According to the amyloid hypothesis (Hardy and Higgins, 1992), the primary cause of AD is the accumulation and deposition of oligomeric or fibrillar Aβ peptides. The Aβ peptide consists of 38–42 amino acids that are derived from amyloid precursor protein (APP); a transmembrane protein that has two competing pathways. In the non-amyloidogenic pathway, it is cleaved by α-secretase, to produce the secretory fragment sAPPα (see Figure 2). The candidate enzymes for α-secretase which are from the a-disintegrin and metalloprotease (ADAM) family include ADAM17, ADAM9, and ADAM10 (Buxbaum et al., 1998; Lammich et al., 1999; Fahrenholz et al., 2000; Asai et al., 2003). Among these enzymes, ADAM10 is suggested as the most physiologically relevant α-secretase in neurons (Anders et al., 2001; Kuhn et al., 2010). In the amyloidogenic pathway, APP is initially cleaved by β-secretase (BACE1), then γ-secretase, resulting in the generation of Aβ peptides (Gandy et al., 1994). The non-amyloidogenic pathway is beneficial since sAPPα has neuroprotective activity (Corrigan et al., 2011). In contrast, in the amyloidogenic pathway, an over-production of Aβ and its accumulation results in cytotoxicity (Chasseigneaux and Allinquant, 2012; Paroni et al., 2019). The length of the Aβ peptide influences this toxicity, where Aβ₄₂ (42 amino acids) is more cytotoxic than Aβ₄₀ and Aβ₄₃ (Fu et al., 2017). This is because Aβ₄₂ has a higher hydrophobicity and thus higher propensity to aggregate by hydrophobic bonding into toxic oligomers compared with Aβ₄₀ peptide (Vion et al., 2018). Several inherited and environmental factors such as APP, presenilin 1 (PSEN1), and presenilin 2 (PSEN2), gene mutations, deficit Aβ clearance, oxidative stress and mitochondrial dysfunction might be contributing factors to the over-production and accumulation of Aβ (Mao and Reddy, 2011; Hernández-Zimbrón and Rivas-Arancibia, 2015; Zuo et al., 2015; Paroni et al., 2019). Although amyloid deposition is always seen in AD patients, its pathogenic role is still unclear (Modrego and Lobo, 2019). While many questions still remain unanswered regarding the pathogenesis of AD, the amyloid hypothesis is still the most accepted theory to describe the associated neuropathological events.

Tau proteins are phosphoproteins present in all cells of the central nervous system (CNS; Lionnet et al., 2018). The main function of tau is the modulation of microtubule stability which forms the main pathway for intracellular protein trafficking (Mandelkow, 1998; Buée et al., 2000). But in AD, abnormal hyperphosphorylation of tau leads to its dysfunction, resulting in impairment of the transport system, the cytoskeleton, intracellular signaling, and mitochondrial integrity (Mandelkow, 1998; Iqbal et al., 2005). Hyperphosphorylated tau proteins dissociate from the microtubule (Figure 3) and bind with each other, forming paired helical filaments (PHFs). These accumulate, resulting in the characteristic NFTs seen in AD pathology (Gamblin et al., 2003).

There are other alternative hypotheses where Aβ plaques and NFTs may be formed independently and may be the products of dementia and not the cause (Hardy and Selkoe, 2002; Selkoe and Hardy, 2016). To date, there is no satisfactory hypothesis that can fully explain the exact mechanism of Aβ and tau accumulation, aggregation, and subsequent toxicity.

