lncRNAs: key player in Aβ deposition

ABSTRACT

Alzheimer’s disease (AD) is a typical neurodegenerative disorder, characterized by the deposition of β-amyloid (Aβ) plaques. β- and γ-secretases generate Aβ by cleaving amyloid precursor protein. The imbalance between its production and clearance leads to Aβ accumulation, causing neuronal damage through mechanisms such as inducing oxidative stress and inflammatory responses. Long non-coding RNAs (LncRNAs), composed of more than 200 nucleotides, usually do not encode proteins and are involved in processes such as gene expression regulation, chromatin remodelling, and cell cycle control. Studies have shown that LncRNAs play a key role in brain development and the maintenance of neuronal function, especially by influencing Aβ deposition to affect the progression of AD. This review summarizes the pathways by which LncRNAs affect Aβ deposition, classifies them according to their modes of action, discusses the existing problems in current research, and summarizes and prospects their role in the treatment of AD.

1. Introduction

1.1. The incidence and course of Alzheimer’s disease

Alzheimer’s disease (AD) is a typical neurodegenerative disease. In 1907, German psychiatrist and neuropathologist Alois Alzheimer first described the disease, so the disease was named after him [1]. Every three seconds, someone in the world is diagnosed with AD [2]. According to the World Health Organization, approximately 55 million people worldwide had dementia in 2019, and this number is projected to increase to 139 million by 2050,and in 2019, the annual cost of dementia was estimated at 1.3 trillion US dollars, and it is expected to more than double to 2.8 trillion US dollars by 2030. With the ageing of the world’s population, dementia is becoming one of the leading causes of death globally [2]. As the disease progresses, the pathological features of AD gradually emerge, and the early symptoms vary [1]: The most common initial symptom of patients is short-term memory impairment. As the disease progresses, the patient’s mental function declines significantly, often accompanied by cognitive difficulties [1]. In addition, patients may experience a decline in numerical ability and cognitive function, also known as impaired executive thinking ability [3]. In the late stage, patients are completely bedridden and dependent on others for care. The average course of the disease is about 5.40 years, and the disease eventually leads to death [1]. The preclinical stage of AD is characterized by amyloid deposition and early neuroinflammatory changes in the patient’s brain. As the disease progresses, elevated tau phosphorylation levels become an important pathological feature of AD, and synaptic dysfunction, synaptic loss, and neurodegeneration worsen with the progression of the disease [3]. The manifestation and severity of AD clinical symptoms can be staged using the Clinical Dementia Rating scale, where 0 points represent normal cognition, and 0.5, 1, 2, and 3 points represent suspected, mild, moderate, and severe dementia, respectively [3].

1.2. Excessive extracellular deposition of amyloid β is a key pathological feature of AD

Excessive extracellular deposition of amyloid β (Aβ) is a key pathological feature of AD: this protein comes from the amyloid precursor protein (APP), which is cleaved by β-secretase (BACE) and γ-secretase to produce Aβ; abnormal phosphorylation of Tau protein leads to neurofibrillary tangles. In addition, pathological characteristics such as loss of neuronal synapses and pyramidal neurons are also closely related to the disease [3]. However, the exact cause of AD remains difficult to determine. Based on the above pathological characteristics, researchers have proposed several hypotheses, including the Aβ hypothesis, the Tau protein hypothesis, and the cholinergic hypothesis [2–6]. Among them, excessive extracellular deposition of Aβ amyloid protein is the most important pathological feature of AD. The imbalance between Aβ production and clearance is considered to be a pathogenic factor in the progression of AD [4,5]. However, the underlying molecular mechanism that explains the imbalance between Aβ production and clearance is still unclear. β and γ secretases cleave APP to produce Aβ [6,7], which is cleared by glial cell-mediated autophagy [8,9].

1.3. The relationship between LncRNAs and deposition of Aβ

LncRNAs are non-coding nucleic acids consisting of more than 200 nucleotides. Studies have shown that changes in LncRNAs levels can affect the occurrence and progression of a variety of diseases, including lung cancer [10], breast cancer [11,12], colorectal cancer [13,14], etc. In addition, recent studies have shown that LncRNAs play a key role in brain development, neuronal function and maintenance, and their impact on neurodegenerative diseases is becoming increasingly evident, such as Huntington’s disease [15,16], Parkinson’s disease [17–19] and amyotrophic lateral sclerosis [20,21]. Notably, multiple studies have identified a close link between LncRNAs and several pathological features of AD, such as Aβ deposition [22,23], Tau protein hyperphosphorylation [24] and neuroinflammation [25,26]. LncRNAs can also serve as potential AD biomarkers [7–9,26].

