Oxidative stress has long been thought to play a major role in the pathogenesis of AD [18]. The proposal that the beneficial action of EGb761 is mainly due to its free-radical scavenging action is supported by numerous in vitro and in vivo studies [7]. For example, pretreating cerebellar granule cells with EGb761 effectively attenuated oxidative damage triggered by H₂O₂/FeSO₄ [31]. In another study using two AD models, Aβ-expressing neuroblastoma cell lineN2a and Aβ-expressing transgenic Caenorhabditis elegans, EGb761 was found to be able to attenuate the basal as well as the induced levels of H₂O₂-related reactive oxygen species (ROS) [7,32].

In addition to direct attenuation of ROS, EGb761 may also stabilize the cellular redox state by up-regulation of the protein level and activity of antioxidant enzymes [5]. For example, EGb761 was found to be able to increase the protein level and activity of superoxide dismutase (SOD) and catalase in rat hippocampus and rat ileum [5,33,34]. Moreover, activity of glutathione (GSH) reductase and gamma-glutamylcysteinyl synthetase, two enzymes critical for reduction and synthesis of GSH, were also enhanced by EGb761 [5,35,36].

The flavonoid fraction is suggested to be mainly responsible for the antioxidant properties of EGb761. It is proposed that, the flavonoid fraction evokes antioxidant effects via direct scavenging ROS, chelating prooxidant transitional metal ions and increase in antioxidant proteins such as SOD and GSH [7,37–40]. The polyphenol structure of flavonoids (Figure 1A) is thought to be responsible for their antioxidant actions [7,41]. Quercetin [42] and myricetin, two flavonoid constituents with such structure, especially effectively inhibit oxidation of tert-butylhydroperoxide [5]. In contrast, the antioxidant activity of terpene lactones is still in dispute [5]. There are conflicting data regarding the superoxide scavenging activity of bilobalide and the ginkgolides B, C and J [5,43,44]. Ginkgolide A has been demonstrated to lack the ability to scavenge the superoxide [5]. The discrepancy on the antioxidant activity of the terpene lactones may be explained as being due to differences in the type of oxidative stress used as well as the experimental models [5].

Abnormalities in mitochondrial function are suggested to be associated with the pathological changes seen in AD [45]. Recently, EGb761 has been proposed to have direct protective effects on mitochondria. This may also contribute to its antioxidant effects, as mitochondrial respiratory chain is both the major target and the major source of ROS. Using SH-SY5Y cells, we reported that, EGb761 prevented amyloid β peptide (Aβ)-induced mitochondrial dysfunction, and thus reduced intracellular ROS generation [3]. This protective effect was observed with ginkgolide B but not with quercetin [3]. In another study using mitochondria isolated from PC12 cells, EGb761 and bilobalide up-regulated the gene expression of mitochondrial NADH dehydrogenase and decreased stage 4 respiration, whose increase is indicative of oxidative damage in mitochondria [46]. Protective effects of EGb761 on mitochondrial functions have also been demonstrated in in vivo studies. Using two animal models of aging, the senescence accelerated prone 8 mouse strain (SAMP8) and ovariectomized rats, we recently reported that EGb761 treatment effectively prevented the decrease of cytochrome c oxidase (COX) activity, mitochondrial ATP content and mitochondrial GSH content in hippocampi of aged SAMP8 mice and ovariectomized rats [47,48]. Despite these evidences, mechanisms underlying the protective effects of EGb761 and its constituents on mitochondrial function are still unclear and necessitate further studies.

Apoptosis has been implicated in the pathogenesis of various neurodegenerative diseases like AD [7,23,49]. As summarized by Smith and Luo in a review, the anti-apoptotic actions of EGb761 are multifactorial and may act synergistically upon multiple intracellular signaling pathways involved in apoptosis [4,7]. As for possible mechanisms underlying its anti-apoptotic action, EGb761 may maintain the integrity of the mitochondrial membrane; prevent cytochrome c release from the mitochondria, thereby blocking the formation of the apoptosome and the apoptotic caspase cascade; enhance the transcription of antiapoptotic Bcl-2-like protein; attenuate the transcription of pro-apoptotic caspase-12; inactivate pro-apoptotic c-Jun N-terminal kinase (JNK), thereby “turning off” downstream target c-Jun; inhibit the cleavage of the key effector protease caspase-3, thereby blocking the execution of apoptosis and prevent nuclear DNA fragmentation, the molecular hallmark of apoptosis [3,7,50].

