Effects of Celecoxib and PGE2 on SYP/PSD95 and cAMP/PKA/CREB/BDNF signaling in N2a cells.
To identify the effects COX-2 and PGE2 in vitro, we treated N2a cells with COX-2 selective inhibitor Celecoxib. As shown in RT-qPCR result, the inhibition ability of Celecoxib for COX-2 was dose-dependent, while 10μm was the most efficient among 1μm, 5μm, 10μm and 20μm these four concentration gradients (Fig. 7A). After N2a cells were treated with 10μm Celecoxib for 24h, evidently, we found not only the expression of COX-2 and PSD95 protein decreased but also cAMP and its downstream signal pathway p-PKA/PKA, p-CREB/CREB, BDNF all downregulated. But no change was found in the SYP expression after the treatment of Celecoxib. While all of these cytokines significantly increased after the treatment of PGE2 (1μmol/L) for 4 hours (Fig. 7B, C). These results suggest that Celecoxib has the effect of reducing PSD95 and cAMP/PKA/CREB/BDNF signaling by inhibiting the releasing of PGE2 in N2a cell lines.
DISCUSSION
Evidence suggests the constitutive expression of ”inducible” COX-2 in the brain[25], kidney[26, 27], and a few other organs[28], with basal activity involvement in cellular functions. Previous studies have demonstrated that neuronal COX-2 and its metabolites are essential in memory and learning. For the first time, Yamagata established that hippocampal neuronal COX-2 correlates with neural activity and memory consolidation in rats [29]. COX-2 is also implicated in various cognitive functions, such as acquiring[30], consolidating [31], and passively avoiding [32] memories. Celecoxib is a selective COX-2 inhibitor and could impair the spatial memory of mice due to infusion into the hippocampal CA1 region [33, 34] without any effect on COX-1 inhibition[21]. Our data also demonstrated the importance of COX-2 in mice’s memory, consistent with these findings.
Many studies have examined the molecular mechanisms associated with synaptic modifications during memory development and mediated by various proteins [35-38]. Furthermore, COX-2 is enriched in neuronal dendritic spines with active synapses[39,40]. Synapses form the structural basis for neuronal communication and facilitate memory functions. PSD95 and SYP regulate the structure and function of synapses[41]. PSD95, a significantly available scaffold protein in the postsynaptic region, helps in the signal transmission of neurons [42]. A complete PSD95 knockout in mice causes severe deficits in learning and memory of spatial areas, establishing the significance of PSD95 in developing synapses[43, 44]. Therefore, COX-2 signaling plays a crucial role in the postsynaptic region, regulating PSD95 expression and contributing to memory formation.
Learning and memory are associated with activity-induced changes in dendritic spine density, maturity, size, and shape[45]. New memories are directly related to the density of dendritic spines inside the CA1 region of the hippocampal structure [46-49]. Dendritic spines mature into mushroom spines to achieve spatial working memory[45]. Decreased spine density or maturity is correlated with mental retardation, memory loss, and Alzheimer’s disease[50-52]. Dendritic tree structure and arborization provide insight into how neurons receive and transmit synaptic inputs[53]. The dendritic diameter in the cerebellum of C57BL/6 male mice positively correlated with the total postsynaptic density (PSD). Moreover, a greater diameter is associated with a greater PSD area [54]. Dendrite thickness regulates postsynaptic potential transmission by affecting electrical resistance, with faster signal propagation among thicker dendrites[55, 56]. Evidence indicates COX-2/PGE2 signaling alters dendrite and dendritic spine morphology [57]. This study supported the importance of dendritic spine density and maturity in memory formation. The study also significantly reduced dendritic spine density in COX-2KO mice. Thus, OX-2 was crucial in mediating memory-related neuronal structure in the hippocampal CA1 region.