Therefore, in this review, we focus on summarizing the mechanism by which LncRNAs affect the progression of AD by regulating Aβ deposition. We have classified the mechanisms discovered in current research into roles such as competitive endogenous RNAs (ceRNAs), natural antisense transcripts (NATs), and histone modification enzyme receptor that influence the excessive deposition of Aβ, and have drawn relevant pictures and made summaries. By summarizing the current application of LncRNAs in AD, we aim to provide more ideas for the future prevention and treatment of this disease.

2. LncRnas’ functions in ceRNAs of Aβ deposition

MicroRNAs (miRNAs) are important post-regulatory regulator [27]. In the classical miRNA regulatory mechanism, it binds to the 3’UTR of the target and, therefore, acts post-regulatory. In addition, a single miRNA represses multiple genes that can compete with the miRNA for binding [27]. LncRNAs is one of the competitors of various genes, so ceRNAs comes from the above concept. ceRNAs are important direction for studying RNA regulation which refers to the fact that in cells, certain non-coding RNAs (such as LncRNAs) and mRNA share the same miRNAs binding sites, thus forming a mutually competitive relationship and regulating the expression levels in various places [28]. In previous studies, researchers have repeatedly found that LncRNAs competitively bind to miRNAs (Table 1), thereby affecting the accumulation of Aβ (Figure 1).

Table 1.

LncRNAs functions as ceRNAs deposition of Aβ.

LncRNAs	Expression	Action	Target	Function	Reference
NEAT1	up	harmful	miR-107	Promoting the secretion of Aβ.	[30]
NEAT1	up	harmful	miR-124/BACE1	Promoting the secretion of Aβ.	[31]
NEAT1	up	harmful	miR-27a-3p	Promoting the secretion of Aβ.	[32,33]
NEAT1	up	harmful	has-miR-27a-3p	Changing the levels of IGF1R and related GPCRs.	[34]
NEAT1	up	harmful	miR-291a-3p	Activating the NF-κB pathway.	[37]
NEAT1	up	harmful	miR-361-3p	Activating the NF-κB pathway.	[38]
XIST	up	harmful	miR-124/BACE1	Promoting AD-related BACE1 changes.	[48]
MALAT1	down	protective	miRs-200a,-26a,-26a/RTK	Targeting downstream EPHA2 prevents Aβ-induced. Cytotoxicity	[55]
RPPH1	up	harmful	miR-330-5p/CDC42	enhancing the expression level of CDC42.	[59–61]
RPPH1	up	harmful	miR-122/RPPH1, Wnt13’UTR	Activating the Wnt/β-catenin signalling pathway.	[60]
LINC01311	up	harmful	has-miR-146a-5p	Increasing Aβ production.	[63]
BACE1-AS	up	harmful	miR-132-3p	Regulating the expression of ATG5 in a longitudinal manner.	[70]
SOX21-AS1	up	harmful	miR-107	Alleviating Aβ-induced neuronal damage.	[73]
NORAD	up	harmful	miR-26b-5p	Inhibiting Aβ clearance.	[75]

Figure 1.

LncRNAs functions as ceRnas in deposition of Aβ. This figure summarizes the roles of LncRNAs and miRnas in regulating Aβ production and deposition as described above. Both XIST and NEAT1 competitively bind to miR-124, thereby affecting the level of BACE1. LncRNAs also influence NF-κB, Wnt and other pathways, as well as RTK, IGF1R, GPCR and other receptors, affecting Aβ deposition by competitively binding to miRnas. The above interactions indicate that miRnas and LncRNAs are potential therapeutic targets for regulating Aβ levels in AD. Additionally, BC200 and 51A directly affect BACE1 and SORL1, respectively, influencing Aβ deposition.

2.1. NEAT1

Nuclear Paraspeckle Assembly Transcript 1 (NEAT1) plays a significant role in the pathogenesis of Alzheimer’s disease (AD), and its involvement in multiple cell types and time points has been extensively studied [29]. Firstly, the role of NEAT1 in neurons has been widely investigated: as early as 2019, researchers found that NEAT1 expression was enhanced in the early stage of AD, and knockdown of NEAT1 could inhibit Aβ-induced cell viability. Through luciferase activity assays and RNA immunoprecipitation (RIP) detection, miR-107 was identified as the target site of NEAT1, suggesting that NEAT1 has a protective effect on neurons in the early pathological process [30]; however, its upregulation in the later stage promotes Aβ accumulation through sponge action: luciferase reporter gene assays indicated that NEAT1 exacerbates Aβ production through the miR-124/BACE1 axis, highlighting its crucial role in the Aβ generation process [31]. Additionally, researchers confirmed the targeted relationship between LncRNAs NEAT1 and micro-27a-3p through dual luciferase reporter gene and RNA pull-down experiments: by detecting the expression of NEAT1 and miR-27a-3p in serum and cerebrospinal fluid, it was further demonstrated that NEAT1 downregulates miR-27a-3p to promote the progression of AD, emphasizing its role in neuronal damage [32,33]. In the AD model constructed in SH-SY5Y cells, NEAT1 can alter the levels of insulin-like growth factor 1 receptor (IGF1R) and related G protein-coupled receptors (GPCRs) by regulating hsa-miR-15a-5p and hsa-miR-16-5p. This is the first exploration of the regulation of IGF1R by non-coding RNA from the perspective of neurodegeneration through bioinformatics [34].