The flavonoid fraction of EGb761 may be partly responsible for its anti-apoptotic properties. Despite the evidence that flavonoids prevent cell apoptosis induced by various oxidants [37–41], beneficial effects of EGb761 may go beyond its free-radical scavenging properties. Recently, evidences have accumulated to show that anti-apoptotic effects of flavonoids may be associated with modulation of specific proteins central to intracellular apoptotic signaling cascades such as the mitogen-activated protein kinase (MAPK) cascade. Quercetin, one of the major flavonoid constituents in EGb761 for example, did not inhibit JNK activity and apoptosis induced by hydrogen peroxide and 4-hydroxy-2-nonenal [51–53]. It is proposed that, quercetin exerts its anti-apoptotic effects by inactivation of the peroxide-induced JNK-c-Jun/AP-1 pathway and extracellular signal-regulated kinase (ERK)-c-Fos/AP-1 pathway [53,54]. But low concentrations of quercetin could also promote cellular survival by activation of the MAPK pathway (ERK2, JNK1, and p38), leading to expression of downstream survival genes (c-Fos, c-Jun) and defensive genes (phase II detoxifying enzymes; GSH S-transferase, quinone reductase) [55].

The terpene fraction of EGb761 may also contribute to its anti-apoptotic properties. Bilobalide, ginkgolide B, and ginkgolide J were demonstrated to be able to attenuate apoptosis in chick embryonic neurons caused by 24-h exposure to serum deprivation [56,57]. Bilobalide could also reverse apoptotic damage induced by 12-h staurosporine treatment of chick neurons [57]. In mixed cultures of neurons and astrocytes from neonatal rat hippocampus, bilobalide rescued the neurons from serum deprivation-induced apoptosis, and both bilobalide and ginkgolide B attenuated staurosporine-triggered apoptotic damage [57]. However, ginkgolide A failed to block apoptotic damage either in serum-deprived or in staurosporine-treated neurons [58]. Analysis of DNA fragmentation and the activities of caspase-1- and caspase-3- like protease suggest that bilobalide can block neuronal apoptosis in the early stage by attenuating the elevations of c-myc, p53, and Bax and the activation of caspase-3 [57]. Similarly, the anti-apoptotic effects of ginkgolides are also suggested to be associated with blockage of early signaling events in apoptosis. A recent study demonstrated that ginkgolide B inhibited ethanol-induced apoptotic cell death via suppression of activation of JNK and caspase 3 [59]. Our recent study supported this finding by showing that ginkgolide B protected against Aβ-induced activation of JNK, ERK1/2 and Akt signaling pathways and DNA fragmentation in SH-SY5Y cells [3]. However, there are also contrasting data showing that the terpene constituents of EGb761 have no protective effect on cell apoptosis. For example, bilobalide failed to protect primary adult rat hippocampal neurons against apoptosis caused by a peroxyl radical-generator, 2,2′-azobis-2-amidinopropane [57,60]. In another study, the terpenes of EGb761 could not block hydroxyl radical-induced apoptosis in rat cerebellar granule cells [61]. This discrepancy may be due to the different types of cells and different methods for inducing apoptosis [57].

Despite the anti-apoptotic effects of EGb761 and its pharmacologically active components, the pro-apoptotic action of EGb761 and its constituents (e.g., quercetin and ginkgolide B) has also been demonstrated [50]. Treatment dosage may be one of the vital factors that determine the specific action of EGb761 and its constituents on apoptosis [50] (discussed below).

Inflammation has been implicated in the pathology of AD [62]. Cytokines, acute phase reactants, and other inflammatory mediators have been found to be up-regulated in pathologically vulnerable regions of AD brains [62]. EGb761 has been demonstrated to have anti-inflammatory effects [63–65]. These effects may be attributed to the combined actions of its ginkgolide and flavonoid constituents [65].

The anti-inflammatory action of ginkgolides may be associated with their platelet-activating factor (PAF) -antagonist activity. Substantial evidence suggests the role of PAF as a regulator of cytokines in inflammatory responses [4,66]. Intracerebroventrical administration of PAF in rats could stimulate the synthesis of pro-inflammatory mediator leukotriene, particularly leukotriene C4 [4,67]. PAF can be synthesized in neurons following stimulation with neurotransmitters such as N-methyl-d-aspartic acid (NMDA) and glutamic acid and plays various roles in neuronal functions and brain development [68]. However, increased concentrations of PAF in the brain are also implicated in neurodegenerative diseases such as AD [69,70]. Ginkgolides display very specific and potent antagonist effects against PAF [65]. Intracerebroventrical administration BN-52021 (ginkgolide B) in rats significantly attenuated PAF-induced rise in cerebrospinal fluid peptidoleukotriene levels [4,67]. BN-52021 could also reduce PAF-induced production of the eicosanoid and thromboxane B in a fetal rat brain [4,71]. In support of these findings, our recent study has shown that ginkgolide B could completely block PAF-induced decrease of cell viability in SH-SY5Y cells [3]. In addition to PAF-antagonizing activity, the inhibitory effects of ginkgolides A and B on pro-inflammatory cytokines tumour necrosis factor-alpha and interleukin-1 production in lipopolysaccharide-stimulated rat microglial cultures were also observed [4,72].

