Introduction

Although cardiopulmonary resuscitation (CPR) and the restoration of spontaneous circulation (ROSC) can be obtained after cardiac arrest (CA), outcomes remain largely unfavorable. The majority of survivors experience neurological impairment, varying from mild cognitive impairment to being in a minimally conscious or persistent vegetative state. Only a small percentage of patients (5–17%) successfully regain full consciousness1,2. Significant efforts were made over time to comprehend the pathophysiological mechanisms underlying post-cardiac arrest brain injury (PCABI). These mechanisms include mitochondrial damage, disrupted cellular metabolism, the activation of neuroinflammation and reactive oxidative species (ROS), and energy-metabolic disorder. Ultimately, these factors lead to neurological impairment3,4,5. Despite advancements in therapeutic procedures, the overall efficacy of comprehensive therapy remains unsatisfactory. Thus, it is crucial to identify the fundamental pathogenic molecules associated with PCABI disorders in order to uncover the underlying processes that cause harm or neuroprotection.

Long non-coding RNA (lncRNA) represents a group of non-coding RNA molecules, known as non-coding RNAs (ncRNA), that are longer than 200-100000nt and do not possess the ability to code proteins6. MicroRNAs (miRNAs) are small RNA molecules that consist of 19–25 nucleotides, and are capable of post-transcriptional regulation of gene function in disease pathogenesis7. Recent studies have emphasized the role of these ncRNAs in thedevelopment and progression of neurological disorders, especially under conditions of hypoxia or anoxia8,9,10. Within particular cells and tissues, lncRNAs operate as competitive endogenous RNAs (ceRNA). These lncRNAs contain ‘seed sequences’ that are targets for specific miRNAs and act as sponges, binding to these miRNAs. This interaction indirectly regulates the expression levels of genes involved in development, thereby affecting a variety of biological processes11. However, research specifically investigating the regulating mechanisms of ceRNA in the context of neurological impairment resulting from CA remains limited.

The rapid development of bioinformatics techniques enables us to gain a deeper understanding of disease pathobiology at the genetic level12. In light of this, the goal of this research is to examine and analyze the ceRNA regulatory network linked to PCABI via the use of biochemical methods and bioinformatics analysis. Initially, we conducted RNA-Sequence investigation to detect genes that were expressed differently between the rat model of CA/CPR group and the sham group. Afterwards, the ceRNA was constructed using computational biology, based on the ceRNA hypothesis. In addition, Further analysis involved examining the hub ceRNA network regulated by lncRNA and conducting enrichment studies to elucidate the signaling pathways and fundamental mechanism. Ultimately, these discoveries were verified via quantitative real-time PCR (qRT-PCR) and luciferase assays. Hence, the proposed hub ceRNA network inour study provides a deeper insight into the fundamental regulatory mechanism during PCABI and may offer a theoretical basis for the development of novel therapeutic therapies.

Materials and methods

Statement

This study was approved by the Medical Ethics Committee of Hangzhou First People’s Hospital. The experiment followed the ARRIVE guidelines and the Guide for the Care and Use of Laboratory Animals approved by the National Institute of Health. All methods were carried out in accordance with relevant guidelines and regulations.

Animals

Male Sprague-Dawley rats, weighing between 350 and 450 g and aged between 8 and 12 weeks, were obtained from Hangzhou medical College. Prior to the commencement of the CA/CPR paradigm, all rats had 12 h of fasting, while still being able to freely consume water and food. The core temperature was regulated at 37.5 ± 0.3 °C via heating lamps throughout the process and for an additional 4 h thereafter.

Establishment of CA/CPR model

The rats were randomly allocated to the experimental regimen. The CA/CPR model was established as previously documented13,14. The rats were induced by a 45 mg/kg intraperitoneal injection of pentobarbital sodium, and further doses (10 mg/kg) were given when needed. Subsequently, the animals underwent mechanical ventilation. Following a 5-minute stabilization period, hypoxia was produced by removing the breathing tubing from the ventilator. The established criterion for CA was a reduction in systolic arterial pressure to a level below 25 mmHg. CPR was commenced 5 min after CA, using efficient mechanical breathing and intravenous administration of epinephrine. The requirements for ROSC were (1) reestablishment of supraventricular heart rhythm, and (2) maintenance of a mean artery pressure (MAP) above 60 mmHg for a minimum of 10 min. The sham group had identical anesthetic and surgical procedures as the CA groups, with the exception that they did not experience produced asphyxial CA and subsequent CPR. The experimental design is outlined in Fig. 1.

