Cellular response to β‐amyloid neurotoxicity in Alzheimer's disease and implications in new therapeutics

Abstract

β‐Amyloid (Aβ) is a specific pathological hallmark of Alzheimer's disease (AD). Because of its neurotoxicity, AD patients exhibit multiple brain dysfunctions. Disease‐modifying therapy (DMT) is the central concept in the development of AD therapeutics today, and most DMT drugs that are currently in clinical trials are anti‐Aβ drugs, such as aducanumab and lecanemab. Therefore, understanding Aβ's neurotoxic mechanism is crucial for Aβ‐targeted drug development. Despite its total length of only a few dozen amino acids, Aβ is incredibly diverse. In addition to the well‐known Aβ_1‐42, N‐terminally truncated, glutaminyl cyclase (QC) catalyzed, and pyroglutamate‐modified Aβ (pEAβ) is also highly amyloidogenic and far more cytotoxic. The extracellular monomeric Aβ_x‐42 (x = 1–11) initiates the aggregation to form fibrils and plaques and causes many abnormal cellular responses through cell membrane receptors and receptor‐coupled signal pathways. These signal cascades further influence many cellular metabolism‐related processes, such as gene expression, cell cycle, and cell fate, and ultimately cause severe neural cell damage. However, endogenous cellular anti‐Aβ defense processes always accompany the Aβ‐induced microenvironment alterations. Aβ‐cleaving endopeptidases, Aβ‐degrading ubiquitin‐proteasome system (UPS), and Aβ‐engulfing glial cell immune responses are all essential self‐defense mechanisms that we can leverage to develop new drugs. This review discusses some of the most recent advances in understanding Aβ‐centric AD mechanisms and suggests prospects for promising anti‐Aβ strategies.

Many variants of Aβ exist, created by a combination of N‐terminal truncates and C‐terminal truncates. QC enzyme catalyzed pEAβ_x‐42, plays the role of “ignition” as a seed to cause Aβ to aggregate into plaques. Extracellular Aβ activates various Aβ receptors on cell membranes, triggering numerous downstream cell signaling cascades, ultimately impacting cell metabolism, gene expression, cell cycle, and differentiation. On the other hand, Aβ activates endogenous anti‐Aβ mechanisms, and these self‐protection mechanisms are promising targets for developing future AD drugs.

1. ALZHEIMER'S DISEASE AND AMYLOID HYPOTHESIS

Alzheimer's disease (AD) is a chronic neurodegenerative disease hallmarked by cognitive defects, specifically episodic memory impairment. ¹ The predominant risk factor for AD is aging since AD appears predominantly in late adulthood. The risk of developing the disease rises exponentially after the sixth decade of life. ² The most critical in vivo pathological hallmarks of the disease are extracellular β‐amyloid (Aβ) accumulation ³ and intracellular tau propagation. ⁴ Early deficits of AD typically exhibit as mild and can slowly progress for as long as several decades. The aggregation of Aβ peptide and microtubule‐associated protein tau begins early and initiates the disease, marking the biochemical phase. ⁵ Later, neuroinflammation and immune responses mediated by excessive proliferation of microglia and astrocytes ⁶ , ⁷ arise during the cellular phase. ⁵ MRI changes, hippocampal shrinkage, memory loss, and dementia appear when the disease advances to the clinical phase. ⁵

Amyloid‐containing tissue was first documented in 1639 with a description of the appearance of the lardaceous liver and white stone‐containing spleen tissue, and characterization of the protein has continued over the subsequent four centuries. ⁸ Over time, the concept of “amyloidosis” was established, and Aβ found in the AD brain was added to the growing list of substances triggering amyloidosis. The “amyloid hypothesis” implies that increases in amyloid plaques or soluble Aβ are the primary events causing the onset and progression of the disease. ³ Aβ pathology directly involves accumulation of Aβ oligomers and Aβ plaque, as well as indirectly leading to cellular metabolic disorders (e.g. ER stress and oxidative stress), ⁸ , ⁹ synaptic deterioration, ¹⁰ and neuronal apoptosis ¹¹ (Figure 1). This toxicity mechanism is conserved across a wide range of species, as supported by the observation that increased Aβ levels can lead to diffused amyloid deposits in the brain and severe cognitive defects such as memory deficits in mice and fruit flies. ¹² , ¹³

FIGURE 1.

