The Pathogenesis of Alzheimer’s Disease (AD)
Patients with Alzheimer's disease (AD) accumulate abundant Aβ plaques and NFTs in their brains, accompanied by a range of pathological processes such as
neuroinflammation, synaptic dysfunction, mitochondrial and bioenergetic disturbances, as well as vascular abnormalities. The pathogenesis of AD is currently explained by three core mechanistic hypotheses: the cholinergic hypothesis, the amyloid hypothesis, and the tau protein hypothesis. Emerging evidence further implicates neuroimmune dysregulation,
oxidative stress, metabolic disturbances, biometal imbalance, gut-brain axis dysfunction, etc. as co-contributors to disease progression
[2].
Cholinergic Hypothesis
As the earliest proposed mechanism, degeneration of cholinergic neurons in the nucleus basalis of meynert (NBM) leads to reduced choline acetyltransferase (ChAT) activity in projection areas such as the
cerebral cortex and the
hippocampus (regions associated with learning and memory).
Acetylcholine (ACh) is synthesized from choline and
acetyl-coenzyme A by ChAT, and then it is transported into the synaptic vesicles via the
vesicular acetylcholine transporter (VAChT). Upon arrival of a neural signal, ACh is released and binds to muscarinic and nicotinic acetylcholine receptors (
mAChRs and
nAChRs) on the postsynaptic membrane, facilitating signal transmission. ACh in the synaptic cleft is then degraded into choline by
acetylcholinesterase (AChE) and reabsorbed into presynaptic cholinergic neurons. The decline in ChAT activity, along with the detrimental effects of Aβ on nutritional imbalance, ACh synthesis, release, and degradation, leads to decreased ACh levels. This decrease impairs the physiological functions of ACh in learning, memory, motor control, and sleep cycle regulation. Moreover, neuronal damage is accompanied by a significant increase in the density of senile plaques. The cholinergic hypothesis posits a strong association between cholinergic deficits in the basal forebrain and cognitive impairment in AD patients.
Figure 1. AD pathogenesis: diagram of the cholinergic hypothesis[2].
Amyloid Hypothesis
The amyloid hypothesis proposes that the primary cause of AD is the abnormal accumulation and deposition of amyloid proteins in the brain. The buildup of Aβ is considered the hallmark pathology in both autosomal dominant and sporadic late-onset forms of AD, which have been extensively investigated.
Aβ is generated through the proteolytic processing of the transmembrane glycoprotein
amyloid precursor protein (APP), which is cleaved by β-secretase and γ-secretase to produce Aβ fragments of varying lengths—predominantly Aβ40 and Aβ42. The hydrophobic C-terminus of Aβ42 facilitates conformational transition in the β-sheet and promotes aggregation into the core component of senile plaques. Aβ may contribute to AD pathology by losing its physiological function during aggregation.
It is important to note that the amyloid cascade hypothesis remains controversial. It faces challenges in explaining multiple pathological features, shows only a weak correlation between Aβ and cognitive decline, and has failed to demonstrate efficacy in numerous clinical drugs targeting Aβ.
Figure 2. AD pathogenesis: schematic of the amyloid hypothesis[2].
Tau Protein Hypothesis
Tau is a microtubule-associated protein that is mainly expressed in the axons of neurons, with lower expression levels in dendrites, soma, and glial cells. It contains numerous phosphorylation sites across the N-terminal, C-terminal and repeat regions, which are regulated by a balance of various kinases and phosphatases to maintain normal neuronal physiology. Under pathological conditions, an imbalanced activity of these enzymes results in tau hyperphosphorylation. This hyperphosphorylation causes tau protein to detach from microtubules, undergo conformational changes, mislocalization, and aggregate into tau oligomers, paired helical filaments (PHFs), and NFTs in neurons and dendrites. These pathological changes disrupt neuronal function and ultimately lead to cell death. Notably, the spatial and temporal distribution of tau protein—a major component of NFTs—correlates strongly with clinical symptoms and serves as a highly specific pathological biomarker in AD patients.
Figure 3. AD pathogenesis: schematic of the tau protein hypothesis[2].