Fig. 1
figure 1

Animal preparation and experimental procedure.

Full size image

Neurological evaluation

Methods for assessing neurological damage are detailed in our team’s recent publication and include the Neurologic Deficit Score (NDS), cognitive function evaluations, pathology, electroencephalography, nerve injury biomarkers, magnetic resonance imaging, and microdialysis. Our analysis primarily utilized the NDS and pathological assessments15. The NDS measures many aspects such as awareness, breathing, walking, standing, muscle tone, motor response, and constraint. The NDS scale spans from 0 to 400, representing the spectrum from normal brain function to complete brain death, respectively. We designated the 24-hour mark post-ROSC as the pivotal timepoint for assessing early post-resuscitation neurological function. The NDS was assessed by three investigators who were unaware of the research details, and the ultimate rating for each rat was determined by calculating the average of all component scores.

Haematoxylin–eosin (HE), Nissl staining and immunohistochemistry

Following the completion of the behavioral studies, we euthanized the rats using intraperitoneal injection of pentobarbital sodium 160 mg/kg under deep anesthesia and extracted brain tissue for subsequent experiments. NE, Nissl staining and immunohistochemistry was performed as previously described16,17. In brief, the hippocampal brain tissues, measuring 0.4 cm in thickness, were promptly extracted and preserved in paraformaldehyde for 48 h. Subsequently, the tissues were embedded in paraffin and cut into slices measuring 5 μm in thickness. These sections underwent HE and Nissl’s staining to evaluate neuronal integrity. Neuronal injury in the CA1 region of the hippocampus was assessed microscopically; neurons were considered viable and healthy only if they exhibited clear nuclei and nucleoli. Subsequently, the brain tissue slices were stained with NeuN and 4′-6-diamino-2-phenylindole (DAPI) antibodies, and the resulting fluorescence images were obtained from the CA1 area utilizing an Inverted Fluorescence Stereomicroscope.

RNA-Seq

Hippocampus tissues were employed to capture total RNA using Trizol reagent (Invitrogen, China). The library’s quality was evaluated using the Agilent 2100 Bioanalyzer, while its quantitative analysis was conducted using the Quant-iT PicoGreen dsDNA Assay Kit. The sequencing was conducted via the Illumina Hiseq 2000 instrument (LC Bio, China) following the procedure indicated by the vendor. Subsequently, a procedure of information analysis was conducted.

LncRNA identification

Transcripts overlapping known mRNAs or shorter than 200 bp were excluded. The coding potential of the remaining transcripts was assessed using CPC18 and CNCI19. Transcripts with a CPC score < -1 and a CNCI score < 0 were eliminated. The surviving transcripts were classified as lncRNAs.

Differential expression (DE) analysis of mRNAs and lncRNAs

Expression levels of mRNAs and lncRNAs were quantified in terms of FPKM20 using StringTie21. Differentially expressed mRNAs, miRNAs and lncRNAs were identified based on a log2 (fold change) greater than 1 or less than − 1, with statistical significance established at a p-value < 0.05 using the R package DESeq2.

Target gene prediction

To investigate the function of lncRNAs, we predicted cis-acting targets within 100,000 bases upstream and downstream of each lncRNA using a Python script. Afterwards, the TargetScan (v5.0) and miRanda (v3.3a) software platforms were employed to identify target genes for DEmiRNA. Target predictions were filtered using specific scoring criteria from each platform. For TargetScan, gene targets were excluded if their context score percentile was below 50. For miRanda, gene targets were discarded if their maximum free energy of binding exceeded − 10 kcal/mol. The convergence of the predictions from both tools defined the final set of target genes. The comprehensive results are documented in the Supplementary Table S1 and S2.