NEAT1 also plays an important role in glial cells. Transcriptome analysis indicates that overexpression of NEAT1 is associated with hAPP/Aβ-induced astrocyte dysfunction in J20 mice [35], and NEAT1 regulates ROR1 through the miR-146a axis, alleviating Aβ [1–42]-induced cytoskeletal instability, further confirming its role in glial cells [36].

In terms of therapeutic strategies, targeted regulation of NEAT1 is considered a potential intervention approach. Tanshinone IIA effectively blocks this process by inhibiting the NEAT1/miR-291a-3p/Rab22a axis and the NF-κB signalling pathway [37]. Notoginseng regulates the NEAT1/miR-361-3p/TRAF2 axis, thereby inhibiting the pathological features of decreased cell viability, increased apoptosis, and exacerbated inflammation in LPS-induced HMC3 cells, and activating the NF-κB signalling pathway [38]. Its role in improving Aβ aggregation, inflammatory response, and cognitive function further reveals the therapeutic potential of NEAT1 intervention.

In conclusion, the multi-level regulatory role of NEAT1 in neurons and glial cells plays a key role in the occurrence and progression of AD. It regulates Aβ production, neuroinflammatory responses and cytoskeletal stability by interacting with different miRNA axes, such as miR-107, miR-27a-3p, miR-124, miR-146a and miR-291a-3p. With further research on the NEAT1 regulatory network, therapeutic strategies targeting NEAT1 are expected to provide new directions for clinical intervention in AD.

2.2. XIST

X-inactive specific transcript (XIST) is a LncRNAs encoded by the XIST gene, whose main function is to be responsible for X chromosome inactivation in mammals. Previous studies have indicated that XIST may exert biological effects on gene expression by altering the stability of heterochromatin [39–41]. A large amount of current evidence suggests that XIST plays a crucial role in the differentiation, proliferation, and genomic maintenance of human cells: such as lung cancer [42], bladder cancer [43], and pancreatic cancer [44] through a mechanism involving ceRNAs. Specifically, there is connectivity between XIST and miR-124, which can regulate the occurrence and development of pancreatic cancer and bladder cancer [43–45].

Previous studies have confirmed that miR-124 also gradually decreases in AD and plays a key role in the regulation of BACE1 gene expression [44,46]. The researchers confirmed the direct interaction between XIST and miR-124, and between BACE1 and miR-124 through luciferase reporter gene assay [47]. They also found that co-transfection of miR-124 could reverse the downregulation of BACE1 and Aβ caused by knocking down XIST [47], which is consistent with the results that miR-124 can regulate BACE1 expression [48].

2.3. MALAT1

Receptor tyrosine kinases (RTKs) are associated with neurological damage in AD and are thus considered important signalling molecules and potential therapeutic targets for this condition. Among the various families of human RTKs, TYRO3 protein tyrosine kinase(Tyro3) [47], Neurotrophic Receptor Tyrosine Kinase 2 (NTRK2) [49,50], Insulin Receptor(IR), and IGF1R [51,52] are closely related to AD. Additionally, members of the Ephrin family of RTKs, such as Ephrin type-A receptor 4(EPHA4) and Ephrin type-B receptor 2(EPHB2) [53,54], are now believed to be associated with the progression of this disease.

Metastasis Associated Lung Adenocarcinoma Transcript 1 (MALAT1) is an evolutionarily conserved lncRNAs that is highly expressed in mammals. By RNA FISH and RIP techniques, MALAT1 was identified as a responder and regulator of miRs-200a, −26a, and −26b, which are elevated in AD and inhibit the RTK Ephrin type-A receptor 2(EPHA2) and several of its downstream effectors; EPHA2 detoxification in AD causes overexpression of its downstream effectors cAMP Response Element-Binding Protein (CREB), p38 Mitogen-Activated Protein Kinase (p38), and synaptophysin, thereby conferring protection against Aβ toxicity [55].

2.4. RPPH1

Ribonuclease P RNA Component H1 (RPPH1) plays a broad regulatory role in numerous diseases. For example, in diabetic nephropathy, RPPH1 is significantly overexpressed in renal tissues, where it enhances the release of tumour necrosis factor-α and monocyte chemotactic protein-1 [56]. In colorectal cancer, RPPH1 interacts with tubulin β3 class III and promotes exosome-mediated macrophage M2 polarization, which in turn drives the metastasis of colorectal cancer [57]. Furthermore, RPPH1 is known to increase the migration, proliferation, and invasion ability of human acute myeloid leukaemia cells by downregulating miR-330-5p [58].

