MicroRNA-153-3p targets repressor element 1-silencing transcription factor (REST) and neuronal differentiation: Implications for Alzheimer's disease

Abstract

Introduction: Small non-coding microRNAs (miRNAs) play essential roles in Alzheimer's disease (AD) pathogenesis. Repressor element 1-silencing transcription factor (REST) is involved in AD, though its regulation remains unclear.

Methods: We performed real-time quantitative polymerase chain reaction (qPCR) in autopsied brain tissues to determine miR-153-3p and AD associations. A reporter-based assay measured the activity of REST mRNA 3'-untranslated region (3'-UTR). Induced pluripotent stem cells (iPSC)-derived neurons and human cell lines were applied to determine miR-153-3p regulation of endogenous proteins.

Results: Elevation of miR-153-3p is associated with a reduced probability of AD, while elevated REST is associated with a greater probability of AD. The 3'-UTR functional assay pinpointed the miR-153-3p binding sites. miR-153-3p treatment reduced REST, amyloid precursor protein (APP), and α-synuclein (SNCA) 3'-UTR activities and protein levels. miR-153-3p treatment altered REST and neuronal differentiation in iPSC-derived neuronal stem cells. RNA-sequencing and proteomics revealed miR-153-3p-associated networks.

Discussion: miR-153-3p reduces REST, APP, and SNCA expression, pointing toward its therapeutic and biomarker potential in neurodegenerative diseases.

Highlights: With the increased emphasis on comorbidities of Alzheimer's disease (AD) and other neurodegenerative diseases, we identified that miR-153-3p, as a master regulator, reduced a group of neurodegeneration related proteins: REST, amyloid precursor protein (APP) and α-synuclein (SNCA) levels. The elevation of miR-153-3p levels is associated with reduced probability of AD in posterior cingulate cortex (PCC), while REST, by contrast, is associated with a greater probability of AD. miR-153-3p treatment alters REST protein levels and neuronal differentiation in induced pluripotent stem cells (iPSC) derived neuronal cells. RNA sequencing proteomics and interactome analysis revealed the role of miR-153-3p in axonal guidance.

Keywords: REST; UTR activity; amyloid; biomarker; dementia; drug target; non‐coding.

FIGURE 1

The probability of AD increases as REST mRNA levels increase. The likelihood of AD is reduced as levels of miR‐153‐3p increase, but REST mRNA levels partially reverse this effect. RNA was extracted from human brain samples and subjected to qRT‐PCR as described in the text. Ordinal logistic regression models were built to test the effects of REST mRNA and of miR‐153‐3p on the probability of NCI, MCI, and AD. Ignoring potential effects of miR‐153‐3p, elevated REST mRNA levels were associated with greater probability of AD and a reduction in NCI. If miR‐153‐3p and APP RNA levels were taken into account, the miRNA effect was stronger than either REST or APP effects, but each had significant contributions in the model. (A–C) REST RNA only, ages shown are 17th percentile (%ile), median, and 83rd %ile. (D–I) Models with miR‐153‐3p. APP mRNA is shown at the 17th, median, and 83rd %iles while REST is shown at its 25th and 75th %ile. AD, Alzheimer's disease; APP, amyloid‐β precursor protein; MCI, mild cognitive impairment; NCI, no cognitive impairment; qRT‐PCR, quantitative real‐time polymerase chain reaction; REST, repressor element 1‐silencing transcription factor.

FIGURE 2

Predicted miR‐153‐3p sites in REST 3’‐UTR and effects of miR‐153‐3p mimics on REST 3’‐UTR activity. The REST 3’‐UTR (NM005612) was compared with the miR‐153‐3p sequence via the StarMir utility. Four putative sites were found. Multiz alignments with 11 other mammalian species are shown. (A) REST mRNA sequence. (B) Site between 4262‐4290 (starting from transcription start site). (C) Site between 4328‐4354. (D) Site between 4458‐4490. (E) Site between 4512‐4534. (F) Structure of Gluc/seAP vector with REST 3’‐UTR. (G) Transfection of empty GLuc/seAP vector or vector with REST 3’‐UTR, co‐transfected with mock, miR‐153‐3p, miR‐153‐3p antagomiR, miR‐153‐3p + antagomiR, REST siRNA, and NCM (negative control), in HeLa cells. miR‐153‐3p drove reduced expression when REST 3’‐UTR was present. (H) Time course of effects of miR‐153‐3p or miR‐153‐3p antagomir on GLuc/seAP ratio in HeLa cells. Over a period of 36‐96 h, a single transfection of miR‐153‐3p significantly reduced GLuc/seAP ratio, while a single transfection of the miR‐153‐3p antagomiR significantly increased such signal. The interval from 36‐96 h was modeled as a repeat‐measure ANOVA. (I) REST 3’‐UTR was mutated as shown. (J) Transfection of empty GLuc/seAP vector and vector with native or mutant REST 3’‐UTR, co‐transfected with mock, miR‐153‐3p, miR‐153‐3p antagomiR, and NCM (negative control miRNA), in HeLa cells. Symbols indicate differing statistical groups at p ≤ 0.05. Samples sharing symbols did not significantly differ. ANOVA, analysis of variance; REST, repressor element 1‐silencing transcription factor; UTR, untranslated region.