The function of all differentially expressed genes (DEGs) was determined using Gene Ontology (GO) annotations and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, which were accessed via OmicStudio tools at https://www.omicstudio.cn/tool. A p-value < 0.05 was considered as significant.

Construction of hub lncRNA-miRNA-mRNA regulatory network

The ceRNA hypothesis suggests that intersecting molecules engage in ceRNA relationships. To identify these relationships, we applied stringent criteria that included directional regulation and a requirement that the absolute value of log2 (fold change) exceeds the median for both lncRNAs and mRNAs. This criterion was critical given the substantial number of genes with a log2 (fold change) exceeding 2. After filtering the DERNAs, we utilized metrics such as MCC, Closeness, and EPC in cytoHubba to delineate the hub ceRNA networks. Key regulatory components within the network were identified through an intersection-based method. Additionally, we conducted Pearson correlation analyses to investigate the expression patterns of lncRNAs, miRNAs, and mRNAs, further elucidating the regulatory dynamics within this network.

QRT-PCR validation

qRT-PCR was employed for validating the DEGs implicated in the ceRNA networks. Total RNA extraction was performed via TRIzol reagent. The RNA that was obtained was converted into cDNA using reverse transcription and then measured using ChamQ SYBR qPCR Master Mix. The gene expression level was determined by the 2-ΔΔCT technique. The primers were generated via the primer design feature inside the National Center for Biotechnology Information (NCBI) database. Supplementary Table S3 demonstrates the specific primers.

Dual-luciferase reporter assay

We used a dual-luciferase assay to identify the relationships between lncRNA MSTRG.13871 and miR-155-5p, as well as Grip1 and miR-155-5p. In summary, the wild-type (WT) lncRNA MSTRG.13871 (lncRNA MSTRG.13871-3’ untranslated region [UTR]-WT), mutant (Mut) lncRNA MSTRG.13,871 (lncRNA MSTRG.13871-3’UTR Mut), wild-type Grip1 (Grip1-3’UTR-WT), and mutant Grip1 (Grip1-3’UTR Mut) were inserted into the reporter vector designed to be targeted by miR-155b-5p. Either a miR-155-5p mimic or NC mimic, together with 3’UTR-WT or 3’UTR-Mut, were co-transfected into 293T cells via a dual Luciferase Assay (Promega, Corp., Fitchburg, WI, USA) 48 h after transfection.

Statistical analysis

The statistical analysis and graph production were performed using GraphPad Prism software version 9.5.0 (GraphPad Software, Inc., San Diego, CA, USA). The measurement results were presented as the average value plus or minus the standard deviation. The Student’s t-test employed for comparing data between two groups that were paired. A value < 0.05 was significant.

Result

Basic physiological characteristics

There was no statistically significant disparity in initial results between the two groups (see Supplementary Table S4 online). This study includes measurements such as body temperature, end-tidal carbon dioxide levels, mean arterial pressure, heart rate, and weight (P > 0.05). The average duration of CA was 299.6 ± 2.40 s, while the duration of CPR was 102.6 ± 22.72 s. Out of the experiment, 3 rats successfully survived 24 hourspost-resuscitation. The revived rats were subsequently evaluated for neurological impairments using a battery of assessments: NDS, NE and Nissl staining, immunofluorescence staining, RNA-seq, qRT-PCR confirmation, and dual-luciferase Assay.

Evaluation of neurological function

The NDS scores were substantially assessed in the CA/CPR group compared to the sham one (P < 0.0001, see Supplementary Fig. S1 online). After conducting behavioral trials, we assessed neuronal cell death to confirm the effective establishment of the CA/CPR model. This assessment especially focused on hippocampal CA1 neurons, which are notably vulnerable to damage during global cerebral ischemia. HE staining and Nissl’s staining of the CA1 regions revealed neuronal injury in the CA/CPR group, characterized by tissue edema, cytoplasmic loosening, nuclear shrinkage, and intense eosinophilic staining. In contrast, the sham group displayed cortical neurons with big, circular nuclei, clear and intact double nuclear membranes, and Nissl corpuscles in the cytoplasm. This suggests that the CA/CPR procedure led to significant neuronal cell deaths. Immunohistochemistryanalysis demonstrated a substantial reduction in the number of surviving NeuN-positive cells in the CA/CPR group compared to the sham group (see Supplementary Fig. S2 online). Such outcomes suggest that brain tissues can be a robust basis for further RNA-seq analysis.

