 –3018. 651 https://doi.org/10.1111/jnc.16167 652 5. Bayoumi, M., & Munir , M. (2021). Evolutionary conservation of the DRACH signatures of 653 potential N6 -methyladenosine (m6A) sites among influenza A viruses. Scientific Reports , 654 11(1). https://doi.org/10.1038/s41598-021-84007-0 655 6. Bertagna, N. B., dos Santos, P . G. C., Queiroz, R. M., Fernandes, G. J. D., Cruz, F . C., & Miguel, T . T . 656 (2021). Involvement of the ventral, but not dorsal, hippocampus in anxiety -like behaviors in 657 mice exposed to the elevated plus maze: participation of CRF1 receptor and PKA pathway. 658 Pharmacological Reports, 73(1), 57–72. https://doi.org/10.1007/s43440-020-00182-3 659 7. Bourin, M., & Hascoe t, M. (2003). The mouse light/dark box test. European Journal of 660 Pharmacology, 463(1–3), 55–65. https://doi.org/10.1016/S0014-2999(03)01274-3 661 8. Briggs, J. A., Wolvetang, E. J., Mattick, J. S., Rinn, J. L., & Barry, G. (2015). Mechanisms of Long 662 Non-coding RNAs in Mammalian Nervous System Development, Plasticity, Disease, and 663 Evolution. Neuron, 88(5), 861–877. https://doi.org/10.1016/j.neuron.2015.09.045 664 9. Brosens, N., Simon, C., Kessels, H. W ., Lucassen, P . J., & Krugers , H. J. (2023). Early life stress 665 lastingly alters the function and AMPA-receptor composition of glutamatergic synapses in the 666 hippocampus of male mice. Journal of Neuroendocrinology , 35(12). 667 https://doi.org/10.1111/jne.13346 668 10. Chang, R., Zhu, S., Peng, J., Lang, Z., Zhou, X., Liao, H., Zou, J., Zeng, P ., & Tan, S. (2023). The 669 hippocampal FTO -BDNF-TrkB pathway is required for novel object recognition memory 670 reconsolidation in mice. Translational Psychiatry, 13(1). https://doi.org/10.1038/s41398 -671 023-02647-4 672 11. Du, K., Zhang, L., Lee, T ., & Sun, T . (2019). m 6 A RNA Methylation Controls Neural Development 673 and Is Involved in Human Diseases. In Molecular Neurobiology (Vol. 56, Number 3, pp. 1596–674 1606). Humana Press Inc. https://doi.org/10.1007/s12035-018-1138-1 675 12. Ernst, C., & Morton, C. C. (2013). Identification and function of long non-coding RNA. Frontiers 676 in Cellular Neuroscience, 7(OCT), 1–9. https://doi.org/10.3389/fncel.2013.00168 677 13. Fatica, A., & Bozzoni, I. (2014). Long non-coding RNAs: New players in cell differentiation and 678 development. Nature Reviews Genetics, 15(1), 7–21. https://doi.org/10.1038/nrg3606 679 14. Hanson, J. L., Nacewicz, B. M., Sutterer , M. J., Cayo, A. A., Schaefer , S. M., Rudolph, K. D., Shirtcliff, 680 E. A., Pollak, S. D., & Davidson, R. J. (2015). Behavioral problems after early life stress: 681 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 36 Contributions of the hippocampus and amygdala. Biological Psychiatry , 77(4), 314 –323. 682 https://doi.org/10.1016/j.biopsych.2014.04.020 683 15. Ji, J., Mu, X., Xu, S., Xu, X., Zhang, Z., Yao, L., Xie, Q., & Bi, Y . (2024). Conservation and distribution 684 of the DRACH motif for potential m6A sites in avian leukosis virus subgroup J. Frontiers in 685 Veterinary Science, 11. https://doi.org/10.3389/fvets.2024.1374430 686 16. Jia, G., Fu, Y ., Zhao, X., Dai, Q., Zheng, G., Yang, Y ., Yi, C., Lindahl, T ., Pan, T ., Yang, Y . G., & He, C. 687 (2011). N6-Methyladenosine in nuclear RNA is a major substrate of the obesity -associated 688 FTO. Nature Chemical Biology, 7(12), 885–887. https://doi.org/10.1038/nchembio.687 689 17. Jiang, Y ., Zhang, T ., Yang, L., Du, Z., Wang, Q., Hou, J., Liu, Y ., Song, Q., Zhao, J., & Wu, Y . (2023). 