Other Hypotheses
The pathogenesis of AD is complex and multifactorial, and its exact pathogenesis remains incompletely understood. Beyond the key roles of acetylcholine, Aβ and tau protein, a range of additional factors may contribute to AD pathogenesis, including neuroinflammation, oxidative stress, biometallic imbalance, glutamate dysregulation, insulin resistance, abnormalities of the gut microbiota, disturbances in cholesterol homeostasis, mitochondrial dysfunction and abnormalities in
autophagy.
Figure 4. AD pathogenesis: schematic of other emerging hypotheses[2].
Neuroinflammation Hypothesis:
Neuroinflammation is usually characterized by a chronic inflammatory response in the
central nervous system (CNS), typically involving the activation of glial cells and release of pro-inflammatory factors. Microglia, which function as immune defense macrophages in the CNS, can become hyperactivated by the accumulation of β-amyloid in the brain, leading to the release of proinflammatory factors that contribute to the progression of AD.
Oxidative Stress Hypothesis:
Reactive oxygen species (ROS), reactive nitrogen species, and other highly reactive and unstable substances are produced during regular metabolic processes. Under physiological conditions, these substances are normally kept at low levels by an effective antioxidant defense system to protect cells from oxidative damage. However, in the brain of AD patients, metal accumulation, overexpression of related enzymes (e.g.,
NADPH oxidase), and mitochondrial dysfunction contribute to the overproduction of ROS. When ROS levels exceed the capacity of the endogenous antioxidant system, oxidative imbalance occurs, damaging neuronal membrane lipids, proteins and nucleic acids, ultimately leading to neuronal cell death.
Metal Ion Hypothesis:
Under physiological conditions, trace metals are essential for maintaining the microenvironmental homeostasis of neuronal cells. This balance may be disrupted by inappropriate deposition or misdistribution of metal ions, with the dyshomeostasis of Fe2+, Cu2+, and Zn2+ closely associated with AD. The accumulation of these biometals in Aβ plaques and NFTs plays a key role in pathological protein deposition. For example, they can modulate the activity of essential enzymes, alter protein conformation, or interfere with clearing pathways.
Glutamatergic Excitotoxicity:
Glutamates are the primary excitatory neurotransmitters for glutamatergic neurotransmission in the CNS. Their receptors include ionophilic glutamate receptors, including
NMDA receptors,
AMPA receptors, and kainate receptors, as well as
metabotropic glutamate (mGlu) receptors. Glutamate interacts primarily with NMDA receptors to control the influx of sodium and calcium to neurons. Under physiological conditions, magnesium ions block the cationic channel of the NMDA receptors, preventing excessive ion influx. However, in AD, an overstimulation of NMDA receptors displaces magnesium, allowing excessive sodium and calcium ions to enter neurons. The entry of sodium ions into the neuron leads to temporary swelling, while elevated intracellular calcium levels trigger a cascade of Ca
2+-dependent processes, including ROS production, mitochondrial dysfunction, and activation of necrotic/apoptotic pathways, ultimately resulting in permanent excitotoxic neuronal damage to the neurons.
Microbiota-Gut-Brain Axis Hypothesis:
In recent years, the microbiota-gut-brain axis hypothesis has gained significant attention as a potential avenue for novel therapeutic strategies. Compromise of the intestinal epithelial barrier allows harmful substances and microorganisms from the intestinal tract to enter the bloodstream, triggering immune responses that may lead to systemic inflammation. This systemic inflammation may enable inflammatory mediators to cross the blood-brain barrier and affect microglia, further exacerbating neuroinflammation. This process is accompanied by disruptions in neurotransmission, ultimately leading to neuronal degeneration and damage.
Abnormal Autophagy:
Autophagy is a highly conserved metabolic degradation process that maintains cellular homeostasis by delivering intracellular protein aggregates and damaged organelles to lysosomes for degradation and recycling. In AD, defects in autophagy disrupt protein homeostatic networks (Aβ production and extracellular secretion, as well as abnormal aggregation of tau protein) and result in the accumulation of damaged organelles, such as dysfunctional mitochondria.
Drug Development Pipeline for AD
With growing insights into AD pathogenesis and therapeutic targets, drug development has accelerated in recent years, with an increasing focus on disease-modifying therapies (DMTs) that aim to tackle underlying pathology rather than merely alleviating symptoms. Disease-targeted therapies (DTTs)—which act directly on key mechanisms such as amyloid, tau, and neuroinflammation—constitute the primary path toward effective DMTs. This emphasis is evident in the current AD pipeline, where most candidates focus on these core disease processes.