Differential expression profile of DEGs

A total of 784 lncRNAs (478 down-regulated and 306 up-regulated), 16 miRNAs (8 down-regulated and 8 up-regulated), and 1858 mRNAs (1245 down-regulated and 613 up-regulated) had significant changes in the CA/CPR groups compared to the sham groups (Fig. 2, Supplementary Table S5-S7).

Fig. 2
figure 2

Screening of the DERNAs in the hippocampus tissues of CA/CPR group. (A-C) Volcano plot of DE lncRNAs (A), DE miRNAs (B), and DE mRNAs (C) (red, up-regulated; blue, down-regulated). (D-F) Cluster heat map of DE lncRNAs (D), DE miRNAs (E), and DE mRNAs (F) (red, up-regulated; blue, down-regulated).

Full size image

ceRNA regulatory network construction and functional analysis

Subsequently, to elucidate the fundamental mechanism and interaction of DEGs, we established a ceRNA network including lncRNA-miRNA-mRNA. A comprehensive ceRNA regulation network consisting of 16 miRNAs, 123 lncRNAs (61 downregulated and 62 upregulated), and 31 mRNAs was constructed, as seen in Fig. 3A and Supplementary Table S8.

Fig. 3
figure 3

(A) Construction of the lncRNA-miRNA-mRNA regulatory network based on DE-mRNAs, DE-lncRNAs and DE-miRNAs. Red and blue represent up-regulated and down-regulated RNAs, respectively. Diamonds, triangles, and ellipses represent miRNAs, mRNAs, and lncRNAs, respectively. (B) Construction of the hub ceRNA regulatory network, including lncRNA MSTRG.13,871, 9 miRNAs and 15 mRNAs. The blue diamonds around lncRNA MSTRG.13,871 represent miRNA; the yellow triangles represent mRNA. (C) Classification of top 10 significant GO terms for hub ceRNA network. The results of enrichment cover three sets: biological process (BP), cellular component (CC), and molecular function (MF). (D) Significant KEGG pathways for hub ceRNA network.

Full size image

By Cytohub, crucial RNA components were identified as significant node molecules within the network. Figure 3B illustrates that lncRNA MSTRG.13,871 is the sole lncRNA present and plays a crucial role in the formation of the regulatory network. The network consisted of a substructure of 9 miRNAs and 15 mRNAs, centered around the critical lncRNA MSTRG.13,871 (see Supplementary Fig. S3 online). Based on expression profiles, we conducted a co-expression network analysis. The analysis confirmed that the correlations between MSTRG 13,871, miR-155-5p, and mRNA conformed to our expectations (see Supplementary Fig. S4 online). In addition, KEGG and GO functional studies were conducted to comprehend the roles of the DE mRNAs in the created ceRNA network. The findings of GO research revealed that biological processes (BP) such as cellular response to unfolded protein, epidermal growth factor receptor signaling pathway, and cellular response to topologically incorrect protein were significantly enriched. Subsequently, photoreceptor outer segment membrane, apical plasma membrane and nuclear nucleosome were significantly enriched in cellular component (CC). The molecular function (MF) includes dystroglycan binding, CARD ___domain binding and ATP-dependent microtubule motor activity, plus-end-directed (Fig. 3C and Supplementary Table S9). Furthermore, the study revealed considerable enrichment of KEGG pathway keywords, including Glycosaminoglycan biosynthesis, Glycosphingolipid biosynthesis, and Glycosylphosphatidylinositol (Fig. 3D and Supplementary Table S10).

The complex method of analysis was employed to meticulously delineate the lncRNA-miRNA-mRNA ceRNA network and identify their crucial components. This approach provided insights into the regulatory dynamics within the hippocampus in the context of PCABI.

