Novel LncRNA Gm44763 Regulates Morphine-Induced Reward Memory via MiR-298-5p-Mediated eIF4E Translation Control

Abstract

Drug-associated reward memory underlies both the development and relapse of addiction, yet its molecular basis remains poorly understood. Here, transcriptomic profiling and functional validation identified a novel long non-coding RNA (lncRNA), Gm44763, as a critical regulator of morphine-induced reward memory specifically in neurons of the medial prefrontal cortex (mPFC). Behavioral and molecular analyses demonstrated that Gm44763 functions as a sponge for miR-298-5p, thereby relieving the repression of the downstream target gene, eukaryotic translation initiation factor 4E (eIF4E), and modulating both the acquisition and retrieval of reward memory. Golgi staining and fiber photometry further revealed that Gm44763 normalized morphine-induced alterations in synaptic structure and neuronal excitability. miR-298-5p bidirectionally regulated morphine-induced reward memory and reversed both behavioral and neuronal effects mediated by Gm44763. Mechanistically, the downstream effector eIF4E modulates translation via its interaction with eIF4G, thereby contributing to morphine-induced memory regulation. This process can be effectively modulated by 4EGI-1, a selective inhibitor of the eIF4E/eIF4G interaction. In summary, this study characterized lncRNA expression profiles in the mPFC of mice with morphine-induced conditioned place preference. We identified and validated Gm44763 as a novel lncRNA regulator of morphine-induced reward memory and synaptic plasticity. We further delineate a previously uncharacterized Gm44763/miR-298-5p/eIF4E axis that may represent a novel regulatory pathway linking transcriptional and translational control to drug-associated memory formation.

Fig. 1.

Morphine-induced CPP model establishment in mice after 5 rounds of training and identification of differentially expressed lncRNAs. (A) Experimental schedule for morphine-induced CPP. Mice first underwent a 7-day habituation period, then underwent pre-test on day 0 and exposed to 5 days of consecutive training (days 1 to 5). From days 7 to 12, mice were housed in the home cage without drug administration. The retrieval test was performed on day 13 and mPFC tissue was extracted for RNA sequencing. (B) CPP scores of pre-test (day 0), post-test 1 (day 6), and post-test 2 (day 13). Data are expressed as mean ± SEM (n = 8). ****P < 0.0001 vs. pre-test; ^{^^^^}P < 0.0001 vs. saline group (2-way ANOVA followed by Bonferroni post hoc test). (C to E) Moving distance, mean speed, and shuttle times of pre-test, post-test 1 and post-test 2. ****P < 0.0001 vs. pre-test; ^^^^P < 0.0001 vs. saline group. n = 8 mice/group. (F) Representative motor paths and heatmap images of mice from both groups during the tests. The white chamber on the left represents the morphine-paired compartment. (G) Flowchart for validating and analyzing DElncRNAs. (H) Proportion of different types of lncRNAs among the identified DElncRNAs. (I) Heatmap showing clusters of top 100 DElncRNAs between saline and morphine samples. Red color gradients correspond to positive fold-change values, whereas blue gradients correspond to negative values. (J) Volcano plot illustrating DElncRNAs. Up-regulated lncRNAs are marked in red, while down-regulated ones are marked in blue. (K) KEGG enrichment analysis of DEmRNAs revealed significant enrichment in the cGMP-PKG signaling pathway and morphine addiction.

Fig. 2.

Gm44763 expression is reduced in the mPFC of mice with morphine-induced CPP. (A) The qPCR analysis of 4 up-regulated lncRNAs randomly selected from the sequencing data. Among these, 1700110k17rik was significantly up-regulated, while Gm28192and Gm30934 showed an upward trend. n = 6 to 8. *P < 0.05, Student’s t test, compared with the saline group. (B) qPCR analysis of 4 down-regulated lncRNAs randomly selected from the sequencing data. Gm44763, Gm42854, Gm26971, and Gm47405 were significantly down-regulated. n = 6 to 8, *P < 0.05, **P < 0.01 compared with the saline group. Student’s t test. (C) RNA FISH assays confirmed that the expression of Gm44763 in mPFC of morphine-treated mice were significantly lower than in the saline group. n = 6. **P < 0.01, Student’s t test. (D) Representative images show the colocalization of Gm44763 (red) with the neuronal marker NeuN (green), the microglial marker Iba1 (green), and the astrocytic marker S100β (green). Subcellular localization revealed that Gm44763 surrounds NeuN-positive nuclei and is distributed within the cytoplasm, while exhibiting no overlap with Iba1 or S100β. (E) RNA FISH assays were performed to analyze the subcellular distribution of Gm44763 in N2a and CATH.a cells; scale bar = 10 μm. (F) lncLocator and iLoc-LncRNA predictions supported the cytoplasmic enrichment of Gm44763. (G) Regional expression analysis revealed that Gm44763 showed no significant changes in the NAc, BLA, or Hipp of mice following morphine-induced CPP. n = 6. (H) Chromosomal location of Gm44763 and PhyloCSF scores for this region. (I) The CPAT and CPC scores of Gm44763 were significantly lower than those of the coding transcripts (Syn2a, Iqck, and Gapdh) as with known lncRNAs (Neat1 and Gas5). (J) Evolutionary conservation across 60 vertebrate species was assessed using PHAST package, and comparative genomic alignments between mouse, rat, and human genomes were visualized.