Aβ pathology constitutes an essential mechanism for AD. Aβ oligomers, or Aβ plaques, exert their neurotoxic effects by causing various cellular metabolic disorders, such as ER stress, oxidative stress, and cell cycle disturbances. These neural function disruptions lead to further synaptic deterioration and neuronal apoptosis. This figure was created with BioRender.com with the publication agreement number BZ24RS5K5N.

Because of the neurotoxic effects of Aβ, AD patients can exhibit a variety of cognitive dysfunctions, such as slow response times, impaired short‐term memory, repetitive speech, and decreased comprehension and expression capabilities. To date, there is still no effective drug for AD. In the past few decades, pharmaceutical companies have invested heavily in R&D to develop drugs that target Aβ, many of which have successfully entered clinical trials. Unfortunately, all have failed in clinical trials. ¹⁰ So why are Aβ‐targeted drugs ineffective? Is the amyloid hypothesis wrong? One possible explanation is that due to the insidious onset of AD and the prolonged course of the disease, Aβ may indeed initiate the disease at the very initial stage, but it also causes changes in other molecules in the cell, such as phosphorylation of tau protein at the synaptic site, ¹⁴ and gradually continues to affect to more downstream molecules. When the clinical symptoms of AD have begun to appear, the progression of the disease is likely no longer dependent solely on Aβ, so trying to reduce the level of Aβ is not enough to significantly improve the patient's cognitive impairment. ¹⁴ Even so, scientists keep developing drugs that target Aβ using a new mainstream concept, disease‐modifying therapy (DMT). ¹⁵ DMT is based on direct action of the drug on the pathophysiological mechanism of the disease, that is, treatment of the cause, but not the symptom. Currently, the DMT drugs that account for the most significant proportion of clinical trials are anti‐Aβ drugs, such as aducanumab and lecanemab, which demonstrates academic acceptance of the amyloid hypothesis. Therefore, understanding Aβ's neurotoxic mechanism is critical for Aβ‐targeted drug development.

2. Aβ PRODUCTION AND AGGREGATION

Aβ consists of peptides of 37–49 amino acids produced by sequential proteolytic cleavage of amyloid precursor protein (APP) by β‐secretase and γ‐secretase. ¹⁶ The γ‐secretase can cleave at different C‐terminal positions, creating Aβ fragments of different lengths. ¹⁷ Aβ plaques also contain N‐terminally truncated Aβ peptides, giving Aβ species even more variation in sequence and length. ¹⁸ Among all these Aβ species, Aβ_1‐42, Aβ_4‐42, and Aβ_11‐42 are proven to be the most amyloidogenic forms of the peptide. ¹⁹ Aβ molecules can aggregate to form soluble oligomers, which can further misfold to form, first, Aβ fibrils and, ultimately, Aβ plaques. ⁸ Many neurodegenerative diseases, including AD, involve prion mechanisms. Protein conformational disorders (PCD) derive from protein misfolding, leading to conformational change which might promote toxicity and subsequent disease. The misfolded protein can self‐associate and becomes deposited in Aβ deposits. The critical events in Aβ PCD are the turn formations and β‐pleated‐sheet conformations ²⁰ , ²¹ , ²² that stimulate this aggregation in AD. ²³ Importantly, the turn position plays a pivotal role in the ability of Aβ to induce aggregative and cytotoxicity. The difference in cytotoxicity between Aβ₄₂ and Aβ₄₀ can be explained by the difference in the turn position in the aggregates; the former has a turn at positions 22 and 23, ²⁰ and the latter has a turn at positions 25 and 26 ²¹ due to the difference in the salt bridges between Lys28/Ala42 ²⁴ and Asp23/Lys28. ²¹