FIGURE 3

Predicted miR‐153‐3p sites in APP and SNCA 3’‐UTR as well as effects of miR‐153‐3p mimics on APP and SNCA 3'‐UTR activity. (A) SNCA (NM001375287) mRNA sequence. (B) Site between 1367‐1402 (starting from transcription start site). (C) Site between 1688‐1719. (D) Transfection of empty GLuc/seAP vector or vector with SNCA 3’‐UTR, co‐transfected with mock, miR‐153‐3p, miR‐153‐3p antagomiR, miR‐153‐3p + antagomiR, SNCA siRNA, and miR‐216 (alleged negative control), in HeLa cells. miR‐153‐3p drove reduced expression when SNCA 3’‐UTR was present. (E) Time course of effects of miR‐153‐3p or miR‐153‐3p antagomir on GLuc/seAP ratio in HeLa cells. Over a period of 48–120 h, a single transfection of miR‐153‐3p significantly reduced GLuc/seAP ratio. Symbols indicate differing statistical groups at p ≤ 0.05. Samples sharing symbols did not significantly differ. (F) APP (NM000484) mRNA sequence. (G) Site between 2905–2928 (starting from transcription start site). (H) Transfection of empty GLuc/seAP vector or vector with APP 3’‐UTR, co‐transfected with mock, miR‐153‐3p, miR‐153‐3p antagomiR, miR‐153‐3p + antagomiR, APP siRNA, and NCM (negative control), in HeLa cells. miR‐153‐3p drove reduced expression when APP 3’‐UTR was present. (I) Time course of effects of miR‐153‐3p or miR‐153‐3p antagomir on GLuc/seAP ratio in HeLa cells. Over a period of 48–120 h, a single transfection of miR‐153‐3p significantly reduced GLuc/seAP ratio Symbols indicate differing statistical groups at p ≤ 0.05. Samples sharing symbols did not significantly differ. APP, amyloid‐β precursor protein; SNCA, α‐synuclein; UTR, untranslated region.

FIGURE 4

Effects of miR‐153‐3p treatment in HeLa and differentiated neuronal cells on REST, APP, and SNCA. Cultures of (A) HeLa and (B) Neuronal cells were treated with miR‐153‐3p, antagomiR of miR‐153‐3p, miR‐153‐3p plus antagomiR, siRNA versus SNCA, and miR‐216 (alleged negative control), and levels of SNCA, REST, and APP were measured. Each lane was loaded with cell lysate from an independent transfection. Densitometry revealed that in (C) HeLa and (D) Neuronal cells, miR‐153‐3p consistently reduced SNCA, REST, and APP protein signal (adjusted by β‐actin signal). Of potential note, treatment with miR‐216 increased levels of SNCA versus mock treatment. Symbols indicate differing statistical groups at p ≤ 0.05. Samples sharing symbols did not significantly differ. APP, amyloid‐β precursor protein; REST, repressor element 1‐silencing transcription factor; SNCA, α‐synuclein.

FIGURE 5

Effects of miR‐153‐3p treatment in iPSC derived neural stem cells on REST, nestin and doublecortin as well as variation of effect by iPSC neuronal differentiation into npc, neurons, and astrocytes. (A) Western blot of iPSC‐derived neural stem cell transfected with miR‐153‐3p, REST siRNA and other oligomers as described in the main text. Each lane represents an independent transfection. (B) Western blot of REST, NSE, doublecortin, and β‐actin proteins during differentiation of an iPSC cell line derived from a familial AD patient. Each lane represents an independent differentiation from one iPSC cell line. (C) Densitometry of REST, nestin and doublecortin proteins, as well as levels of CTG (measurement of cell viability). Treatment with miR‐153‐3p and REST siRNA was significantly associated with reductions in REST, nestin protein, and CTG. Treatment with miR‐153‐3p but not REST siRNA increased doublecortin protein. (D) Densitometric quantification of REST, NSE, and doublecortin proteins. (E) REST mRNA levels of four distinct stages of cells; endogenous miR‐153‐3p levels in iPSC, npc, neurons, and astrocytes. Symbols indicate differing statistical groups at p ≤ 0.05. Samples sharing symbols did not significantly differ. iPSC, induced pluripotent stem cell; npc, neural progenitor cells; NSE, neuron‐specific enolase; REST, repressor element 1‐silencing transcription factor.

FIGURE 6

RNA‐sequencing analysis of miR‐153‐3p transfection in neuronal cells. Differentiated human neuroblastoma cells were transfected with miR‐153‐3p or miR‐153‐3p antigomiR. mRNA samples were subjected to RNA‐seq assay followed by pathway analysis. (A) Count of differentially expressed genes. (B) Kegg pathway enrichment, mock versus miR‐153‐3p. (C) Volcano plot of differentially expressed genes.

FIGURE 7

Proteomic analysis of miR‐153‐3p transfection in neuronal cells. (A) Proteomic analysis was performed, and data filtered as described in 2.13. Red points represent proteins with miR‐153‐3p ÷ mock abundance < 1 and antagomR ÷ mock abundance > 1. Blue points represent proteins with miR‐153‐3p ÷ mock abundance > 1 and antagomR ÷ mock abundance < 1. Solid points have log₂ abundance absolute differences of ≥ 0.2 and p ≤ 0.05. (B–C) We generated networks from proteomic abundance data after treating cells with miR‐153‐3p and antagomir to miR‐153‐3p. (B) Network in hippocampus. (C) Network in cerebral cortex.

FIGURE 8

Roles of miR‐153‐3p in connection to neurodegeneration and AD. Based on our own work and others known to the field, we propose a mechanistic cascade to explain the roles of miR‐153‐3p in AD. In this model, miR‐153‐3p would serve to target the RISC (among others) to REST, APP, and SNCA mRNA 3’‐UTR sequences. This blockade would reduce levels of APP and resultant Aβ and levels of SNCA. Blockade of REST would influence neuronal differentiation. Dysregulation of miR‐153‐3p would perturb normal regulation, resulting in CNS dysfunction. Figure created with BioRender.com. AD, Alzheimer's disease; APP, amyloid precursor protein; CNS, central nervous system; REST, repressor element 1‐silencing transcription factor; SNCA, α‐synuclein; 3’‐UTR, 3′‐untranslated region.