Fig. 3.

Overexpression of Gm44763 in the mPFC attenuated the rewarding effects of morphine. (A) Schematic representation of LV-Gm44763 constructs. (B) (Left) Fluorescence microscopy images demonstrate localized GFP expression in the mPFC. Scale bar = 500 μm. (Right) Fluorescence verification of mPFC-specific transduction. Scale bar = 50 μm. (C) qPCR validation of Gm44763 overexpression efficiency in the mPFC. n = 10, *P < 0.05 (Student’s t test). (D) Schematic timeline of the experimental procedure. LV-Gm44763 or LV-Control was stereotactically injected into the mPFC 14 days before CPP pre-test (day 0), followed by morphine/saline conditioning (days 1 to 5) and post-test 1 (day 6). Post-test 2 was conducted on day 13, after which mPFC tissue was collected for molecular analyses. (E) During both post-test 1 and post-test 2 phases, the CPP scores of the LV-Gm44763 + Morphine group were significantly lower than those of the LV-Control + Morphine group. Two-way ANOVA followed by Bonferroni post hoc test, ****P < 0.0001 vs. pre-test score of the same group; ##P = 0.006, ###P = 0.0008, morphine-induced differences in the LV-Gm44763 group of mice vs. LV-Control group in the post-tests, n = 8. (F) Representative locomotor traces and heatmap images of mice from the LV-Gm44763 and LV-Control groups during the CPP procedure following morphine conditioning. The white chamber on the left represents the morphine-paired compartment. (G) (Left) Representative locomotor activity traces in the OFT. (Right) No significant differences were observed between the LV-Gm44763 and control groups in total distance traveled or time spent in the center and periphery zones. n = 20, Student’s t test. (H) (Left) Representative locomotor activity traces in the EPM. (Right) No significant differences between LV-Gm44763 and control groups in either the number of arm entries or the time spent in open and closed arms. n = 20, Student’s t test. (I) (Left) Representative locomotor activity traces in the Y-maze. (Right) No significant difference in the alternation triplet in the LV-Gm44763 group of mice compared with the LV-Control group of mice. n = 17, Student’s t test. (J) (Left) Representative locomotor activity traces in the NOR. (Right) No significant differences were observed in the recognition index for either the short-term memory test (test 1) or the long-term memory test (test 2) between LV-Gm44763 and LV-Control mice. n = 17 to 20, Student’s t test. (K) In the sucrose preference test (SPT), mice showed a natural preference for sucrose, consuming more than 70% sucrose. n = 11 to 12, Student’s t test.

Fig. 4.