Pyroglutamate (pGlu) modified N‐terminally truncated Aβ (pEAβ) likely initiates aggregation, called “seeding”, because the Aβ plaque core mainly contains pEAβ_x‐42 (x = 2–11). ²⁵ Further, the researchers discovered that pEAβ_3‐42 can co‐oligomerize with excess Aβ_1‐42 to form misfolded metastable low‐n oligomers (LNOs) that are far more cytotoxic to cultured neurons than comparable LNOs made from Aβ_1‐42 alone and can facilitate the seeding of new LNOs. ²⁶ Because of its seeding capability and convincing preclinical trial evidence, pEAβ has recently become a new potential drug target in AD. ²⁷ In addition, because glutaminyl cyclase (QC) catalyzes the formation of pEAβ, ²⁸ developing drugs targeting QC may also be promising, as supported by a series of pilot investigations. For example, studies using QC knockout transgenic mice ²⁹ or QC inhibitors ³⁰ showed that decreasing QC levels can efficiently rescue the AD phenotype.

While Aβ exists both intracellularly and extracellularly, most is present in the extracellular environment. ³¹ In the first instance, Aβ, a secreted protein, is cleaved from APP and secreted into the extracellular space. ⁹ Extracellular Aβ can be taken up into the cell via endocytosis. ⁸ , ³² Extracellular Aβ aggregates, but intracellular Aβ does not because Aβ aggregation requires a seeding event mediated by Aβ and cell‐surface interactions, ³³ and membrane lipid‐mediated seeding takes place only in the extracellular space. ³⁴ , ³⁵ Specifically, gangliosides are located in the outer leaflet of the plasma membrane. ³⁶ Monosialotetrahexosyl ganglioside (GM1), which is implicated in raft structure, ³⁶ binds tightly to Aβ₄₀ ³⁷ and Aβ₄₂ ³⁸ and facilitates their aggregation. ³⁷ , ³⁹ Lipid raft‐associated Aβ can serve as a platform and acts as an adaptor to initiate extracellular Aβ aggregation into Aβ plaques. ³⁴

3. RECEPTOR‐COUPLED SIGNAL PATHWAYS THAT MEDIATE Aβ CYTOTOXICITY

The soluble extracellular Aβ dimers or oligomers can activate numerous cellular signal transduction pathways coupled to multiple types of well‐recognized Aβ receptors, causing pleiotropic downstream effects. ⁹ So far, many Aβ receptors have been discovered, including receptor of advanced glycation endproducts (RAGE), ⁴⁰ p75 neurotrophin receptor (p75NTR), ⁴¹ and nicotinic acetylcholine receptors (nAChRs), ⁴² , ⁴³ among others. In particular, p75NTR, also known as nerve growth factor receptor (NGFR), functions at the molecular nexus of cell death, survival, and differentiation and thus serves as a new therapeutic candidate for intervention in many diseases, including AD. ⁴⁴ Blocking Aβ binding to p75NTR can reduce Aβ‐induced neurodegeneration, neuroinflammation, and cognitive deficits. ⁴⁴ Drug developers have carried out animal studies in in vivo AD models to test LM11A‐31, a p75NTR inhibitor, and observed its excellent performance in reducing Aβ oligomer‐induced dendritic loss and synaptic failure. ⁴⁵ Clinical randomized controlled trials (RCT) of LM11A‐31 in AD are currently ongoing and hopefully will give us positive outcomes soon.

Aβ can not only alter cellular signal transduction pathways by directly binding Aβ receptors but can also interfere with the binding of other ligands and receptors. New studies have shown that before insoluble Aβ plaque formation, soluble Aβ dimers cause glutamate excitotoxicity by blocking glutamate reuptake in the synaptic cleft. ⁴⁶ , ⁴⁷ The excessive glutamate stimulation of multiple types of glutamate receptors and their downstream cell signaling transduction cascades, such as CamKII ⁴⁸ and MAPK/ERK, ⁴⁹ , ⁵⁰ can cause neurotoxicity. ⁵¹ Another study has shown that Aβ can cause insulin receptor redistribution, leading to insulin resistance, and disturbing ERK/MAPK, PI3K/AKT, and GSK3 pathways. ⁵²

