Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by insidious onset and gradual cognitive decline. The pathological hallmarks of AD include extracellular amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles composed of hyperphosphorylated tau protein (Lane et al., 2018). Clinically, AD patients present with memory impairment and progressive cognitive dysfunction, often accompanied by emotional and behavioral abnormalities that severely affect quality of life (Scheltens et al., 2021; Bessey and Walaszek, 2019). Despite extensive research efforts, effective disease-modifying therapies for AD remain unavailable (Weuve et al., 2015).

Increasing evidence indicates that synaptic dysfunction, rather than neuronal loss, is the strongest pathological correlation of cognitive decline in the early stages of AD (Taddei and Duff, 2025; Kenaan and Alshehabi, 2025). Soluble Aβ species, especially Aβ₁₋₄₂, can induce oxidative stress and disrupt synaptic transmission long before plaque deposition becomes evident (Butterfield et al., 2002; Walsh and Selkoe, 2007). Oxidative damage triggered by Aβ not only impairs neuronal viability but also directly interferes with synaptic signaling and plasticity, ultimately leading to learning and memory deficits (Lin and Beal, 2006).

Long-term potentiation (LTP) in the hippocampus is widely recognized as the electrophysiological basis of learning and memory. Impairment of hippocampal LTP, particularly in the CA1 region, has been consistently reported in both AD patients and transgenic mouse models and is considered an early functional manifestation of synaptic pathology (Shankar et al., 2008; Palop and Mucke, 2010). At the structural level, synaptic plasticity is closely associated with dendritic architecture and dendritic spine density, which determine the strength and stability of excitatory synaptic connections (Citri and Malenka, 2008). Loss of dendritic complexity and spine density in the hippocampus has been shown to correlate strongly with synaptic failure and cognitive impairment in AD models (Dorostkar et al., 2015; Barrantes, 2024).

Sulforaphane (SFN) is a naturally occurring isothiocyanate derived from cruciferous vegetables and has attracted increasing attention for its antioxidant and neuroprotective properties (Bobermin et al., 2022; Schepici et al., 2020). Pharmacokinetic studies have demonstrated that SFN is rapidly absorbed and capable of crossing the blood–brain barrier, suggesting its potential application in central nervous system disorders (Gasper et al., 2005; Jazwa et al., 2011). Previous studies have reported that SFN can attenuate Aβ-induced oxidative stress, reduce neuronal injury, and improve learning and memory performance in cellular and animal models of AD (Lee et al., 2013; Zhang et al., 2014). However, most existing studies have focused primarily on behavioral outcomes or molecular markers, and the synaptic mechanisms underlying SFN-mediated cognitive improvement remain incompletely understood.

Notably, it is still unclear whether SFN can systematically restore cognitive behavior, hippocampal synaptic plasticity, and synaptic structural integrity within the same AD model, and whether these functional and structural changes are mechanistically linked. In particular, direct evidence connecting behavioral improvement with recovery of hippocampal LTP, dendritic architecture, and synapse-associated proteins following SFN treatment is lacking.

Therefore, the present study aimed to comprehensively investigate the neuroprotective effects of SFN at multiple levels, ranging from cellular oxidative injury to synaptic function and behavior. Using an Aβ₁₋₄₂–induced oxidative injury model in vitro and APP/PS1 transgenic mice in vivo, we evaluated the effects of SFN on oxidative damage, spatial learning and reference memory, hippocampal LTP, dendritic complexity, dendritic spine density, and the expression of synaptic proteins including PSD-95 and synaptophysin. By systematically correlating behavioral performance with electrophysiological and structural synaptic alterations, this study provides new insight into the synaptic basis of SFN-mediated cognitive improvement and highlights its therapeutic potential for AD.