In 2017, researchers constructed the first ceRNAs network of the APPswe/PS1ΔE9 transgenic mouse model. Based on this model, researchers found that RPPH1 was upregulated, dual-luciferase reporter gene assay results showed that miR-330-5p was inhibited, and overexpression of miR-330-5p reduced Cell Division Control protein 42(CDC42) mRNA levels, proving that RPPH1 regulates CDC42 expression through miR-330-5p. RPPH1 increases the expression level of CDC42 and promotes the formation of dendritic spines by competing with endogenously expressed miR-330-5p [59]. The above regulatory circuits demonstrate potential compensatory mechanisms early in AD pathogenesis [60,61]. Additionally, Dual-luciferase reporter assays revealed that miR-122 directly targets the 3’UTR of RPPH1 and Wingless-type MMTV Integration Site Family, Member 1(Wnt1).RPPH1 activates Wnt/β-catenin signalling and ameliorates amyloid β-induced neuronal apoptosis in SK-N-SH cells by directly targeting miR-122 [60].

2.5. LINC01311

The RNA type of the LINC01311 gene (LINC01311) is a novel long non-coding RNA (LncRNAs), which exerts an inhibitory effect on the progression of thyroid cancer by regulating the miR-146b-5p/IMPA2 axis [62]. Researchers have found in Alzheimer’s disease (AD) cell models that LINC01311 May act as a ‘sponge’ or competitively bind miR-146a-5p, thereby regulating the accessibility of miR-146a-5p and its downstream target genes, promoting Aβ accumulation and leading to neuronal damage. It has been confirmed that the LINC01311/miR-146a-5p epigenetic axis may be a functional regulatory factor for neuronal damage in AD [63].

2.6. BACE1-AS

Beta-Secretase 1 (BACE1) transcripts are regulated by the BACE1 antisense (BACE1-AS), which is transcribed from the antisense strand of the BACE1 gene. BACE1-AS is a 2 kb transcript originating from the antisense strand of the BACE1 locus on chromosome 11 at position 11q23.3 [64]. BACE1 plays a crucial role as a key enzyme in the onset and advancement of AD [65]. Research findings have revealed that the levels of BACE1-AS are elevated in serum samples from AD patients, brain tissues of AD transgenic (Tg) mice, and SH-SY5Y cells treated with Aβ1–42. The tandem interaction among BACE1-AS, autophagy-related protein 5 (ATG5), and miR-214-3p was verified by the dual-luciferase reporter gene assay: BACE1-AS regulates the expression of ATG5 in a longitudinal manner by binding to miR-214-3p. The upregulation of BACE1-AS in AD indirectly inhibits the expression of ATG5 by binding to miR-214-3p, leading to excessive Aβ deposition. Rapamycin, an inhibitor of miR-214-3p, can eliminate the neuroprotective effect of shBACE1-AS on Aβ-induced cell damage, and silencing BACE1-AS alleviates neuronal damage in AD by regulating autophagy through the miR-214-3p/ATG5 signalling axis [66].

BACE1-AS1 is also an effective therapeutic target for AD: β-Asarone inhibits the expression of Presenilin 1, Aβ, BACE1, APP, and Sequestosome 1(SQSTM1), while promoting the expression of SYN, LC3 I/II, and Beclin-1, thus showing the potential to alleviate Aβ deposition, thereby promoting autophagy by inhibiting the inhibitory effect of BACE1 on autophagy [67]. In addition, there are reports that isoflurane, as an anaesthetic, can exacerbate the above process [68].

Memantine treatment of streptozotocin-induced AD rats: increased BACE1-as levels and BACE1 gene expression levels were found [69].

Furthermore, studies have shown that berberine exerts neuroprotective effects in AD: Berberine protects neuronal cells from Aβ 25–35 at least partially through the BACE1-AS/miR-132-3p axis. The combination of berberine treatment and BACE1-AS depletion may provide new insights for the treatment of AD [70].

2.7. SOX21-AS1

Long non-coding RNA SOX21 antisense RNA 1 (SOX21-AS1) is a LncRNAs located on chromosome 13q32.1 and is transcribed into a 2986-nucleotide transcript [71]. Literature reports that down-regulation of SOX21-AS1 can reduce oxidative stress in AD mice and inhibit neuronal apoptosis [72]. Cellular models have shown that silencing of SOX21-AS1 can alleviate Aβ-induced neuronal damage by sponging miR-107, which provides a potential therapeutic approach for AD [73].