Gm44763 overexpression reverses morphine-induced synaptic remodeling and neuronal hyperactivity. (A) Synaptic protein dysregulation (PSD-95, VGluT1) in morphine-induced mice was rescued by Gm44763 overexpression. Syn1 protein levels showed no significant variation across groups. n = 4. *P < 0.05, ***P < 0.001, compared with LV-Control-Saline; #P < 0.05, ###P < 0.001 compared with LV-Control-Morphine. (B) Golgi–Cox staining visualized neuronal morphology in the mPFC. Sholl analysis was performed using concentric circles with 10-μm intervals and a radius range of 0 to 100 μm to quantify dendritic complexity. (C) The Sholl intersection profile revealed that morphine exposure enhanced dendritic complexity in the mPFC, which was normalized following Gm44763 overexpression. The X-axis represents the radial distance from the soma, while the Y-axis indicates the number of intersections at each concentric circle. n = 5. *P < 0.05, **P < 0.01 compared with LV-Control-Saline; #P < 0.05, ###P < 0.001 compared with LV-Control-Morphine. (D) Quantification of total dendritic branch length showed that morphine conditioning significantly increased branch length, while Gm44763 overexpression restored dendritic length to baseline levels. n = 10. ****P < 0.0001 compared with LV-Control-Saline; ####P < 0.0001 compared with LV-Control-Morphine. (E) Representative images of dendrites in the mPFC of mice. Scale bar = 25 μm. (F) Quantification of dendritic spine density (spines per 10 μm) showed that morphine exposure increased spine density, while Gm44763 overexpression normalized this effect. n = 10. ****P < 0.0001 compared with LV-Control-Saline; ####P < 0.0001 compared with LV-Control-Morphine. (G) Schematic illustration of the experimental setup showing coinjection of the red calcium indicator AAV-hSyn-JRGECO1a and the green LV-Gm44763 into the mPFC, followed by fiber implantation for in vivo calcium dynamics recording during the CPP test. (H) Fluorescence validation showing coexpression of the 2 viruses in the mPFC. Top: scale bar = 200 μm; bottom: scale bar = 20 μm. (I) Gm44763 overexpression partially reversed the morphine-induced neuronal hyperactivity, restoring calcium signaling toward baseline levels. (a) Calcium signals (ΔF/F) recorded from mPFC neurons of mice during pre-test, post-test 1, and post-test 2, showing calcium activity 2 s before and 6 s after entry events. Bottom: Time-aligned heatmaps of calcium signals (−2 to +6 s) relative to chamber entry. Each row represents an individual entry event. Color intensity reflects the normalized ΔF/F amplitude. (b) Quantification of the average calcium transients during chamber entry events. Overexpression of Gm44763 significantly attenuated morphine-induced neuronal hyperexcitability. n = 5. *P < 0.05, ****P < 0.0001 compared with pre-test; ^{^^^^}P < 0.0001 compared with LV-Control-Saline; ###P < 0.001, ####P < 0.0001 compared with LV-Control-Morphine. Data analyzed using 2-way ANOVA with Bonferroni post hoc test.

Fig. 5.

Identification and validation of the Gm44763-mediated ceRNA regulatory network. (A) Gm44763-mediated ceRNA network. In this network, red diamond represents lncRNA, purple rectangles represent miRNAs, and green ellipses represent mRNAs. (B and C) qPCR analysis showed that miR-298-5p expression was significantly up-regulated in mice following morphine exposure, while miR-370-3p expression remained unchanged. Both eIF4E and Nrxn1 were significantly down-regulated after morphine treatment. n = 6 to 7. *P < 0.05, ***P < 0.001 compared with the saline group. (D) (a) RIP experimental schematic of Gm44763 and eIF4E as RISC target genes. (b and c) RIP was performed on N2a cell lysates with anti-Ago2 or IgG antibodies. In the same precipitates, anti-Ago2 enriched miR-298-5p, Gm44763, and eIF4E. n = 3, *P < 0.05, **P < 0.01 compared with IgG-enriched group. (E) Schematic diagram of the binding site between Gm44763 and miR-298-5p. (F and G) Schematic diagram of the binding site between eIF4E and miR-298-5p, with the longest consecutive pairing consisting of 8 bases. The binding sites are indicated by red text or shaded gray areas. (H) Dual-luciferase assays verified the direct binding of Gm44763 to miR-298-5p. miR-298-5p mimics significantly reduced luciferase activity in both WT1 and WT2 reporters. Data were shown as the relative luciferase activity. n = 3. **P < 0.01, ****P < 0.0001 compared with mimics NC. (I) Targeted binding between eIF4E and miR-298-5p was demonstrated by dual-luciferase reporter assays. n = 3, **P < 0.01 compared with mimics NC group. Data analyzed using Student’s t test. (J) The FISH experiment showing the colocalization of Gm44763 (red) and miR-298-5p (green) in the mouse mPFC. Scale bar = 10 μm.

Fig. 6.