As of January 1, 2025, there are 138 AD drug candidates across 182 clinical trials worldwide. This included 48 trials assessing 31 drugs in Phase 3; 86 trials assessing 75 drugs in Phase 2; and 48 trials assessing 45 drugs in Phase 1 (Figure 5). Of these 182 trials, 16 are long-term extensions of agents in prior trials[3].
The AD drug development pipeline is comprised primarily of DTTs, which account for 102 candidates, representing 74% of drugs in clinical trials. Additionally, 20 agents (14%) of the pipeline are putative cognitive enhancers, and 15 drugs (11%) target neuropsychiatric symptoms (NPS) in participants with AD. Among the DTTs, 60 (59%) are small molecules and 42 (41%) are biological therapies. In terms of clinical development phases, DTTs make up 20 drugs (65%) in Phase 3, 60 drugs (81%) in Phase 2, and 33 drugs (73%) in Phase 1. Notably, repurposed agents originally approved for non-AD indications comprise 33% (46) of current drug candidates and are involved in 37% (68) of current trials.
Figure 5. Agents in clinical trials for treatment of AD[3].
Table 1. Some representative clinical AD drugs and their mechanism of action.
| Drug name |
Highest clinical trial phase |
Mechanism of action |
| Aducanumab |
Approved |
Human monoclonal antibody targeting aggregated Aβ |
| Donanemab |
Approved |
Humanized IgG1 antibody against N-terminal pyroglutamate Aβ |
| Gantenerumab |
Phase Ⅲ |
Fully humanized anti-Aβ IgG1 monoclonal antibody |
| Bapineuzumab |
Phase Ⅲ |
Anti-APP monoclonal antibody |
| Solanezumab |
Phase Ⅲ |
Humanized IgG1 antibody targeting mid-domain of Aβ peptide |
| Crenezumab |
Phase Ⅲ |
Fully humanized anti-Aβ monoclonal antibody |
| Semorinemab |
Phase Ⅱ |
Anti-Tau humanized IgG4 monoclonal antibody |
| Bepranemab |
Phase Ⅱ |
Humanized full-length IgG4 monoclonal antibody |
| Semagacestat |
Phase Ⅲ |
γ-secretase inhibitor |
| Avagacestat |
Phase Ⅱ |
γ-secretase inhibitor |
| Umibecestat |
Phase Ⅲ |
BACE-1 reversible inhibition |
| Elenbecestat |
Phase Ⅲ |
BACE-1 reversible inhibition |
| ALZ-801 |
Phase Ⅲ |
Prevent Aβ42 from forming oligomers |
| Varoglutamstat |
Phase Ⅱ |
Glutaminyl cyclase inhibitor |
| Tideglusib |
Phase Ⅱ |
Tau protein kinase inhibitor with neuroprotective and anti-inflammatory effects |
| TRx0237 |
Phase Ⅲ |
Tau protein aggregation inhibitor |
| ALZT-OP1 |
Phase Ⅲ |
Promote microglia recruitment to plaques, and phagocytosis of Aβ deposits |
Experimental Mouse Models for AD
Preclinical models that faithfully replicate AD pathology are essential for translating mechanistic insights into effective therapies. Among these, experimental mouse models remain the most widely used and well-characterized in vivo systems. They include transgenic models carrying familial AD–associated mutations and non-transgenic models that mimic key pathological features through direct administration of amyloid-β (Aβ). These models enable investigation into hallmark processes such as amyloid plaque formation, tau aggregation, and neuroinflammation, while also serving as critical platforms for preclinical drug screening and therapeutic evaluation.
Aβ Injection Mouse Model
Preparation of Aβ solution: Aβ is dissolved in sterile normal saline, diluted, divided into frozen storage in the refrigerator for use (incubate in an incubator before use).
Anesthesia: Before preparing for surgery, use
isoflurane to anesthetize the mouse or intraperitoneal injection of chloral hydrate.
Positioning the mouse: The anesthetized mouse is placed in a prone position on a brain stereotaxic apparatus. The hair on the top of the mouse head is cut off with scissors, and the skin is cut open with a scalpel after disinfection with iodophor.