2.8. NORAD

Non-coding RNA (NORAD) activated by DNA damage is a long-chain non-coding RNA that has been found to maintain the stability of mitosis and DNA transcription by sequestering Pumilio proteins from the cytoplasm [74]. Researchers used the StarBase database to predict the binding sites between NORAD and miR-26b-5p, as well as between miR-26b-5p and membrane metallo-endopeptidase (MME). In Aβ1–42-treated PC12 cells, the expression of NORAD was significantly decreased, while the expression of miR-26b-5p was significantly increased, and the expression of MME was also inhibited, which would lead to the deposition of Aβ. Dual-luciferase reporter gene assays confirmed the existence of interaction binding sites between miR-26b-5p and NORAD; miR-26b-5p also targeted the 3’ UTR of MME. Therefore, NORAD acts as a ‘sponge’ for miR-26b-5p, thereby reducing the inhibition of MME by miR-26b-5p, enhancing the expression of MME, and improving the ability to degrade Aβ, which helps to slow down cell damage [75].

In the above-mentioned ceRNA cell model, researchers hypothesized that the interaction between miRNA and long non-coding RNA (LncRNA) was basically a simple competitive relationship. However, the interaction between miRNA and its targets may not be so simple: in previous studies, some researchers analysed the relationship between miRNA and the abundance of its target binding sites through the mouse liver system. The results showed that the number of binding sites between miRNA and its targets was huge, and competitive RNA was difficult to significantly regulate the function of miRNA in vivo. Therefore, researchers were cautious about the ceRNA model [76,77]. Therefore, researchers suggested that it was not enough to rely solely on basic experiments at the cellular level for verification. It was also necessary to supplement data on the expression levels of LncRNA-miRNA, the number of miRNA response element (MRE) sites, and affinity. In addition, researchers needed to consider multiple factors such as interspecies and intraspecies specificity in different cells or tissues [78] in order to accurately report the function of LncRNA in ceRNA.

3. LncRnas functions as a NATs in Aβ accumulation

NATs refer to RNA molecules transcribed from the DNA strand complementary to the coding gene in the genome [79]. These RNA molecules are complementary in sequence to the mRNA of the coding gene and can form double-stranded structures. NATs are widespread, with an estimated 70% of genes having antisense transcripts, and the majority of them are non-coding RNAs. Among them, antisense LncRNAs exert their functions through various mechanisms, including [80]: (1) Transcriptional regulation: The transcription process itself may have an impact on the sense-strand genes, which is unrelated to long non-coding RNA molecules but only related to transcriptional events [81]. (2) RNA-DNA binding: LncRNAs bind to DNA modification or histone modification proteins, influencing the epigenetics of gene loci and thereby promoting or inhibiting the expression of the sense strand genes [81]. (3) RNA-RNA binding: LncRNAs bind to the mRNA of the sense strand gene through partial base complementary pairing, affecting the alternative splicing, processing, stability and translation efficiency of the mRNA [79]. (4) Coding function: Some antisense LncRNAs can encode small peptides with independent functions, which are unrelated to the sense strand genes [79].

Current studies have shown that NATs have a lot of roles in Tau protein hyperphosphorylation, such as: the increase in SOX21 Antisense RNA 1(SOX21-AS1) expression was accompanied by a decrease in miR-107 levels, which elevated p-Tau levels, reduced the viability of SH-SY5Y and SK-N-SH cells, and triggered apoptosis [73]. Another team reported that the upregulation of SOX21-AS1 in AD mice models exacerbated neuronal oxidative stress injuries by downregulating Frizzled Class Receptor 3/5(FZD3/5) via the Wnt signalling pathway [72]. This upregulation of EBF3 Antisense RNA (EBF3-AS) can lead to decreased expression of Early B-Cell Factor 3(EBF3), which, in turn, induces autophagy and promotes cell death [82]. These observations indicate that Microtubule-Associated Protein Tau Antisense RNA 1(MAPT-AS1) has the potential to serve as a biomarker for AD and a target for future gene therapy interventions [83]. A study indicated that overexpression of Zinc Finger and BTB Domain-Containing Protein 20 Antisense RNA 1 (ZBTB20-AS1) suppresses Zinc Finger and BTB Domain-Containing Protein 20(ZBTB20) expression, diminishes the inhibitory effect of ZBTB20 on Glycogen Synthase Kinase-3 Beta (GSK-3β) gene transcription, leads to elevated GSK-3β expression, and promotes Tau phosphorylation, thereby exacerbating the progression of AD [84]. However, little is known about its effects on Aβ clearance: the RNA type for the Low-Density Lipoprotein Receptor-Related Protein 1 Antisense RNA (LRP1-AS) represents one of the NATs for LRP1 and is also considered as a novel LncRNAs. In the context of AD, LRP1-AS interacts with High Mobility Group Box 2(HMGB2) and inhibits the Srebp1a-dependent transcription of LRP1, thereby disrupting the Aβ clearance mediated by LRP1 which LRP1-AS can facilitate APP internalization, increase Aβ production, and decrease the clearance rate of Aβ. These effects collectively lead to significant cytotoxicity. These observations indicate that LRP1-AS may serve as a biomarker of AD and a potential target for future gene therapy strategies [85]. Silencing the expression of BACE1-AS leads to a disruption of BACE1’s ability to cleave APP and delays the formation of APP-amyloid plaques in AD [86].