The neuroinflammation is a response of the innate immune system within the brain as shown by an increased level of activated microglia and astrocytes, activated complement proteins and cytokines (Heneka et al., 2015; Zhang and Jiang, 2015). In AD, Aβ plaques and NFTs exacerbate any chronic inflammatory state, resulting in the increased action of cytokines (interleukin 1, tumor necrosis factor), prostaglandins, growth factors, thromboxanes, and ROS. These, in turn, enhance the APP processing, increasing Aβ₄₂ levels in brain (Meraz-Ríos et al., 2013). Aβ also activates proinflammatory cytokines and some pro-inflammatory enzymes, such as cyclooxygenase-2 (COX-2), inducible nitric oxide synthase (iNOS) and nuclear factor kappa B (NF-κB). This Aβ-linked inflammatory response has been claimed to lead the neuronal damage in AD (Meraz-Ríos et al., 2013; Heneka et al., 2015; Zhang and Jiang, 2015). Accumulating evidence suggests that neuroinflammation is a major contributor of AD onset and progression (Heneka et al., 2010). Long-term consumption of nonsteroidal anti-inflammatory drugs (NSAIDs) can delay the onset or progression of AD, which also supports the role of neuroinflammation in AD (Ali et al., 2019).

A growing body of literature indicates that oxidative stress is another pathophysiological feature of the AD brain (Good et al., 1996; Agostinho et al., 2010). Oxidative stress is the imbalance between the formation and detoxification of reactive oxygen species (ROS; Adwas et al., 2019). These ROS are normally produced as by-products of oxygen metabolism which utilizes both free radicals and non-free radical oxygen intermediate species, such as hydrogen peroxide (H₂O₂), superoxide (O2−), hydroxyl radical (^•OH), and singlet oxygen (¹O₂). These ROS are known to damage many biomolecules including DNA, RNA, protein, and lipids (Pham-Huy et al., 2008). Studies suggest that oxidative stress could: (a) be a consequence of Aβ deposition; (b) induce the production of Aβ; (c) be a combination of both: (a) and (b) (Sonnen et al., 2008; Tamagno et al., 2012). ROS are produced in vivo during oxidation and are involved in the progression of various health problems including cellular aging, mutagenesis, cardiovascular problems, diabetes, and neurodegeneration (Halliwell and Gutteridge, 1999; Moskovitz et al., 2002).

A high intake of foods rich in antioxidants may be beneficial to attenuate the ROS-associated problemsbased on the result of human dietary intervention studies (Lobo et al., 2010; Liu et al., 2018).

Mitochondria are responsible for energy homeostasis in cells. Their dysfunction may contribute to the progression of several diseases including cancer, cardiovascular diseases, diabetes and neurodegenerative diseases (de Moura et al., 2010; Wen et al., 2016). A large body of research indicates that dysfunctional mitochondria play an important role in the pathogenesis of AD (Bhatti et al., 2017).

The initial level of mitochondria functions as well as its rate of decline influences AD onset and progression (Swerdlow et al., 2014). When mitochondrial function falls below a critical threshold, abnormal tau phosphorylation processes, amyloid plaque generation, synaptic degeneration, and oxidative stress can result (Lezi and Swerdlow, 2012). Several essential mitochondrial functions such as biogenesis, fission/fusion, and bioenergetics are also associated with AD. This makes mitochondrial dysfunction an important factor to consider in AD pathogenesis and its prevention (Lezi and Swerdlow, 2012; Nicolson, 2014; Flannery and Trushina, 2019).

There is compelling evidence that smoking, high blood pressure, diabetes, high cholesterol, and obesity significantly increase the risk of AD (Prince et al., 2014). Hormones (testosterone and estrogen) can have a neuroprotective effect through regulating Aβ, thus, age-related decline in these hormones can affect cognitive ability and therefore increase the risk of developing AD (Verdile et al., 2014). Traumatic brain injury (TBI) is also reported to increase AD risk (Alzheimer’s Association, 2018).

Lifestyle factors including a healthy diet; adequate physical exercise, good sleep hygiene and cognitive training have been shown to reduce the risk of developing AD (Bauer and Morley, 2018). Conversely psychological factors (e.g., depression, anxiety, and stress) and vascular risk factors contribute to an increased risk of AD. A diet with high PP and high antioxidant activity can thus be considered as an approach to assist the prevention of chronic diseases, especially AD.