690 Downregulation of FTO in the hippocampus is associated with mental disorders induced by 691 fear stress during pregnancy. Behavioural Brain Research , 453(July), 114598. 692 https://doi.org/10.1016/j.bbr .2023.114598 693 18. Kaech, S., & Banker , G. (2006). Culturing hippocampal neurons. Nature Protocols, 1(5), 2406–694 2415. https://doi.org/10.1038/nprot.2006.356 695 19. Kisliouk, T ., Rosenberg, T ., Ben-Nun, O., Ruzal, M., & Meiri, N. (2020). Early -life m6A RNA 696 demethylation by fat mass and obesity -associated protein (Fto) influences resilience or 697 vulnerability to heat stress later in life. ENeuro, 7(3), 1 –14. 698 https://doi.org/10.1523/ENEURO.0549-19.2020 699 20. Lee, S. L., Lew, D., Wickenheisser , V ., & Markus, E. J. (2019). Interdependence between dorsal 700 and ventral hippocampus during spatial navigation. Brain and Behavior , 9(10), 1 –14. 701 https://doi.org/10.1002/brb3.1410 702 21. Leger , M., Quiedeville, A., Bouet, V ., Haelewyn, B., Boulouard, M., Schumann-Bard, P ., & Freret, 703 T . (2013). Object recognition test in mice. Nature Protocols , 8(12), 2531 –2537. 704 https://doi.org/10.1038/nprot.2013.155 705 22. Levin, R. Y ., & Liu, R. T . (2021). Life stress, early maltreatment, and prospective associations 706 with depression and anxiety in preadolescent children: A six -year , multi-wave study. Journal 707 of Affective Disorders , 278(September 2020), 276 –279. 708 https://doi.org/10.1016/j.jad.2020.09.072 709 23. Liau, W . S., Zhao, Q., Bademosi, A., Gormal, R. S., Gong, H., Marshall, P . R., Periyakaruppiah, A., 710 Madugalle, S. U., Zajaczkowski, E. L., Leighton, L. J., Ren, H., Musgrove, M., Davies, J., Rauch, S., 711 He, C., Dickinson, B. C., Li, X., Wei, W ., Meunier , F . A., … Bredy, T . W . (2023). Fear extinction is 712 regulated by the activity of long noncoding RNAs at the synapse. Nature Communications, 713 14(1). https://doi.org/10.1038/s41467-023-43535-1 714 24. Lin, L. Y ., Zhang, J., Dai, X. M., Xiao, N. A., Wu, X. L., Wei, Z., Fang, W . T ., Zhu, Y . G., & Chen, X. C. 715 (2016). Early -life stress leads to impaired spatial learning and memory in middle -aged 716 ApoE4-TR mice. Molecular Neurodegeneration , 11(1), 1 –16. 717 https://doi.org/10.1186/s13024-016-0107-2 718 25. Livneh, I., Moshitch -Moshkovitz, S., Amariglio, N., Rechavi, G., & Dominissini, D. (2020). The 719 m6A epitranscriptome: transcriptome plasticity in brain development and function. Nature 720 Reviews Neuroscience, 21(1), 36–51. https://doi.org/10.1038/s41583-019-0244-z 721 26. Loedige, I., Baranovskii, A., Mendonsa, S., Dantsuji, S., Popitsch, N., Breimann , L., Zerna, N., 722 Cherepanov, V ., Milek, M., Ameres, S., & Chekulaeva, M. (2023). mRNA stability and m6A are 723 major determinants of subcellular mRNA localization in neurons. Molecular Cell , 83(15), 724 2709-2725.e10. https://doi.org/10.1016/j.molcel.2023.06.021 725 27. Ma, L., Bajic, V . B., & Zhang, Z. (2013). On the classification of long non -coding RNAs. RNA 726 Biology, 10(6), 924–933. https://doi.org/10.4161/rna.24604 727 28. Malovic, E., & Pandey, S. C. (2023). N 6-methyladenosine (m6A) epitranscriptomics in synaptic 728 plasticity and behaviors. Neuropsychopharmacology, 48(1), 221 –222. 