Local population activities or the assembly of neurons across multiple brain regions cause a rhythmic fluctuation known as neural oscillation. It is detected using electrocorticography (ECoG), local field potentials (LFP), magnetoencephalography (MEG), and electroencephalography (EEG) under various frequencies. These include delta (0.5–3 Hz), theta (4–12 Hz), alpha (13–15 Hz), beta (16–30 Hz), low gamma (30–60 Hz), and high gamma(60-120 Hz) [58-60]. They regulate neuronal spike timing on the microscale and coordinate macroscale data dispersed in the cortical region to provide information concerning temporospatial localizations [61]. Group coordination is necessary for effective communication between presynaptic and postsynaptic groups [62]. Selective COX-2 inhibitors significantly reduced LTP induction in hippocampal dentate granule neurons in vitro . In contrast, COX-1 inhibitors were ineffective, reversed by PGE2 treatment, but not by other PGs [63]. Additionally, activating neuronal COX-2 was necessary to induce LTP in learningin vivo [21]. Gamma oscillations have fast rhythms to escape inhibition through network excitation for the time being. This elevates the effectiveness, overall precision, and selection of communication among various network areas [64]. The changing gamma oscillation power could be a disease biomarker to investigate novel therapies for treating cognitive impairment in diseases such as Alzheimer’s [65, 66]. A strong correlation was observed between the in vivo gamma oscillation power and the cognitive functions of the test animals[67]. The present study depicts that COX-2KO in mice effectively decreases gamma oscillation power. Thus, COX-2 could affect learning and memory ability by modulating gamma oscillations in the hippocampal CA1 area.
Our findings indicate that only PGE2 expression significantly decreased in COX-2KO mice among all the five PG types. One possible explanation for the altered PG profile would be that COX-2 predominantly metabolizes AA to PGE2. On the other hand, COX-1 could be responsible for other PGs [68]. PGE2 is characterized as the lipid mediator, enabling nervous system development. Meanwhile, various studies demonstrate that abnormal PGE2 signaling leads to neurodevelopmental disorders [69, 70]. The COX-2/PGE2signaling pathway is associated with several vital processes in developing neurons. This includes proliferation, migration[71, 72], synaptogenesis [73, 74], and the formation and plasticity of dendritic spines[72, 75]. Some rat studies have revealed that uncontrolled elevation or reduction of PGE2 expression affects cerebellar [76]and hippocampal[77] dendritic arborization during critical periods after delivery. In contrast, inhibiting PGE2production by locally injecting COX-2 inhibitors can reduce the arborization of the dendritic tree and the density of excited synapses[78]. As depicted in the COX-2KO mouse model[79-81], many developing genes have differential expression due to the reduced PGE2 levels. This disrupts synaptic transmission pathways, long-term nerve potentiation, navigating the neurites, and developing the dendritic spine [82, 83], consistent with our results.
PKA is the primary downstream kinase of cAMP and could phosphorylate CREB. Thus, parts of the transcription factor are recruited within the promoter regions, including the BDNFgene [84]. The cAMP/PKA/CREB/BDNF signaling pathway promotes the survival of neurons, modulates synaptic morphology, and strengthens synaptic transmission[85]. The pathway modulates survival, development, cognition, and mood across epilepsy, depression, and Alzheimer’s disease animal model neurons [86, 87]. Our study observed lower phosphorylation expression in the COX-2KO group of cAMP, PKA, and CREB with BDNF expression than in WT. Thus, COX-2 could participate in the synaptic plasticity modulation of hippocampal neurons using cAMP/PKA/CREB/BDNF signaling.
Our findings revealed the COX-2 signaling pathway as a new mechanism for regulating the hippocampal dendritogenic process and gamma oscillation to affect learning and memory ability. The exogenous application of PGE2 effectively reversed the actions caused by the inhibition of COX-2. COX-2 could participate in hippocampal synaptic plasticity by regulating COX-2/PGE2/cAMP/PKA/CREB/BDNF signaling in the postsynaptic membrane PSD95 modulation. Therefore, this study can elevate our understanding of COX-2 physiologic function and its role in memory formation.