4. LncRNAs affect Aβ accumulation through their interaction with histone modifying enzymes

LncRNAs respond to a variety of histone modification enzymes, such as E1A Binding Protein p300/CREB Binding Protein (P300/CBP), General Control Non-repressible 5 Like 2(GCN5L2), and Enhancer of Zeste Homolog 2(EZH2)(Table 2), thereby affecting the expression of genes related to Aβ endocytosis and autophagy, resulting in the production of Aβ and the scavenging of free radicals, leading to excessive Aβ deposition (Figure 2).

Table 2.

LncRNAs affect Aβ accumulation through their interaction with histone modifying enzymes.

LncRNA	Expression	Action	Target	Function	Reference
NEAT1	down(incipient)	protective	P300/CPB	Inhibiting the transcriptional activity and expression of endocytosis-related genes.	[87]
XIST	up	harmful	EZH2	Recruiting EZH2 to mediate the enrichment of HEK27me3 in NEP promoter region.	[89]
EPB41L4A-AS1	down	harmful	GCN5L2	Influencing histone modifications near theTSSs of autophagy-related genes.	[88]

Figure 2.

LncRNAs interact with histone-modifying enzymes in deposition of Aβ. LncRNAs affect Aβ accumulation through their interaction with histone modifying enzymes. LncRNAs interact with multiple histone-modifying enzymes (such as P300/CBP, GCN5L2 and EZH2), thereby influencing the expression of genes related to Aβ endocytosis and autophagy, disrupting the balance between Aβ production and clearance, and ultimately leading to excessive Aβ deposition.

4.1. NEAT1

In the early stage of AD, the level of NEAT1 decreases, leading to an increase in the intracellular concentration of succinyl-CoA, which in turn affects the succinylation and acetylation of H3K27. The interaction between NEAT1 and P300/CBP inhibits the transcriptional activity and expression of the same endocytosis-related genes (CAV2, TGFB2, and TGFBR1), thereby hindering the ability of glial cells (U251) to clear Aβ through endocytosis, revealing the different roles of H3K27Ac and H3K27Cro in gene expression regulation [87].

4.2. EPB41L4A-AS1

The researchers also found that EPB41L4A-AS1 is required to maintain basal autophagy to regulate Aβ clearance: GCN5L2 is an important histone acetyltransferase. EPB41L4A-AS1 regulates the expression of GCN5L2, thereby influencing histone acetylation, crotonylation and lactylation near the transcription start sites of autophagy-related genes, which leads to the alteration of gene expression and the obstruction of Aβ clearance [88].

4.3. XIST

The researchers also verified the interaction between XIST and EZH2 by RIP, and verified by Chromatin immunoprecipitation (ChIP) that XIST recruited EZH2 to mediate the enrichment of HEK27me3 in the Neprilysin (NEP) promoter region, indicating that XIST can regulate neuronal inflammation and injury through the epigenetic regulation of NEP [89].

5. LncRnas affect the accumulation of Aβ through reactions with enzymes related to Aβ metabolism

LncRNAs interact with enzymes related to Aβ metabolism to affect the accumulation of Aβ (Table 3).

Table 3.

LncRNAs affect the accumulation of Aβ through reactions with enzymes related to Aβ metabolism.

LncRNA	Expression	Action	Target	Function	Reference
MIR600HG	up	harmful	NEDD4L	Promoting PINK1 ubiquitination and degradation, autophagy activation.	[92]
51A	up	harmful	SORL1	Impairing APP processing and increasing Aβ formation.	[97]
BC200	up	harmful	BACE1	Accumulating of Aβ and promoting cell death.	[103,104]

5.1. MIR600HG

NEDD4L (neural precursor cell-expressed developmentally downregulated 4-like, also known as NEDD4-2) is a HECT E3 ubiquitin ligase of the NEDD4L family [90]: MIR600HG has been newly identified as a biomarker of pancreatic cancer [91]. The latest research findings on MIR600HG indicate its close association with the overproduction and deposition of Aβ. In AD, sustained mitochondrial dysfunction can lead to memory loss and cognitive decline. MIR600HG expression increases during ageing. Researchers found that the increase in MIR600HG’s direct response to NEDD4L promotes PINK1 ubiquitination and degradation. PINK1 overexpression or MIR600HG knockdown significantly improved cognitive impairment in APP/PS1. Furthermore, the researchers found that MIR600HG induces the production of Aβ and inhibits cytochrome C oxidase activity, a mechanism that leads to decreased cell viability and cell death. AAV-shMIR600HG can restore cytochrome C oxidase activity and inhibit Aβ production [92].