Targeted injection: Identify the bilateral hippocampal CA1 regions as the injection sites. Mark the coordinates, and insert the microsyringe needle vertically through the dura to a depth of 2.0 mm. Slowly inject the prepared Aβ solution into the targeted area. After injection, keep the needle in place for 2 minutes to ensure adequate dispersion of the solution into the hippocampal tissue before slowly withdrawing the needle.
Suture: Suture the incision carefully. Administer
gentamicin at the suture site to reduce the risk of postoperative infection. The modeling procedure is then complete.
Transgenic Mouse Models
Generation of Tg2576 Mouse Model[4-6]
The Tg2576 model expresses mutated human APP695 under the control of the prion protein (PrP)
promoter.
a). Constructe the PrP-hAPP695 plasmid containing the neuron-specific promoter.
b). Microinject the transgenic plasmid into fertilized C57BL/6J mouse eggs.
c). Reimplant the injected embryos into pseudopregnant foster mothers.
d). Screen the offspring and perform genotyping to identify Tg2576-positive mice.
Generation of APP/PS1 (APPSwe/PS1ΔE9) Mouse Model[6]
a). Cross C57BL/6J mice expressing chimeric mouse/human APP695 with the Swedish mutation (K670N/M671L) under the control of the hamster PrP promoter with mice expressing human PS1 with exon 9 deletion (PS1ΔE9), also driven by the PrP promoter.
b). Screen the offspring and perform genotyping to obtain APP/PS1 double-transgenic mice.
Generation of 5×FAD Mouse Model[7]
The 5×FAD mouse model combines the Swedish mutation (K670N, M671L) at the β-cleavage site with the Florida (I716V) and London mutations (V717I) at the γ-cleavage site of APP, and two additional mutations within the PSEN1 gene (M146L and L286V), both under control of the murine Thy-1 promoter. The 5×FAD mice exhibit robust AD pathology.
a). Introduce the FAD mutations APP (K670N/M671L + I716V+V717I) and PS1 (M146L+L286V) into APP (695) and PS1 cDNAs by site-directed mutagenesis.
b). Subclone the mutated constructs into exon 2 of the murine Thy1 transgene cassette.
c). Sequence the constructs following standard procedures.
d). Purify the transgenes, combine them in equal proportions and microinject into the pronuclei of single-cell C57/B6×SJL hybrid embryos.
e). Identify and screen founder transgenic mice.
f). Breed the three highest APP and PS1 coexpressing lines with B6/SJL F1 hybrids for analysis screening to obtain stably expressing 5×FAD transgenic mice.
Generation of 3×Tg-AD Mouse Model[8]
The triple-transgenic mouse model of AD (3×Tg-AD), which combines human mutations in APPSwe, PS1M146V, and tauP301L, and develops both amyloid plaques and NFTs in AD-relevant brain regions.
a). Using the pronuclear microinjection technique, co-inject two independent transgene constructs encoding human APPSwe and tauP301L (both under control of the mouse Thy1.2 regulatory element) into single-cell embryos harvested from homozygous mutant PS1M146V knock-in mice.
b). Reimplant the injected embryos into foster mothers.
c). Screen the offspring and perform genotyping to obtain 3×Tg-AD mice.
Summary
Alzheimer's disease (AD) is a complex neurodegenerative disorder primarily affecting the elderly. Its core pathogenesis involve amyloid plaque formation, tau hyperphosphorylation, cholinergic deficits, and other factors such as neuroinflammation and oxidative stress. Clinically, while symptomatic treatments exist, new DMTs targeting amyloid (like Aducanumab and Donanemab) represent significant advances. Research critically relies on experimental models, particularly Aβ-injected and transgenic mice (e.g., 5×FAD, 3×Tg-AD), to replicate pathology and test interventions.
Recent advances in early detection and precise disease staging have greatly improved timely diagnosis and prevention of AD. With new Aβ-targeted therapies, promptly recognizing risk factors and early symptoms is becoming increasingly important. Using multiple diagnostic tools is essential not only for accurate diagnosis, but also for predicting disease progression and enabling earlier intervention, underscoring the need to continue refining our understanding of AD’s complex pathogenesis to support future therapeutic advances.
LinkedIn
Researchgate
Facebook
X
Instagram
Mail
The Pathogenesis of Alzheimer’s Disease (AD)
Drug Development Pipeline for AD
Experimental Mouse Models for AD
