The long non-coding RNA NEAT1 is responsive to neuronal activity and is associated with hyperexcitability states

Abstract

Despite their abundance, the molecular functions of long non-coding RNAs in mammalian nervous systems remain poorly understood. Here we show that the long non-coding RNA, NEAT1, directly modulates neuronal excitability and is associated with pathological seizure states. Specifically, NEAT1 is dynamically regulated by neuronal activity in vitro and in vivo, binds epilepsy-associated potassium channel-interacting proteins including KCNAB2 and KCNIP1, and induces a neuronal hyper-potentiation phenotype in iPSC-derived human cortical neurons following antisense oligonucleotide knockdown. Next generation sequencing reveals a strong association of NEAT1 with increased ion channel gene expression upon activation of iPSC-derived neurons following NEAT1 knockdown. Furthermore, we show that while NEAT1 is acutely down-regulated in response to neuronal activity, repeated stimulation results in NEAT1 becoming chronically unresponsive in independent in vivo rat model systems relevant to temporal lobe epilepsy. We extended previous studies showing increased NEAT1 expression in resected cortical tissue from high spiking regions of patients suffering from intractable seizures. Our results indicate a role for NEAT1 in modulating human neuronal activity and suggest a novel mechanistic link between an activity-dependent long non-coding RNA and epilepsy.

Figure 1. NEAT1 binds the potassium channel-interacting protein KCNAB2 that is enriched in the cytoplasm upon neuronal activation.

(a) Protein microarray reveal potential NEAT1 interacting proteins with KCNAB2 circled as a high confidence hit. (b) Table representing the highest confidence hits includes KCNAB2 together with other potassium channel-interacting proteins KCNAB1 and KCNIP1 (arrows). (c) KCNAB2 binds directly to NEAT1 as determined via RNA immunoprecipitation in the neuroblastoma cell-line SH-SY5Y. (d) Activation of SH-SY5Y cells with 50 mM KCl results in a significant increase of KCNAB2 protein in the cytoplasm after 3 hours using western blot analysis with a return to baseline after 10 hours. (n ≥ 3, *p value < 0.05; One-way ANOVA test with a Tukey’s multiple comparison post hoc test) (e) Immunohistochemistry in SH-SY5Y cells shows that KCNAB2 (red) is mostly nuclear localized (see arrows) with few cells showing KCNAB2 cytoplasmic staining (see arrowheads). (f) In contrast, 3 hours post KCl activation, KCNAB2 protein is mostly cytoplasmic (see arrowheads).

Figure 2. NEAT1 is down-regulated following neuronal activity and modulates neuronal excitability.

(a) NEAT1 expression increases during cortical-type neuronal differentiation. (b) NEAT1 is acutely and transiently down-regulated following depolarization in human induced pluripotent stem cell (iPSC)-derived neurons (n ≥ 3; One-way ANOVA test with a Tukey’s multiple comparison post hoc test; *p value < 0.05; **p value < 0.01). (c) Antisense oligonucleotides (ASOs) directed at NEAT1 successfully reduced NEAT1 transcript by around 80% and knockdown did not affect the ability of neuronal activation to significantly reduce residual NEAT1 levels. (n = 3; unpaired Student’s t-test; **p value < 0.01; ***p value < 0.001). (d) Antisense oligonucleotide (ASO)-mediated down-regulation of NEAT1 results in enhanced depolarization-induced calcium influx in iPSC-derived neurons. (n ≥ 300 independent wells; One-way ANOVA test with a Tukey’s multiple comparison post hoc test; ****p value < 0.0001).

Figure 3. NEAT1 is associated with ion channel function.

(a) Deep sequencing was performed on iPSC-derived neurons that were subjected to ASO-mediated NEAT1 knockdown followed by KCl-induced activation. Leading edge gene analysis revealed a strong correlation between NEAT1 expression and significantly altered genes with ion channel classifications. Specific genes included in significant gene sets are shown in Supplementary Table 3. (b) Gene set enrichment analysis (GSEA) of activated control ASO-treated iPSC-derived neurons demonstrated that endogenous levels of NEAT1 transcript are inversely correlated with an overall decreased expression of genes contained in ion channel gene sets relative to controls in activated neurons. The x-axis contains all the genes in the particular gene set from up-regulated (red) to down-regulated (blue). False discovery rates (FDR) and enrichment scores (ES) evaluate whether the genes in the gene set are enriched as up- or down-regulated (c) Conversely, ASO-mediated NEAT1-knockdown is sufficient to drive an overall elevated expression of genes contained in ion channel gene sets relative to controls in activated neurons.

Figure 4. NEAT1 expression is up-regulated in human epilepsy samples and in chronically stimulated in vivo rat models.

(a) Regions were resected that displayed aberrant high or low activity in the cerebral cortex of epilepsy patients. (b) Quantitative PCR performed on resected human epilepsy patient tissue samples showed a significant increase in NEAT1 transcript levels between high and low activity regions (n = 14, Student’s t-test; **p value < 0.01). (c) NEAT1 is acutely down-regulated but chronically insensitive in an in vivo rat model of pilocarpine-induced seizures (n ≥ 3, **p value < 0.01, ***p value < 0.001; One-way ANOVA test with a Tukey’s multiple comparison post hoc test). (d) Similarly, NEAT1 is acutely down-regulated but chronically insensitive in an in vivo post-KA induced Status Epilepticus (SE) model. (n = 3, *p value < 0.05; One-way ANOVA test with a Tukey’s multiple comparison post hoc test).