To a large extent, the signal transduction pathways disrupted by Aβ overlap with AD‐related neurodegeneration cellular networks, such as the PKC pathway, ⁵³ , ⁵⁴ the NF‐κB pathway, ⁵⁵ and the Ras/MAPK/ERK pathway, ⁵⁶ , ⁵⁷ etc. These dysregulated cellular signal transduction pathways can potentially alter gene transcription by phosphorylating transcription factors and their coactivators, ⁵⁸ , ⁵⁹ , ⁶⁰ nuclear receptors, ⁶¹ and histones ⁵⁸ (Figure 1). A crucial focus of AD research is neurogenesis in the brain, which may help development of pro‐neurogenic therapies. ⁶² Although Aβ has been proven to inhibit cell proliferation ⁶³ , ⁶⁴ due to disruption of the mitotic spindle and mitotic microtubule motors, ⁶⁵ it seems to positively enhance the neuronal differentiation phenotype ⁶⁶ , ⁶⁷ , ⁶⁸ , ⁶⁹ (Figure 1). The underlying mechanism likely involves the Ras GTPases‐MAPK molecular network induced by Aβ. The small GTPases play essential roles in neurogenesis and cell differentiation and are implicated in AD pathogenesis. ⁵⁷ Studies have shown the positive relationship between Ras superfamily members and Aβ₄₂ production, ⁷⁰ supporting observations of increased neuronal differentiation and the parallel increase in Aβ plaques observed in human AD. ⁷¹ As expected, inhibiting these Ras GTPases decreases Aβ₄₂ levels and improves cognition in AD mouse models, ⁷² making the Ras/MAPK/ERK pathway a promising therapeutic target for AD. Notably, although it plays a neuroprotective role in fighting amyloid pathology by changing APP processing and reducing Aβ₄₂ production, ⁷³ excess neuronal differentiation has other side effects, such as increased neurogenic‐to‐gliogenic fate switch, e.g. astrogenesis ⁷³ , ⁷⁴ and increased APP expression. ⁷³

4. CELLULAR ANTI‐Aβ DEFENSE MECHANISMS

In response to excessive production of Aβ, cellular defense mechanisms are initiated. Leveraging these endogenous anti‐Aβ mechanisms can hopefully help to develop new drugs with different mechanisms for treating AD in the future (Figure 2). Many endopeptidases that exist in cells are also Aβ‐degrading proteases (Figure 2A). Endopeptidases directly cleave Aβ ⁷⁵ , ⁷⁶ , ⁷⁷ or reduce processing of APP to Aβ, ⁷⁸ playing a crucial role in Aβ clearance. Accumulating studies have reported many such endopeptidases. Neprilysin (NEP), a metalloendopeptidase, can directly cleave Aβ, ⁷⁹ , ⁸⁰ , ⁸¹ and its mRNA and protein levels in the hippocampus and temporal gyrus in AD patients are significantly lower than those in the control group. ⁸² Similarly, in the cerebral cortex of transgenic Tg2576 mice with AD‐like pathology, NEP proteins levels were decreased. ⁸³ Genetic linkage analyses on chromosome 10q have been used to find a strong association between the insulin‐degrading enzyme (IDE) gene and late onset AD, ⁸⁴ and IDE can also degrade Aβ in the brain. ⁸¹ , ⁸⁵ Matriptase, a member of the type 2 transmembrane serine protease (TTSP) family that is encoded by the suppression of tumorigenicity‐14 (ST14) gene, cleaves APP and reduces its processing to Aβ. ⁷⁸

FIGURE 2.

Research evidence reveals endogenous anti‐Aβ mechanisms, which provide alternative drug development approaches to suppress Aβ pathologies. These mechanisms primarily include, but are not limited to, endopeptidases, the ubiquitin‐proteasome system (UPS), and glial cell‐mediated immune responses. (A) Many endopeptidases can directly cleave Aβ. (B) The UPS can utilize a sizeable enzymatic system to precisely degrade unwanted proteins. The new proteolysis targeting chimeras (PROTACs) technique has become a promising way to fight AD pathologies. (C) Glial cells are more than glue. They can internalize and clear Aβ just like scavengers. This figure was created with BioRender.com with the publication agreement number RQ24RS581W.

