In this study, we explored the effect of liraglutide, a GLP-1 analog, on the cognitive improvement of 12-month-old AD mice and attempted to delineate the underlying mechanism. We here demonstrated that liraglutide improved the learning and memory abilities of middle-aged AD model mice. Concomitantly, liraglutide increased the number of neurons in AD mice, and improved the ultrastructure of synapses and mitochondrial morphology. We also found that liraglutide significantly reduced the deposition of Aβ, which might be related to the NF-KB/BACE1 pathway, and decreased inflammation. These results offer new insights into the mechanism of action of liraglutide in improving cognitive ability in elderly AD mice.
Of note, 5×FAD mice are known to coexpress a total of five FAD mutations (APP K670N/M671L [Swedish] + I716V [Florida] + V717I [London] and PS1 M146L + L286V) [17]. As such, 5×FAD mice are considered a useful AD model for developing cerebral amyloid plaques at 2 months of age, and exhibit many characteristics of AD, such as reduced synaptic markers, neuronal loss, and memory impairment [17]. A current study showed that the abnormal expression of Aβ was negatively related to the cognitive function of patients with AD [18], and highlighted BACE1 as an important enzyme in the production of Aβ. Under physiological conditions, most Aβ precursor protein (APP) is cleaved by α-secretase to produce N-terminal nontoxic soluble sAPPα fragments and an α-C-terminal fragment (α-CTF). Then, α-CTF is hydrolyzed by the γ-secretase complex, with the whole process not producing Aβ. However, in the brains of patients with AD, BACE1 has been found to be highly activated. Accordingly, APP is cleaved by BACE1 to form an Aβ-containing C-terminal fragment (β-CTF) and a soluble N-terminal fragment. Then, β-CTF is further cleaved by the γ-secretase complex to form Aβ1−42 and Aβ1−40 [19]. Aβ1−42 and Aβ1−40 are the most common fragments of Aβ, and also the most significant pathophysiological characteristics in the brains of patients with AD [20]. Several studies have suggested that pharmacological or genetic reduction in the activity of BACE1 might be a potential strategy for decreasing the brain concentrations of Aβ, thus preventing the progression of AD [21]. In this study, we found that liraglutide improved the cognitive function in 12-month-old AD mice, accompanied by a decrease in the expression of the Aβ protein. In addition, liraglutide was found to reduce the expression of BACE1. Likewise, sitagliptin, a GLP-1 receptor agonist, was reported to enhance the levels of hippocampal GLP-1 and improve the cognitive function of rats with pentylenetetrazol kindling-induced cognitive deficits, potentially corresponding to the modulation of BACE1 [22]. In addition, treatment with GLP-1 was shown to improve learning and memory deficits in type 2 diabetic rats and decrease the mRNA expression of APP and BACE1 [23]. Liraglutide was also found to reduce the activity of BACE1, in turn decreasing the formation of Aβ in insulin-resistant cells [24]. Conclusively, these observations have indicated that liraglutide might be an effective pathway for improving the cognitive function of patients with AD by reducing the expression of BACE1 and Aβ.
Many studies have suggested that in addition to Aβ plaques, the brains of patients with AD exhibit sustained chronic inflammation [25–27], which also contributes to cognitive impairment [28]. In our study, administration of liraglutide in AD mice not only reduced Aβ plaque deposition, but also decreased the levels of the IL-1β and IL-6 proinflammatory cytokines, and the cyclooxygenase-2 (COX-2) inflammatory mediator in the brain and blood. This finding suggested that the regulation of immunoregulatory functions is involved in the mechanism by which liraglutide improves cognition in AD mice. A meta-analysis of 40 studies revealed the detection of increased levels of IL-1β, IL-6, and TNF-α in patients with AD [29]. Besides, treatment with liraglutide was reported to protect against progressive neurodegeneration, chronic inflammation, and amyloid plaque deposition in 2-month-old APP/PS1 mice [30]. Furthermore, another study suggested that inflammation was reduced by 30%, accompanied by a decrease in plaque load and memory improvement in 7-month-old APP/PS1 liraglutide-treated mice [9]. The relationship between inflammation and amyloid plaques is reciprocal. It is known that Aβ binds to microglial-expressed receptors, such as CD36 or CD47, resulting in their activation and production of inflammatory factors and chemokines [31]. The activated microglia surround and take up Aβ, and when their intake capacity is exceeded they resolve to cell death, resulting in the accumulation of released Aβ in the extracellular space, which contributes to Aβ plaque growth and the generation of new plaque clusters in the vicinity of existing plaques, thus further exacerbating inflammation [32]. Liraglutide is known to suppress the production of proinflammation factors by decreasing the activation of microglia [33], or targeting the expressions of proteins, such as IL-1β and TNF-α [34].
The NF-κB pathway has been shown to play an important role in regulating the propagation and elaboration of inflammation by responding to proinflammatory stimuli, such as IL-1 or TNF-α [35]. Activated NF-κB has been found in neurons and glial cells surrounding areas of Aβ plaque deposition in AD brains [36]. Many studies have suggested that Aβ increases the activity of NF-κB, and decreases the activity of NF-κB, resulting in the improvement of the cognitive ability through the decrease in the production of cytokines and chemokines [37–39]. Furthermore, NF-κB is considered an important regulator of the transcription of BACE1, as abundant NF-κB binding sites have been found near the BACE1 promoter [11]. A recent study showed that the levels of expression of NF-κB p65 and BACE1 in the brain tissue of patients with AD were significantly increased [40]. Several studies have also demonstrated that an NF-κb inhibitor could decrease the TNF-α-induced BACE1 transcription, resulting in lower deposition of Aβ [41, 42]. Nonsteroidal anti-inflammatory drugs (NSAIDs), such as indomethacin and flurbiprofen, have been shown to decrease the activity of NF-κB, in turn decreasing the Aβ burden [43, 44]. In our study, the expression of NF-κB, BACE1, and Aβ in AD mice was increased, whereas liraglutide party reversed this trend. Similar to our research, the GLP-1 analog, exendin-4, was reported to promote neuroprotection by reducing neuroinflammatory responses, such as decreasing the activation of NF-κB in lipopolysaccharide-stimulated microglia [45]. Furthermore, studies have shown that liraglutide decreased the activity of BACE1, resulting in the reduced accumulation of Aβ in insulin-resistant cells [46]. These observations indicated that the ability of liraglutide to improve cognition might involve the regulation of the NF-κB/BACE1 pathway, which subsequently reduces the formation of Aβ.