729 https://doi.org/10.1038/s41386-022-01414-1 730 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 37 29. Mattick, J. S., Amaral, P . P ., Carninci, P ., Carpenter , S., Chang, H. Y ., Chen, L. L., Chen, R., Dean, C., 731 Dinger , M. E., Fitzgerald, K. A., Gingeras, T . R., Guttman, M., Hirose, T ., Huarte, M., Johnson, R., 732 Kanduri, C., Kapranov, P ., Lawrence, J. B., Lee, J. T ., … Wu, M. (2023). Long non -coding RNAs: 733 definitions, functions, challenges and recommendations. Nature Reviews Molecular Cell 734 Biology, 24(June), 430–447. https://doi.org/10.1038/s41580-022-00566-8 735 30. Mercer , T . R., Dinger , M. E., & Mattick, J. S. (2009). Long non -coding RNAs: Insights into 736 functions. In Nature Reviews Genetics. https://doi.org/10.1038/nrg2521 737 31. Mercer , T . R., Dinger , M. E., Sunkin, S. M., Mehler , M. F ., & Mattick, J. S. (2008). Specific 738 expression of long noncoding RNAs in the mouse brain. Proceedings of the National Academy 739 of Sciences of the United States of America , 105(2), 716 –721. 740 https://doi.org/10.1073/pnas.0706729105 741 32. Mitsuhashi, H., & Nagy, C. (2023). Potential Roles of m6A and FTO in Synaptic Connectivity 742 and Major Depressive Disorder . International Journal of Molecular Sciences , 24(7). 743 https://doi.org/10.3390/ijms24076220 744 33. Pan, G., Chen, Z., Zheng, H., Zhang, Y ., Xu, H., Bu, G., Zheng, H., & Li, Y . (2018). Compensatory 745 Mechanisms Modulate the Neuronal Excitability in a Kainic Acid -Induced Epilepsy Mouse 746 Model. Frontiers in Neural Circuits, 12. https://doi.org/10.3389/fncir .2018.00048 747 34. Papatheodoropoulos, C. (2025). Compensatory Regulation of Excitation/Inhibition Balance in 748 the Ventral Hippocampus: Insights from Fragile X Syndrome. In Biology (Vol. 14, Number 4). 749 Multidisciplinary Digital Publishing Institute (MDPI). 750 https://doi.org/10.3390/biology14040363 751 35. Pattwell, S. S., & Bath, K. G. (2017). Emotional learning, stress, and development: An ever -752 changing landscape shaped by early -life experience. Neurobiology of Learning and Memory , 753 143, 36–48. https://doi.org/10.1016/j.nlm.2017.04.014 754 36. Ping, X. L., Sun, B. F ., Wang, L., Xiao, W ., Yang, X., Wang, W . J., Adhikari, S., Shi, Y ., Lv, Y ., Chen, Y . 755 S., Zhao, X., Li, A., Yang, Y ., Dahal, U., Lou, X. M., Liu, X., Huang, J., Yuan, W . P ., Zhu, X. F ., … Yang, 756 Y . G. (2014). Mammalian WTAP is a regulatory subunit of the RNA N6 -methyladenosine 757 methyltransferase. Cell Research, 24(2), 177–189. https://doi.org/10.1038/cr .2014.3 758 37. Pupak, A., Singh, A., Sancho -Balsells, A., Alcala -Vida, R., Espina, M., Giralt, A., Martí , E., Ørom, 759 U. A. V ., Gine s, S., & Brito, V . (2022). Altered m6A RNA methylation contributes to hippocampal 760 memory deficits in Huntington’s disease mice. Cellular and Molecular Life Sciences, 79(8), 1–761 21. https://doi.org/10.1007/s00018-022-04444-6 762 38. Rodriguez A, Zhang H, Klaminder J, Brodin T , Andersson PL, Andersson M (2018) ToxTrac: A 763 fast and robust software for tracking organisms. Methods Ecol Evol 9:460–464. 