Another Aβ degrading mechanism is the ubiquitin‐proteasome system (UPS), which degrades unwanted or damaged proteins in the cytoplasm (Figure 2B). Ubiquitination first acts as a signal for protein degradation, mediated by three different enzymes: the ubiquitin‐activating enzyme (E1), the ubiquitin‐conjugating enzyme (E2), and the ubiquitin‐ligase enzyme (E3). ⁸⁶ Then ubiquitin‐tagged proteins are directed to the 26S proteasome for degradation. ⁸⁶ Studies have proved that UPS can degrade Aβ, ⁸⁷ and UPS defects are evident in AD. ⁸⁸ A study using postmortem human cerebral cortex revealed that the trypsin‐like proteolytic activity of the 26S proteasome was reduced, and E1 and E2 activity was significant decreased in an AD sample. ⁸⁹ Another mass spectrometry‐coupled liquid chromatography investigation found that ubiquitin C (UBC) is a key protein that interacts with a variety of pathophysiological molecular factors associated with AD and suggested that the reduction in UPS is one of the causative factors of AD. ⁹⁰ A recently developed technique, proteolysis targeting chimeras (PROTACs), ⁹¹ can utilize UPS to efficiently and precisely degrade target proteins, such as phosphorylated tau, ⁹² to battle tauopathies. Whether this technique can also efficiently reduce Aβ pathologies by targeting Aβ and Aβ‐associated molecules is a fascinating question for future research.

Neural immune responses mediated by brain immune cells – microglia and astrocytes – in response to Aβ‐induced inflammation play a crucial role in AD development. ⁶ , ⁹³ A recent transcriptomic study utilizing laser capture microscopy (LCM) discovered the upregulation of microglia‐related genes in the Aβ plaque tissue compared to non‐Aβ plaque tissue in the APP mouse, suggesting a robust immune response at the molecular level under the omic scale. ⁹⁴ Both microglia ⁹³ and astrocytes ⁹⁵ can take up amyloid (Figure 2C), which primarily implicates Aβ‐binding scavenger receptors. The microglial phagocytosis‐mediated clearance of Aβ ⁷ helps to alleviate Aβ pathology. Class B scavenger receptors such as CD36, expressed on microglia, can recognize and bind Aβ and mediate macrophage recognition of Aβ. ⁹⁶ , ⁹⁷ In addition, class A scavenger receptors can mediate Aβ phagocytosis by astrocytes and lead to Aβ clearance. ⁹⁵ Based on these neuroprotective functions, researchers anticipate that glial cells may become new therapeutic targets in AD treatment. Still, neuroinflammation is a double‐edged sword because the hyperactivation of glial cells can also be detrimental. In most cases, Aβ can disrupt the ability of microglia to clear the protein and cause microglia hyperactivation. ⁹⁸ In addition, senescent astrocytes showed increased secretion of senescence‐associated secretory phenotype (SASP) factors in AD, reducing the glial cells' ability to clear Aβ. ⁹⁹

5. SUMMARY AND FUTURE DIRECTIONS

Currently, the main approaches to AD drug development include Aβ‐targeted drugs, tau‐targeted drugs, and drugs using other mechanisms. Over the past 20 years, hundreds of drugs, including Aβ‐targeted medicines, have been effective in animal experiments, but most have failed in clinical trials. Therefore, developing new powerful drugs for treating AD is urgent and requires a more profound understanding of AD mechanisms. This review has discussed some of the most recent advances in understanding AD mechanisms and how this further knowledge helps contribute to new drug development. Future work will soon investigate these Aβ‐centric mechanisms, such as the endogenous Aβ degrading system, in more depth and bring us more effective treatment methods for AD.

AUTHOR CONTRIBUTIONS

HZ, FE, and JW conceived and drafted the manuscript. HZ, XL, and JX designed and made the graphics. XL and XW helped find, collect, and sort references.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest. Haolin Zhang is an Editorial Board member of AMEM and a co‐author of this article. To minimize bias, he was excluded from all editorial decision‐making related to the acceptance of this article for publication.

Zhang H, Li X, Wang X, Xu J, Elefant F, Wang J. Cellular response to β‐amyloid neurotoxicity in Alzheimer's disease and implications in new therapeutics. Anim Models Exp Med. 2023;6:3‐9. doi: 10.1002/ame2.12313