In vitro validation of Gm44763 regulation of eIF4E expression via competitive binding to miR-298-5p. (A) Experimental schedule to validate the efficacy of overexpression in LV-infected cells by assessing the expression of GFP and Gm44763. (B) Representative plots of GFP expression in LV-Gm44763- or LV-Control-infected CATH.a cells, scale bar = 50 μm. (C) Gm44763 expression was markedly elevated in CATH.a cells transduced with LV-Gm44763. n = 6, ****P < 0.0001 compared with the LV-Control group. Student’s t test. (D) Representative plots of GFP expression in LV-Gm44763- or LV-Control-infected N2a cells, scale bar = 50 μm. (E) N2a cells infected with LV-Gm44763 exhibited markedly increased expression of Gm44763, significant down-regulation of miR-298-5p, and a concomitant up-regulation of eIF4E mRNA. n = 6. *P < 0.05, ****P < 0.0001, compared with the LV-Control, Student’s t test. (F) Protein levels of eIF4E were significantly increased in LV-Gm44763-infected N2a cells. n = 4. *P < 0.05, compared with the LV-Control-infected group (Student’s t test). (G) Flowchart of the rescue experiment for transfection of miR-298-5p mimics/inhibitor in N2a cells infected with LV-Gm44763. (H) The expression of miR-298-5p was significantly increased by transfection with miR-298-5p mimics, and this effect was partially rescued by LV-Gm44763. n = 6. **P < 0.01, ****P < 0.0001. (I) The expression of eIF4E was significantly suppressed by transfection with miR-298-5p mimics, and this suppression was partially rescued by LV-Gm44763. n = 6. ****P < 0.0001. (J) Transfection with the miR-298-5p inhibitor significantly reduced the expression of miR-298-5p. ****P < 0.0001. (K) miR-298-5p inhibitor transfection significantly up-regulated eIF4E expression, and this effect was further amplified by LV-Gm44763 in a synergistic manner. n = 6. *P < 0.05, ****P < 0.0001. Data analyzed using 2-way ANOVA with Bonferroni post hoc test.

Fig. 7.

miR-298-5p bidirectionally regulates morphine-induced reward memory and reverses Gm44763-mediated suppression. (A) Schematic timeline of the experiment showing stereotactic delivery of AntagomiR-298-5p into the mPFC 7 days prior to the CPP pre-test. (B) qRT-PCR analysis confirmed that antamiR-298-5p administration specifically reduced miR-298-5p expression in the mPFC of morphine-treated mice. n = 8. *P < 0.05, **P < 0.01. (C) AntagomiR-298-5p significantly suppressed morphine-induced CPP scores at post-test 2. n = 8 to 10. ***P < 0.001, ****P < 0.0001 compared with pre-test; ^#P < 0.05 compared with antamiR-NC-Morphine. (D) Representative trajectory maps and 3D heatmaps of chamber exploration. (E) Fluorescence validation of AAV-miR-298-5p infection efficiency in the mPFC. Scale bars = 200 μm (left) and 50 μm (right). (F) qPCR analysis confirmed that AAV-miR-298-5p overexpression increased miR-298-5p levels in the mPFC. n = 8. *P < 0.05, **P < 0.01, ***P < 0.001. (G) Overexpression of miR-298-5p in the mPFC significantly increased CPP scores during both the post-test 1 and post-test 2. Two-way ANOVA followed by Bonferroni post hoc test, ****P < 0.0001 vs. pre-test score of the same group; ##P = 0.006, morphine-induced differences in the LV-Gm44763 group of mice vs. the LV-Control group in post-test 1, ##P = 0.002, morphine-induced differences in the LV-Gm44763 group of mice vs. the LV-Control group in post-test 2, n = 8. (H) Fluorescence validation of viral transduction efficiency. Scale bars = 200 μm (left) and 50 μm (right). (I) miR-298-5p overexpression reversed the Gm44763-induced increase in CPP scores and eIF4E expression at both mRNA and protein levels. (a) Overexpression of miR-298-5p reversed the significant effects of Gm44763 on both eIF4E mRNA and protein levels. n = 8. *P < 0.05, **P < 0.01, ***P < 0.001. (b) Overexpression of miR-298-5p in the mPFC reversed the inhibitory effect of Gm44763 overexpression on reward memory. n = 8. ****P < 0.0001 vs. pre-test score of the same group; ####P < 0.0001 vs. the control virus-Morphine group; ^^P = 0.002 vs. the LV-Gm44763-Morphine group in post-test 1, ^^P = 0.009 vs. the LV-Gm44763-Morphine group in post-test 2. Two-way ANOVA followed by Bonferroni post hoc test. (J) AAV-miR-298-5p enhanced morphine-induced neuronal hyperexcitability, which was attenuated by Gm44763 overexpression. (a) Fluorescence confirmed coexpression of the red calcium indicator AAV-hSyn-JRGECO1a and green LV-Gm44763 or AAV-miR-298-5p in the mPFC. Scale bar = 50 μm. (b) Top: Real-time calcium transients upon entry into morphine-paired chambers. Bottom: Event-locked calcium signal heatmap aligned to chamber entry onset (−2 to +6 s). (c) Mean ΔF/F during chamber entries showed that AAV-miR-298-5p further amplified morphine-induced neuronal hyperexcitability, while concurrent overexpression of Gm44763 significantly attenuated neuronal activity. n = 6. ****P < 0.0001 compared with pre-test; ^####P < 0.0001 compared with the AAV-298-5p-Morphine group; ^^P < 0.01, ^^^^P < 0.0001 compared with the control virus-Morphine group. Data analyzed using 2-way ANOVA with Bonferroni post hoc test.