764 39. Samaddar Sarbani. (2023). Transcriptomic analysis identifies synapse -enriched lncRNAs 765 required for excitatory synapse development and fear memory. BioRxiv, 1 –50. 766 https://doi.org/https://doi.org/10.1101/2023.07.14.549055 767 40. Santini, E., Huynh, T . N., & Klann, E. (2014). Mechanisms of translation control underlying 768 long-lasting synaptic plasticity and the consolidation of long -term memory. In Progress in 769 Molecular Biology and Translational Science (Vol. 122, pp. 131 –167). Elsevier B.V . 770 https://doi.org/10.1016/B978-0-12-420170-5.00005-2 771 41. Santos-Rodriguez, G., Srivastava, A., Ravindran, A., Oyelami, F ., Ip, K., Gupta, P ., Villanueva, J., 772 King, H., Grootveld, A., Blackburn, J., Gupta, I., Vieira, H., Shirokikh, N., Eyras, E., & Weatheritt, 773 R. (2024). The conserved landscape of RNA modifications and transcript diversity across 774 mammalian evolution. https://doi.org/10.1101/2024.11.24.624934 775 42. Sendinc, E., & Shi, Y . (2023). RNA m6A methylation across the transcriptome. Molecular Cell, 776 83(3), 428–441. https://doi.org/10.1016/j.molcel.2023.01.006 777 43. Schafe, G. E., Ledoux, J. E., & Keck, W . M. (2000). Memory Consolidation of Auditory Pavlovian 778 Fear Conditioning Requires Protein Synthesis and Protein Kinase A in the Amygdala. 779 http://www.jneurosci.org/cgi/content/full/4501 780 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 38 44. Shi, C., Zhang, L., & Qin, C. (2017). Long non -coding RNAs in brain development, synaptic 781 biology, and Alzheimer’s disease. Brain Research Bulletin , 132(5), 160 –169. 782 https://doi.org/10.1016/j.brainresbull.2017.03.010 783 45. Shi, H., Zhang, X., Weng, Y . L., Lu, Z., Liu, Y ., Lu, Z., Li, J., Hao, P ., Zhang, Y ., Zhang, F ., Wu, Y ., 784 Delgado, J. Y ., Su, Y ., Patel, M. J., Cao, X., Shen, B., Huang, X., Ming, G. li, Zhuang, X., … Zhou, T . 785 (2018). m6A facilitates hippocampus -dependent learning and memory through YTHDF1. 786 Nature, 563(7730), 249–253. https://doi.org/10.1038/s41586-018-0666-1 787 46. Simon, P ., Dupuis, R., & Costentin , J. (1994). Thigmotaxis as an index of anxiety in mice. 788 Influence of dopaminergic transmissions. Behavioural Brain Research , 61(1), 59 –64. 789 https://doi.org/10.1016/0166-4328(94)90008-6 790 47. Song, J., Hao, J., Lu, Y ., Ding, X., Li, M., & Xin, Y . (2024). Increased m6A modification of BDNF 791 mRNA via FTO promotes neuronal apoptosis following aluminum -induced oxidative stress. 792 Environmental Pollution, 349. https://doi.org/10.1016/j.envpol.2024.123848 793 48. Sood, A., Pati, S., Bhattacharya, A., Chaudhari, K., & Vaidya, V . A. (2018). Early emergence of 794 altered 5-HT2A receptor-evoked behavior , neural activation and gene expression following 795 maternal separation. International Journal of Developmental Neuroscience , 65(September 796 2017), 21–28. https://doi.org/10.1016/j.ijdevneu.2017.10.005 797 49. Spratt, P . W . E., Alexander , R. P . D., Ben-Shalom, R., Sahagun, A., Kyoung, H., Keeshen, C. M., 798 Sanders, S. J., & Bender , K. J. (2021). Paradoxical hyperexcitability from NaV1.2 sodium 799 channel loss in neocortical pyramidal cells. Cell Reports , 36(5). 