Validation of ceRNA regulatory network

To confirm the reliability of the findings obtained from the bioinformatic analysis, we conducted qRT-PCR to evaluate the levels of expression of lncRNA MSTRG.13,871, miR-155-5p, and miR-27a-5p in the hippocampus tissues of rats in both the CA/CPR group and the sham group. Compared to the sham group, the expression of lncRNA MSTRG.13,871 was notably upregulated (P = 0.01) in the CA/CPR group. Additionally, the expression of miR-27a-5p (P = 0.04) and miR-155-5p (P = 0.007) exhibited a decreasing pattern in the CA/CPR group (Fig. 4A). MiR-155-5p, in particular, holds well-established significance and has been extensively studied in the context of the pathological mechanisms underlying brain injury22,23,24,25,26,27. Then we selected miR-155-5p for more investigation. Within the ceRNA network, two mRNA molecules, Agr2 and Grip1, were specifically found to interact with miR-155. Subsequent evaluations of their expression levels were performed and validated using qRT-PCR. As shown in Fig. 4A, Agr2 (P = 0.01) and Grip1 (P = 0.004) expression patterns were consistent with the sequencing findings. Based on these findings, further research was initiated to explore the interaction mechanism within the regulatory axis MSTRG.13,871-miR-155-5p-Grip1, which may play a role in the progression of the illness.

Fig. 4
figure 4

Validation of ceRNA regulatory network. (A) The relative expression level of lncRNA, miRNAs, and mRNAs involved in the ceRNA network using qRT-PCR. (B) Target relationship prediction between miR-155-5p and MSTRG.13,871 as well as Grip1; (C, D) Validation of the target relationships between miR-155-5p and MSTRG.13,871 as well as Grip1 using dual-luciferase assay. *p < 0.05, **P < 0.01, ****P < 0.0001.

Full size image

Subsequently, we investigated the impact of MSTRG.13,871 on Grip1 expression mediated by miR-155-5p. Alignment analysis revealed potential miR-155-5p binding site within the MSTRG.13,871 sequence, as shown in Fig. 4B. To test this interaction, MSTRG.13,871 wild type containing s the miR-155-5p binding site, and its mutant counterpart were cloned into luciferase reporter vectors. These vectors were then introduced into 293 T cells using transfection. The luciferase activity exhibited a reduction in cells that were co-transfected with the wild-type binding site vector in the presence of the miR-155-5p mimic. Nevertheless, the luciferase activity did not exhibit a reduction in cells that contained the mutant binding site vector. The findings indicate that MSTRG.13,871 has the capability to target miR-155-5p (Fig. 4C). The miR-155-5p sequence was aligned with the Grip1 sequence to establish that the 3’UTR region of Grip1 contains a binding site for miR-155-5p. Figure 4D demonstrated that Grip1 was a specific recipient of miR-155-5p. Collectively, these discoveries provide insight into the regulatory interactions involving mRNA, miRNA, and lncRNA in the context of brain damage following CA. Specifically, they emphasize the critical role of lncRNA MSTRG.13,871 and its potential interactions with miR-155-5p and Grip1.

Discussion

Within the field of neuroscience, comprehending the intricacies of central nervous system damage and subsequent recovery methods remains a significant challenge. More specifically, there is considerable interest in exploring the consequences of PCABI because of its important implications for clinical practice. The objective of this work was to reproduce brain tissue damage following resuscitation in the rat model of CA/CPR. Furthermore, our objective was to identify DEGs associated with these conditions. Based on our investigations, we have identified a detailed ceRNA network comprising 26 DERNAs. Additionally, we have uncovered an important ceRNA link involving lncRNA MSTRG.18,371. Further qRT-PCR validation confirmed the regulatory role of MSTRG.13,871-miR-155-5p-Grip1 axis.Additionally, luciferase assays showed a direct binding association between these molecules.

