*6.1. Alzheimer's Disease*

#### 6.1.1. Inositols and the Amyloid Pathology

AD dementia is the main clinical entity contributing to the increased prevalence of dementia. In the absence of effective therapies to avoid the deposition of β-amyloid (Aβ) plaques, the main pathogenic factor studied in AD, efforts are also focusing on procedures to delay the cognitive decline associated with this anomalous protein accumulation.

One of the most characteristic features of AD development is the aberrant production and deposition of Aβ peptides, either in Aβ<sup>40</sup> or Aβ<sup>42</sup> fragments. Amyloid precursor protein is subject to protease activity by β-secretase (BACE1) and γ-secretase. Presenilin-1, the catalytic subunit of γ-secretase, is regarded as a major contributor of increased Aβ production. Notably, γ-secretase also contributes to the cleavage of other substrates like Notch, which is important for cell differentiation in embryogenesis and adulthood [190]. The accumulation of Aβ oligomers extracellularly leads to the appearance of Aβ-derived diffusible ligands (ADDLs), which are highly neurotoxic and mediate the disruption of neuronal synapses, the depression of signaling, tau hyperphosphorylation (another well studied factor in AD), the disruption of normal autophagic processes, the generation of ROS and RNS, and cell death [107]. The combination of the above factors is seen to be the cause for cognitive impairment and memory loss.

The tendency of Aβ accumulation seems to be facilitated by binding to peroxidized PI lipids as a consequence of oxidative stress, inducing the conformational secondary structure of the β-sheet [191]. Based on this, it was hypothesized that free myo-inositol and inositol stereoisomers might interfere with Aβ aggregation and fibril formation competing for Aβ-PI lipid binding and stabilizing soluble forms of Aβ [23,24]. Early studies in vitro proved that MI, SI, and EI induced formations of stable β-structures of Aβ42, but these structures did not result in Aβ<sup>42</sup> progression to fibrillar structures at a 1:1 ratio (by weight). The same studies showed that non-fibrillar Aβ<sup>42</sup> oligomers, which may cause neuronal toxicity as well, were non-toxic in the presence of EI and SI at a ratio of 1:20 [24]. DCI, however, was unable to avoid Aβ<sup>42</sup> fibrillar conformation and toxicity, which is suggested to be due to the necessity of the hydroxyl groups to be oriented at positions 1, 3, and 5 or alternatively 2, 4, and 6 in order to stabilize the inositol-Aβ<sup>42</sup> complex structure [24].

Further studies in vivo have shown that the prophylactic administration of SI and EI is effective for preventing cognitive decline and Aβ aggregation at early stages in a mouse model of AD (TgCRND8) and 1-month treatment of SI, but EI could not also reverse disease development [192]. Positive effects with SI were corroborated with decreased astrogliosis and an improvement of synaptic transmission, observed by increased levels of synaptophysin [192]. A later study corroborated higher levels of SI in the brain after ad libitum administration to TgCRND8 mice, with reduced Aβ<sup>40</sup> and Aβ<sup>42</sup> brain levels and decreased plaque formation [18]. Moreover, Aβ cerebroventricular-injected rats with ad libitum access to SI in drinking water had improved performance for learning tasks than non-SI treated rats [193]. Since SI showed better results than EI and other inositol stereoisomers, later studies focused on the SI effect in the same and other murine models of AD, showing improvements in damaged cortical microvasculature caused by Aβ accumulation and improved spatial learning [194–196].

These studies have been substantially reviewed and have raised the not-so-feasible properties of SI (and, to a lesser extent, EI) and its ability to decrease Aβ aggregation and neuronal damage. Some complaints have arisen from the concentration of SI or EI required for interaction with Aβ and the disruption of fibril formation [197]. Other studies have shown the inefficacy of SI at a ratio of 1:10 for inhibiting Aβ aggregation and toxicity, as well as cell death for PC-12 cells and mixed primary rat hippocampal neurons mixed with glial cells [198]. Others have reported that SI binds weakly to other minor-contributing amyloid fragments like Aβ25−<sup>35</sup> (formed in aging brains after Aβ<sup>40</sup> cleavage) [199,200]. Despite these concerns, SI levels are dramatically increased in the brain after ad libitum access, thus being able to interfere in Aβ toxicity, as observed in animal models [18].

These results led to a phase II clinical study of SI (named as ELND005) for mild-to-moderate AD. The administration of SI at a 250 mg dose resulted, after 78 weeks of treatment, in a slight improvement in neurological performance, increased brain ventricle volume, and lower levels of CFS Aβ42, accompanied by higher levels of SI in CSF, as observed in animal models [201]. Further clinical trials have focused on SI effects on agitation and aggression in AD. Despite these results, higher doses of SI resulted in early deaths (4 deceased for 1000 mg SI dose and 5 deceased for 2000 mg SI dose) [201]. SI toxicity may have been caused by renal failure, as uric acid was decreased dose-dependently in SI-treated patients [201]. However, in young healthy subjects, 10-day administration of 2000 mg SI every 12 h also resulted in higher SI levels in blood and CSF with no adverse effects [202]. Further concerns about SI safety have led to the development of SI derivatives, aiming to increase brain penetration, allowing for better efficacy with lower doses of SI for treating AD. A recent report has highlighted AAD-66, a guanidine-appended SI derivative that has improved cognitive performance in a 5xFAD mouse model of AD, concomitant with a reduction in Aβ deposition and glial reactivity as compared to free SI when both were administered ad libitum at the same dose [203].

Although high doses of SI have not been approved for clinical use due to the adverse effects, the search for inositol derivatives, especially those from SI, represent a promising reality in the use of substances derived from natural compounds for the treatment of AD. Along with SI, DPIN is also being tested in clinical trials for AD pathology treatments. DPIN has shown to be a γ-secretase inhibitor that is Notch-sparing (that is, it does not affect Notch cleavage) in vitro [204]. These results show that inositols can directly intervene in the Aβ pathology of different targets (Figure 4).

**Figure 4.** Schematic view of inositol interplay in mechanisms leading to progression of Alzheimer's disease. Epi- and scyllo-inositol stabilize β-sheet conformation of β amyloid (Aβ) products of aberrant APP hydrolysis by β-secretase (BACE1) and γ-secretase. Scyllo-inositol also prevents fibril formation from Aβ oligomers. Pinitol is seen to inhibit γ-secretase cleavage of APP but not Notch products. Brain insulin resistance contributes to amyloid pathology and ultimately to tau hyperphosphorylation, forming toxic microtubule neurofibrillary tangles (NFTs). Mitochondrial dysfunction and presence of Aβ-derived diffusible ligands (ADDLs) worsen the insulin response. d-chiro-inositol and d-pinitol improve insulin sensitivity in the brain, counteracting the development of brain insulin resistance. Inositols prevent early impairment in insulin signaling and the development of vascular dysfunction, which promotes oxidative stress and the release of pro-inflammatory cytokines, ultimately contributing to amyloid pathology in the context of brain insulin resistance in Alzheimer's disease.