5.2. 51A

Sortilin-related receptor 1 (SORL1; also known as sorting protein-related receptor 1 and LR11) is a 250 kDa type-1 membrane protein expressed in the nervous system [93]. SORL1 interacts with APP, influencing its transport within the brain and facilitating APP cleavage [94,95]. In the brain tissues of patients with AD, SORL1 expression is diminished [96], leading to an increase in APP production and subsequent accumulation of Aβ [94,95]. Although SORL1 has been identified as a high-risk gene for AD, its exact role in the onset and progression of AD remains to be fully elucidated. The expression of 51A influences the splicing of SORL1, thereby shifting the synthesis from the typical long protein variant A to alternative splicing forms. This shift has been implicated in the development of AD. In the brains of patients with AD, 51A expression is increased it drives the transition of SORL1 from transporting variant A to an alternatively spliced protein form. This transition results in impaired APP processing and the enhanced production of Aβ [97].

5.3. BC200

Brain cytoplasmic 200 RNA (BC200), a long non-coding RNA, is highly expressed in the nervous system [98]. Changes in its levels are reportedly linked to the progression of various diseases, including oesophageal squamous cell [99], breast [100], and oesophageal cancers [101]. BC200 levels have been found to be significantly elevated in the plasma of patients with AD, which suggests its utility as a biomarker for the onset and progression of AD [102]. In AD, BC200 upregulated and targets BACE1. The overexpression of BC200 leads to increased BACE1 levels, resulting in the accumulation of Aβ peptides and promotion of neuronal death. Conversely, knocking down of BC200 can significantly restrain BACE1 expression, improve cell viability, and reduce the rate of cell death. Hence, targeting BC200 with siRNA presents a novel approach for gene therapy in AD [103]. Furthermore, BC200 is involved in regulating neuronal apoptosis and neuroinflammation via the PI3K/AKT pathway in AD [104].

6. LncRnas RNAs affect the influence of microglia on Aβ deposition

Microglia (MIs) are the primary immune cells in the central nervous system (CNS), playing crucial roles in immune surveillance, inflammatory responses, injury repair, and neural support. Recent studies have found that LncRNAs exacerbate the inflammatory response and neuronal apoptosis during the pathological process of AD by influencing microglia, and the above processes may be related to Aβ deposition.

6.1. RP11‑543N12.1

Researchers designed a co-culture model of neurons and microglia and incubated the co-culture model with Aβ. The study found that in the above model, the upregulation of RP11-543N12.1 regulated the expression of miR-324-3p through the ‘sponge effect’, thereby promoting neuronal apoptosis and the inflammatory response of microglia. Therefore, the authors believe that the above mechanism would lead to neuroinflammation and neuronal apoptosis [105].

6.2. XIST

Researchers found that XIST expression was upregulated in the hippocampus in APP/PS1 mice and Aβ1–42-treated BV2 microglial cells. After silencing XIST, they observed improved memory in mice, reduced Aβ1–42 levels in the hippocampus, inhibited activation of microglia, M1 polarization and pro-inflammatory factor levels, and promoted M2 polarization of microglia. The study demonstrated that XIST participates in regulating M1/M2 polarization of microglia through the miR-107/PI3K/Akt axis and affects its neurotoxicity [106].

7. LncRNAs modulates Aβ accumulation through additional regulatory mechanisms

7.1. 17A

Alternative splicing is a crucial component of human brain complexity; however, its mechanisms remain elusive. The expression of LncRNAs 17A is dependent on RNA Polymerase III in the human brain. In 2010, 17A was found to be overexpressed in the brain tissues of patients with AD. 17A is consistently expressed in SH-SY5Y neuroblastoma cells and is induced into an alternative splicing isoform, which stimulates the production of a specific spliceosome for Gamma-Aminobutyric Acid B Receptor Subunit 2(GABAB2). This leads to the inhibition of cAMP accumulation in cells and the activation of potassium ion channels. Such effects have implications for the biological functions of receptors, such as diminishing GABAB2 signal transduction in cells and contributing to deficits in learning and memory. Researchers suggest that these outcomes may be linked to neurodegenerative diseases, including AD [107]. Subsequent studies have further established that overexpression of 17A increases the production of Aβ peptides and the ratio of Aβx-42/Aβx-40 in SH-SY5Y cells. Additionally, other studies indicate that overexpression of 17A can inhibit autophagy and neural growth in AD cell models [107,108].