Recently, an increasing trend was noted in the amount of data highlighting the varied functions of lncRNAs in different disorders, including central nervous system damage6,8,28. LncRNAs, which are more than 200 nucleotides long, are often compared to mRNA-like transcripts and can operate as ceRNAs. They may sequester target miRNAs to counterbalance the activity produced by miRNAs29. One lncRNA, MSTRG.13,871, has attracted little scientific attention. It is located on chr14 and spans a length of 128,873 nucleotides, namely from position 65,549,598 to 65,678,504. Our examination revealed that the GO analysis of all genes controlled by MSTRG.13,871 mostly showed enrichment in words linked to the “epidermal growth factor receptor signaling pathway” and “nuclear nucleosome“(Fig. 3C). Furthermore, the KEGG pathway analysis revealed significant enrichment in pathways related to “Glycosaminoglycan biosynthesis,” “Glycosphingolipid biosynthesis,” and “Glycosylphosphatidylinositol” (Fig. 3D).

Due to the low energy reserves of neurons, they are very susceptible to ischemia, which causes the interruption of aerobic metabolism and induces the depletion of adenosine triphosphate (ATP)3. This emphasizes the crucial need of providing additional energy to guarantee the survival of cells. Furthermore, the following activation of lytic enzymes that are Ca++-dependent exacerbates the damage to neurons5. Additionally, the malfunction of mitochondria that are Ca++-dependent leads to a loss in cellular energy, the release of proteins that promote cell death, and the production of reactive oxygen species30,31,32. These factors further worsen the damage to neurons.

These data indicate that MSTRG.13,871 may exert their protective effects on neuronal cells by increasing energy supply and demonstrating anti-apoptotic properties. Significantly, the enhancement of its central sub-networks consistently emphasizes associations with preventing cell death andpreserving mitochondrial function. The results are consistent with the observations made in rat experiments and were further confirmed using qRT-PCR analysis, which indicated an increase inMSTRG.13,871 expression in rats undergoing CPR.

NcRNAs, specifically miRNAs, being crucial regulators of gene function in several process, are proven to be valuable biomarkers and therapeutic targets in various diseases7. In addition to their interaction with lncRNAs, miRNAs have crucial functions in post-transcriptional regulation across several biological processes. Out of them, miR-155-5p is particularly notable as a versatile regulator that is crucial for maintaining cellular balance. The complex characteristics of miRNAs, which may affect many genes, indicate their significant effect on phenotypes, which is greatly impacted by the fundamental genetic context33.

Research highlighting the role of miR-155-5p in various human disorders has prompted focused studies on brain injury22,23,24,25,26,27. Clinical investigations involving patients with temporal lobe epilepsy post-surgery have demonstrated an increased expression of miR-155-5p, correlating with the presence of pathological symptoms. Conversely, in a mouse model, administering an antagomir targeting miR-155-5p led to a reduction in related pathological features, including decreased apoptosis in the hippocampus34.

The significance of miR-155-5p is further highlighted by analyzing models that include middle cerebral artery occlusion/reperfusion (MCAO/R) in rats and oxygen-glucose deprivation/reoxygenation (OGD/R) in SH-SY5Y cells. The increased expression of miR-155-5p in these animals correlated with heightened cellular damage and inflammation35. Knockdown experiments have shown that miR-155-5p plays a contributory role in neuronal and brain damage under ischemic conditions, as evidenced by its mitigating effects on these outcomes35.

The increased expression of miR-155-5p in extracellular vesicles (EVs) produced from choroid plexus epithelial (CPE) cells played a crucial role in transporting it to neurons36,37. This transfer resulted in reduced neuronal viability, increased apoptosis, and elevated levels of autophagic proteins following neuronal damage induced by oxygen-glucose deprivation (OGD). Moreover, the presence of miR-155-5p in EVs from CPEs played a detrimental role in promoting inflammatory brain injury by regulating the expression of Rheb and activating NLRP3-mediated inflammasomes36. Further examinations indicated that injecting a miR-155-5p mimic led to increased astrocyte proliferation and reduced apoptosis, with the magnitude of these effects being dosage-dependent25. The extensive participation of miR-155-5p in several types of brain damage highlights its complex function as a potential target for therapy, showing potential in reducing the degenerative processes that underlie these disorders.