Fig. 8.

eIF4E-mediated translational regulation is involved in morphine-induced reward memory. (A) eIF4E-mediated PPI network (B) GO enrichment analysis of proteins interacting with eIF4E revealed that the associated biological processes were primarily involved in the regulation of postsynaptic translation and synaptic transmission. (C) Morphine exposure reduces total eIF4E protein levels but increases its phosphorylation at Ser209 in the mPFC. GAPDH was applied as an internal loading control. n = 4, *P < 0.05, **P < 0.01 compared with the saline group. Student’s t test. (D) IP of eIF4E analysis confirmed that morphine exposure enhances the interaction between eIF4E and eIF4G, while reducing its association with 4E-BP in the mPFC. n = 6. **P < 0.01, ***P < 0.001 compared with the saline group, Student’s t test. (E) IP analysis using eIF4G (left) and 4E-BP (right) antibodies show that morphine exposure markedly increases the binding of eIF4E to eIF4G and decreases its interaction with 4E-BP in the mPFC. n = 6. **P < 0.01, ****P < 0.0001 compared with the saline group, Student’s t test. (F) SUnSET assay using puromycin incorporation shows that morphine-treated mice exhibit a significant elevation in overall protein translation in the mPFC. “–“ represents a control sample without puromycin. n = 6. **P < 0.01 compared with the saline group, Student’s t test. (G) SUnSET assay shows that overexpression of Gm44763 significantly reduced the translation rate in the mPFC compared with the LV-Control group. (Left) Representative images for the puromycin experiment. (Right) Analysis of the effect of Gm44763 on protein synthesis in mPFC. n = 6, ***P< 0.001. (H) Schematic workflow of morphine-induced translational activation simulated in N2a cells, inhibited by 4EGI-1, followed by IP and SUnSET assay. (I) 4EGI-1 significantly inhibited the interaction between eIF4E and eIF4G. IP of eIF4E. n = 3. *P < 0.05, ***P < 0.001, ****P < 0.0001, 2-way ANOVA followed by Bonferroni post hoc test. (J) Protein synthesis was blocked by 4EGI-1 treatment. n = 3. ***P < 0.001. (K) Protein–protein docking analysis revealed the potential interaction model between eIF4E and eIF4G. (L) 4EGI-1 impairs both the formation and retrieval of morphine-induced CPP. Coexpression of Gm44763 further enhances the suppressive effect of 4EGI-1 on CPP scores. n = 6. ****P < 0.0001 vs. pre-test; ^P = 0.017, ^^P = 0.007 vs. Vehicle-Morphine group; ^####P < 0.0001 vs. compared with 4EGI-Morphine group. Data analyzed using 2-way ANOVA with Bonferroni post hoc correction. (M) 4EGI-1 suppresses protein synthesis and synaptic plasticity-related protein expression in the mPFC. (a) Representative results of the SUnSET assay showing reduced global protein synthesis in the mPFC following 4EGI-1 administration. “–“ represents a control sample without puromycin. (b) Western blot revealed concomitant reductions in the expression of synaptic plasticity-related proteins, including PSD-95, Syn1, VGluT1, and SYP, consistent with the overall down-regulation of translation. Data are presented as mean ± SEM, n = 6, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, Student’s t test.

Fig. 9.

Mechanistic hypothesis diagram. Gm44763 regulates morphine-induced reward memory via the miR-298-5p/eIF4E axis. Down-regulation of Gm44763 following morphine exposure enhances miR-298-5p-mediated suppression of eIF4E, leading to altered synaptic protein synthesis, increased dendritic complexity, and facilitation of reward memory expression.