7.2. MEG3

Maternal expression gene 3 (MEG3), as indicated in previous studies, acts as a tumour suppressor gene and is downregulated in various cancers. For instance, decreased levels of MEG3 in breast cancer can influence the miR-494/OTDU4 axis, contributing to the proliferation and development of breast cancer cells [109]. Furthermore, MEG3 has been implicated in activation of the PI3K/AKT pathway, which is known to promote the development of liver cancer [110], retinoblastoma [111], and glioma [112]. Recent research on AD has suggested that MEG3 might also facilitate the progression of this disease through the PI3K/AKT pathway. In the hippocampi of mice with AD, a notable downregulation of MEG3 has been observed. This downregulation is associated with activation of the PI3K/Akt signalling pathway, increased levels of Aβ, activation of astrocytes, and elevated levels of pro-inflammatory cytokines, such as Interleukin-1 Beta, Interleukin-6, and Tumour Necrosis Factor Alpha, in the hippocampus. These changes exacerbate inflammatory damage, promote neuronal apoptosis, and negatively affect spatial learning and memory capacity [113].

7.3. BC1

In the brains of mice with AD, the expression of Brain Cytoplasmic RNA 1 (BC1) is increased. BC1 promotes the translation of APP mRNA by interacting with Fragile X Mental Retardation Protein (FMRP), leading to increased production and accumulation of Aβ peptides, which negatively affects spatial learning ability and memory capacity [114]. Knocking down BC1 or disrupting the interaction between BC1 and FMRP can significantly reduce the accumulation of Aβ peptides in the brain. In 5-month-old AD mouse models, specific down-regulation of BC1 expression was achieved by intraventricular injection of LNA-BC1. The results showed that the down-regulation of BC1 could effectively inhibit the expression of APP protein in the hippocampus of AD mice. Additionally, in 180 ± 5-day-old AD mice, intravenous injection of Tat-FMRP_61–79 was used to block the interaction between BC1 and FMRP. Two hours after administration, it was found that the FMRP_61–79 peptide not only effectively inhibited the binding of BC1 and FMRP but also promoted the association of FMRP with APP mRNA, thereby inhibiting the translation process of APP and significantly reducing the aggregation of Aβ peptides. Behavioural assessment was conducted five days after the last administration, and the results indicated that Tat-FMRP_61–79 could significantly alleviate the decline in spatial learning and memory function in AD mice and improve their cognitive performance. The above results suggest that the interaction between BC1 and FMRP plays a key regulatory role in the pathogenesis of AD [115].

7.4. LINC00672

The mechanism by which LINC00672 promotes autophagy by up-regulating GPNMB remains unclear, although it has been established that LINC00672 does indeed enhance autophagy through this process [116].

8. Summary

LncRNAs are associated with the development of AD. This review focuses on the regulatory roles of lncRNAs in the excessive deposition of amyloid-β (Aβ), a pathological factor in AD. We categorize these roles into ceRNAs, NATs, interactions with histone-modifying enzymes, interactions with enzymes related to Aβ formation, and other mechanisms. As ceRNAs, they influence pathways through the LncRNA/miRNA axis by interacting with miRNAs. Despite numerous research findings, the differences between cell experiments and the actual in vivo environment have led to controversies over these results. Quantifying expression levels, the number of miRNA response elements (MRE) sites, and clarifying the molecular affinity might be better approaches to address this issue. As NATs, they regulate pathological processes by targeting upstream genes. Histone-modifying enzymes and other enzymes play roles in the clearance and generation of Aβ. Additionally, in other neuroinflammatory-related diseases such as Parkinson’s disease [17] and cerebral ischaemia-reperfusion injury (CIRI) [117], lncRNA HOXA-AS2 regulates microglial polarization by recruiting PRC2 and epigenetically modifying PGC-1α expression, thereby promoting neuroinflammation. LOC102555978 promotes NLRP3-mediated pyroptosis in RM cells by sponging miR-3584-5p, which may provide a potential therapeutic target for post-CIRI inflammation regulation. This might suggest that long non-coding RNAs could also affect Aβ deposition and clearance by modulating the immune microenvironment.Although numerous studies have confirmed the close relationship between long non-coding RNAs and the accumulation process of Aβ, translating these findings into therapeutic interventions remains challenging, and there are currently no good translational outcomes. The difficulties lie in the following aspects: First, the complex interactions between long non-coding RNAs and adjacent genes mean that altering their levels may disrupt the expression of other genes. Second, for some multi-exon long non-coding RNAs, there may be alternative downstream promoter sequences other than the first exon, which could lead to the failure of long non-coding RNA knockout. Third, some long non-coding RNAs have other non-coding RNAs within their introns, which increases the complexity of their therapeutic applications. Therefore, effectively regulating the levels of long non-coding RNAs is a core challenge and a key area for future research.

Funding Statement

The work was supported by the Shandong Provincial Natural Science Foundation [ZR2020LZL008].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

All authors have accepted responsibility for the entire content of this manuscript and approved its submission. Ruo-Mei Wang: Conceptualization and Writing – original draft, Zi-Qiang Wang: Conceptualization and Writing – original draft.

Data availability statement

Data sharing is not applicable to this article as no new data were created or analysed in this study.