Age is the main risk factor for AD. As people age, the risk of AD increases exponentially, as shown in several population based studies (Corrada et al., 2010). Apolipoprotein E (APOE) ε4 allele is the major genetic risk factor, which increases the probability of developing AD (Thakur et al., 2019). The effects of APOE ε4 on cognitive ability are variable from person to person (Prince et al., 2014; O’Donoghue et al., 2018). The APOE gene is present in chromosome 19 (Dementia Australia, 2018b). In humans, there are three common alleles: ε2, 3 and 4. Each individual carries two apolipoprotein genes which can be the same type (ε2, 2; 3, 3 or 4, 4), or a combination of two types (ε2, 3; 2, 4; 3, 4; Dementia Australia, 2018b). Individuals with at least one ε4 have a 2 to 3-fold risk of AD while those with two ε4 alleles (4, 4) rarely escape the disease. Compared to the other APOE alleles, the higher risk of developing AD in ε4 alleles is associated with an earlier age of AD onset (Alzheimer’s Association, 2018). This higher risk is three fold for one copy of ε4 allele and 12 fold for two copies (Alzheimer’s Association, 2018). In contrast, APOE ε2 carriage has a neuroprotective effect relative to APOE ε3 and APOE ε4. Carrying the double-barrelled APOE ε4 combination is fortunately uncommon, affecting only about 2% of the population, whereas about 25% of people carry a single copy of APOE ε4 (Alzheimer’s Association, 2018).

Polyphenols are naturally occurring compounds and secondary metabolites of plants mostly produced in response to major stress (Pandey and Rizvi, 2009; Isah, 2019). They protect plants against biotic (living beings present in an ecosystem e.g., fungi, bacteria, and protists), and abiotic (non-living components e.g., water, soil, air, sunlight, temperature, and minerals) stressors (Rauf et al., 2019) acting as antioxidants, antimicrobials, and photo-absorption molecules. Thus, they defend plants from pathogens, ultraviolet radiation damage and predators such as insect pests (Beckman, 2000). Moreover, they are involved in the structural strength of plants during growth (Pandey and Rizvi, 2009). Polyphenols have received special attention from researchers due to their antioxidant activities which enable them to scavenge free radicals formed during the pathological processes of diseases such as cancer, cardiovascular diseases, and neurodegenerative disorders (Lakey-Beitia et al., 2015). They also have anti-inflammatory activity that is important in reducing oxidative stress thus conferring potential protective effects against the neurodegenerative process (Masci et al., 2015).

Polyphenols have demonstrated that they provide their neuroprotection through antioxidant, cholinergic, Aβ, and tau aggregation pathways in vitro and in vivo (Omar et al., 2017). The PP attenuate Aβ toxicity and oxidative stress in neurons by decreasing the Aβ aggregation and increasing the scavenging of free radicals, as shown in animal and cell culture studies (Dore et al., 1999; Agostinho et al., 2010; Mathiyazahan et al., 2015; Bai et al., 2017; Hwang et al., 2017). Polyphenols donate electrons to the free radicals to neutralize them, which is important to decrease the levels of ROS within cells (Lobo et al., 2010). Additionally, there is some evidence from cellular and animal model studies that PP may inhibit the Aβ₄₂ toxicity (Bastianetto et al., 2008; Hugel and Jackson, 2015). Decrease in the hyperphosphorylation of tau protein, the formation of NFTs, and inflammation in in vitro and in vivo studies upon addition of PP has also been demonstrated (Mendes et al., 2018).

The basic structure of PP includes two aromatic rings linked through a pyran ring (Ross and Kasum, 2002). There structures are very complex, with the two main categories of PP are flavonoids and non-flavonoid compounds (El Gharras, 2009; see Figure 4). Flavonoids contain 15 carbon atoms. They are soluble in water and characterized by two benzene rings connected through a three-carbon chain. Flavonoids are sub-divided into anthoxanthins (flavones, flavonols, flavanols, isoflavonoids, flavanones), and anthocyanins (Lakey-Beitia et al., 2015). Non-flavonoid PP are phenolic acids, stilbenes, curcuminoids, lignans, and tannins (Lakey-Beitia et al., 2015).

It has been hypothesized that the, the anti-amyloidogenic activity of PP is due to their physicochemical features, including the aromatic rings, molecular planarity, hydrogen bond formation, internal double bonds, and molecular weights below 500 g/mol (Lakey-Beitia et al., 2015). All these features are important for the inhibition of the amyloidogenic APP processing to reduce amyloid load, by activating α-secretase and inhibiting β- and γ-secretase (Lakey-Beitia et al., 2015).