The regulatory impact of miRNAs on gene expression networks is substantial, affecting either individual genes across multiple pathways or numerous genes within a single pathway. MiR-155-5p has been identified as a key regulator, targeting a diverse array of nuclear binding proteins, protein receptors, kinases, and genes involved in transcriptional regulation33. In our hub ceRNA network, only Grip1 and Agr2 mRNA molecules were directly targeted by miR-155-5p.Glutamate receptors are hypothesized to play critical roles in numerous neurological diseases, attracting considerable attention due to their significant impact on excitatory synaptic transmission, neuronal growth, excitotoxicity, and synaptic plasticity38,39. AMPA receptors (AMPARs) are crucial for facilitating fast excitatory synaptic transmission. AMPARs exist as heteromeric complexes, composed of various combinations of four similar subunits (GluR1-GluR4), which form different subtypes of AMPARs40. As the primary glutamate receptors in the brain, AMPARs are essential for controlling fast excitatory communication between neurons40.

Multiple auxiliary proteins have been found within the ___domain of AMPARs. These proteins regulate the trafficking and stability of AMPARs at the dendritic membrane. Significantly, Grip1 is distinguished by its possession of 7 PDZ domains and its ability to form connections with GluA2 and GluA3 AMPAR subunits41. Grip1 is linked to class B of the Eph/ephrin signaling family and the motor protein KIF5, enabling specific targeting of Grip1 and its various payloads to the dendritic compartment. Grip1 has been extensively studied in relation to various cerebral functions42,43. It plays a critical role in regulating synaptic plasticity, which is essential for learning and memory processes44. Additionally, Grip1 contributes to the maintenance of mushroom spines during denervation-induced homeostatic plasticity and influences both excitatory and inhibitory synapses45. Its effects also extend to the early stages of brain development.The involvement of Grip1 in the trafficking of AMPARs has been a topic of discussion in the literature. While several research indicate that Grip1 has a role in transporting AMPARs to the surface of cells and keeping them stable at synapses46,47,48, other conflicting data imply that it may actually be involved in keeping AMPARs within the cell49,50. Recent data indicates the presence of multiple Grip1 pools, each selectively managing the trafficking of AMPARs. Notably, Grip1’s primary function involves facilitating the delivery of AMPARs to the cell surface and stabilizing them at synapses43.Relevant research suggests that elevated levels of glutamate induce the phosphorylation of GluA2 subunits, which subsequently leads to the internalization of AMPA receptors (AMPARs) containing GluA2. The absence of GluA2 enhances the influx of Ca2 + ions, ultimately resulting in cell death due to excitotoxicity51,52. Consequently, the regulation of surface AMPARs demonstrates potential in preventing cell death following hypoxic-ischemic injury53. Therefore, we propose that the neuroprotective effect associated with increased Grip1 expression in brain damage may stem from its role in promoting AMPAR localization to the cell membrane surface. Furthermore, the elevation of Grip1 levels under pathological conditions appears to be influenced by regulatory mechanisms within the ceRNA network.Our study presents certain limitations that warrant further exploration. Initially, we identified regulatory relationships governed by ceRNAs, yet the direct impact of these relationships on cellular behaviors has not been fully investigated. Although there is extensive research on miRNA-155-5p, the specific mechanisms through which lncRNAs and the Grip1 gene provide cellular protection require further elucidation. We have formulated mechanistic hypotheses based on existing literature; however, the complex regulatory dynamics between Grip1 and AMPAR, as well as the detailed mechanisms by which Grip1 influences AMPAR’s subcellular localization and stability, remain insufficiently explored. Currently, our team is dedicated to investigating these complex and potentially significant mechanisms.

Conclusion

In summary, we have successfully established a unique ceRNA network in a rat model of PCABI. Utilizing this network, we identified key ceRNAs regulated by lncRNA and validated their regulatory interactions through qRT-PCR and luciferase assays. Our findings specifically highlighted the critical regulatory pathway involving MSTRG.13,871-miR-155-5p-Grip1. Additionally, GO and KEGG analyses indicated that the hub ceRNA networks are primarily associated with mitochondrial functions and apoptotic processes. Our research provides novel insights into the pathogenesis of PCABI and lays the groundwork for further investigation into its regulatory mechanisms.