800 https://doi.org/10.1016/j.celrep.2021.109483 801 50. Spychala, A., & Ru ther , U. (2019). FTO affects hippocampal function by regulation of BDNF 802 processing. PLoS ONE, 14(2), 1–14. https://doi.org/10.1371/journal.pone.0211937 803 51. Takao, K., & Miyakawa, T . (2006). Light/dark transition test for mice. Journal of Visualized 804 Experiments, 1, 1–3. https://doi.org/10.3791/104 805 52. Walters, B. J., Mercaldo, V ., Gillon, C. J., Yip, M., Neve, R. L., Boyce, F . M., Frankland, P . W ., & 806 Josselyn, S. A. (2017). The Role of The RNA Demethylase FTO (Fat Mass and Obesity -807 Associated) and mRNA Methylation in Hippocampal Memory Formation. 808 Neuropsychopharmacology, 42(7), 1502–1510. https://doi.org/10.1038/npp.2017.31 809 53. Wang, A., Wang, J., Liu, Y ., & Zhou, Y . (2017). Mechanisms of long non -coding RNAs in the 810 assembly and plasticity of neural circuitry. Frontiers in Neural Circuits , 11(October). 811 https://doi.org/10.3389/fncir .2017.00076 812 54. Wang, A., Zou, X., Wu, J., Ma, Q., Yuan, N., Ding, F ., Li, X., & Chen, J. (2020). Early -Life Stress 813 Alters Synaptic Plasticity and mTOR Signaling: Correlation With Anxiety-Like and Cognition-814 Related Behavior . Frontiers in Genetics , 11(December), 1 –11. 815 https://doi.org/10.3389/fgene.2020.590068 816 55. Wang, P ., Doxtader , K. A., & Nam, Y . (2016). Structural Basis for Cooperative Function of Mettl3 817 and Mettl14 Methyltransferases. Molecular Cell , 63(2), 306 –317. 818 https://doi.org/10.1016/j.molcel.2016.05.041 819 56. Widagdo, J. (n.d.). Journal of Neurochemistry - 2018 - Widagdo - The m6A-epitranscriptomic 820 signature in neurobiology from neurodevelopment to.pdf. 821 57. Wimmer , M. E., Hernandez, P . J., Blackwell, J., & Abel, T . (2012). Aging impairs hippocampus-822 dependent long-term memory for object location in mice. Neurobiology of Aging, 33(9), 2220–823 2224. https://doi.org/10.1016/j.neurobiolaging.2011.07.007 824 58. Xie, M., Pallegar , P . N., Parusel, S., Nguyen, A. T ., & Wu, L. J. (2023). Regulation of cortical 825 hyperexcitability in amyotrophic lateral sclerosis: focusing on glial mechanisms. In Molecular 826 Neurodegeneration (Vol. 18, Number 1). BioMed Central Ltd. 827 https://doi.org/10.1186/s13024-023-00665-w 828 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint 39 59. Zaccara, S., & Jaffrey, S. R. (2020). A Unified Model for the Function of YTHDF Proteins in 829 Regulating m6A -Modified mRNA. Cell, 181(7), 1582 -1595.e18. 830 https://doi.org/10.1016/j.cell.2020.05.012 831 60. Zheng, G., Dahl, J. A., Niu, Y ., Fedorcsak, P ., Huang, C. M., Li, C. J., Va ̊ gbø, C. B., Shi, Y ., Wang, W . L., 832 Song, S. H., Lu, Z., Bosmans, R. P . G., Dai, Q., Hao, Y . J., Yang, X., Zhao, W . M., Tong, W . M., Wang, 833 X. J., Bogdan, F ., … He, C. (2013). ALKBH5 Is a Mammalian RNA Demethylase that Impacts RNA 834 Metabolism and Mouse Fertility. Molecular Cell , 49(1), 18 –29. 835 https://doi.org/10.1016/j.molcel.2012.10.015 836 837 838 839 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 30, 2026. ; https://doi.org/10.64898/2026.03.30.715262doi: bioRxiv preprint