On the other hand, the flavonoid fraction of EGb761 reportedly inhibits lipooxygenase that is concerned with the formation of leukotrienes [65]. In addition, in one of our recent studies, we have demonstrated that, a flavonoid constituent of EGb761, quercetin may also be involved in EGb761’s PAF-antagonist activity [3].

The accumulation of Aβ plaques has been proposed to be one of the most prominent mechanisms underlying the pathology of AD [73]. Recently, the role of EGb761 in the protection against the Aβ-induced toxicity has received much attention. A number of recent reports indicate that EGb761 protects against Aβ-induced neurotoxicity by blockage of Aβ-induced events, such as ROS accumulation, glucose uptake, mitochondrial dysfunction, activation of AKT, JNK and ERK 1/2 pathways and apoptosis [3,74,75]. In addition to the protective effects against Aβ, EGb761 has also been shown to prevent amyloidogenesis [76–79]. On hippocampal slices, Colciaghi et al. demonstrated that EGb761 could push amyloid precursor protein (APP) metabolism towards the α-secretase pathway, thereby increasing the release of the soluble form of APP (sAPPα) [76,78]. Using the transgenic AD model Tg2576 mice, the consequence of the capability of EGb761 to influence positively the α-secretase pathway was also assessed in vivo. [78,79]. It was found that, after EGb761 treatment, Tg-2576 mice exhibited an enhancement of spatial learning and memory comparable to wild type mice [78,79]. It has been proposed that, EGb761 inhibits the production of brain Aβ by lowering the levels of circulating free cholesterol, as free circulating and intracellular cholesterol levels could affect APP processing and amyloidogenesis [77,78,80–82]. Despite these evidences, further investigations are needed to identify the major constituents responsible for this antiamyloidogenic effect.

EGb761 could also inhibit the formation of Aβ fibrils [78,86]. It is known that, the β-sheet structure of Aβ fibrils is mainly responsible for the neurotoxicity of Aβ and may also help Aβ escape from clearance via proteolytic degradation [78,83–85]. Thus, inhibiting of formation of β-sheet structure of Aβ fibrils may also help to prevent Aβ toxicity. The interaction of Aβ with transition metal ions, notably iron, zinc and copper, could influence the aggregation state of Aβ. EGb761, through its iron chelating property, may inhibit Aβ fibrils formation [78,86]. EGb761 may also influence the formation of Aβ fibrils by increasing gene expression of transthyretin [87], as transthyretin has been shown to prevent Aβ aggregation in vitro by sequestering Aβ monomers [88]. The inhibitory effect of EGb761 on Aβ aggregation was also observed with bilobalide, ginkgolide J and flavonoid compounds [78,86].

Other mechanisms which may be involved in neuroprotective effects of EGb761 on AD are ion homeostasis, modulation of phosphorylation of tau protein, and induction of growth factor synthesis.

Ca²⁺ dyshomeostasis may be of pivotal importance in mediating neutotoxicity in AD [89,90]. Our unpublished data suggested that, EGb761 could protect against Aβ (1–42)-triggered Ca²⁺ influx via N-methyl-d-aspartic acid receptors. This effect was also observed with its constituents quercetin and ginkgolide B.

In AD brain, hyperphosphorylated microtubule-associated protein tau is aggregated as neuro-fibrillary tangles of paired helical filaments, which is a key event in the pathogenesis of AD [5,91]. Using mRNA microarrays, Watanabe et al. found that EGb761 was able to up-regulate gene expression of microtubuli-associated tau protein as well as of neural protein phosphatase type 1, a serine/threonine protein phosphatase known to dephosphorylate hyperphosphorylated tau protein, in the hippocampus and cortex of normal mice [5,87].

Altered levels of nerve growth factor (NGF) have been detected in AD brains [92–94]. EGb761 was reported to be able to up-regulated mRNA expression of NGF such as growth hormone and prolactin in mouse cortex [5,94]. Furthermore, the mRNA and protein expression of glial-derived neurotrophic factor and vascular endothelial growth factor in cultured rat cortical astrocytes could also be up-regulated by a terpenoid constituent of EGb761, bilobalide [5,95].