Several in vitro and in vivo studies reported that PP-rich extracts from plants, like fruits and herbs, possess neuroprotective activities (Dai et al., 2006; Rossi et al., 2008; Loef and Walach, 2012; Hassaan et al., 2014; Dal-Pan et al., 2017; Omar et al., 2017; Polito et al., 2018). For instance, in vitro, in vivo and clinical studies showed the neuroprotective action of berry fruits through their polyphenolic contents (Vepsalainen et al., 2013; Wong et al., 2013; Subash et al., 2014). Other in vitro and in vivo studies indicate that pomegranate which is also rich in PP has the potential to attenuate AD progression by its anti-inflammatory and anti-Aβ accumulation activity (Hartman et al., 2006; Rojanathammanee et al., 2013). Moreover, extracts of other PP-rich fruits such as apple, banana, orange, grape, citrus fruit, and walnut have been also shown to inhibit Aβ neurotoxicity and oxidative stress as demonstrated by several in vitro studies (Chauhan et al., 2004; Heo et al., 2008; Toda et al., 2011; Lian et al., 2016; Braidy et al., 2017a). In one animal model study, PP-rich fruits such as Palm fruit could attenuate oxidative stress (Subash et al., 2015). Of particular interest to this current review is that several in vivo studies have reported cognition and memory enhancement activities of grapes, citrus fruit, walnut and buckwheat extracts (Wang et al., 2010; Choi et al., 2013; Lian et al., 2016; Braidy et al., 2017a; Pandareesh et al., 2018). in vitro investigation of the effect of a PP-rich extract of perennial buckwheat reported attenuation of Aβ toxicity in plasma (Liang et al., 2017). Another in vitro study using PP-rich extract of the herb Patrinina villosa Juss has shown a significant inhibitory effect on Aβ plaque aggregation (Bai et al., 2017). A cell culture study on twenty different South African medicinal PP-rich plants against AD reported the effectiveness of Xysmalobium undulatum, Cussonia paniculata, and Schotia brachypetala in decreasing the production of Aβ in comparison to other investigated extracts (Thakur et al., 2019). Moreover, based on the dietary intervention animal study of Ingale and Kasture (Ingale and Kasture, 2017), PP rich extract of purple passionflower could enhance cognitive function. Animal model studies, Capparis spinose, Caesalpinia crista, Iris germanica, and Paeonia suffruticosa could attenuate inflammation and Aβ aggregation through their polyphenolic contents, and make positive changes in cognition and memory (Costa et al., 2016; Gu et al., 2016; Borhani et al., 2017).

In vitro studies have reported that dietary drinks such as the crude juice of broccoli sprouts (Masci et al., 2015), tea (Polito et al., 2018), coffee (Ishida et al., 2018), and red wine (Dhir, 2018) are protective against Aβ-induced cytotoxicity and apoptotic cell death. They have been shown to attenuate mitochondrial dysfunction and hyperphosphorylation of tau proteins through their polyphenolic content (Lakey-Beitia et al., 2015; Sawikr et al., 2017; Polito et al., 2018).

Whole grain consumption as part of a healthy diet has been reported to be protective against several chronic diseases (Miller et al., 2000; Slavin, 2003; Aune et al., 2016). The health benefits of whole grains are in part due with their PP and the associated antioxidant activity (Slavin et al., 1997; Miller et al., 2000; Slavin, 2003; Tian et al., 2019).

Among whole grains, sorghum has some unique features that make it very attractive for neuroprotection studies. It is an inexpensive and abundant grain with a wide range of varieties, some of which are very high in PP content (including PP that are very rarely found in other plant food) and antioxidant activity. Several in vitro and in vivo studies have reported beneficial effects of sorghum PP on chronic diseases such as diabetes and cardiovascular disease, both of which are as risk factors of AD (Kim and Park, 2012; Suganyadevi et al., 2013; Stefoska-Needham et al., 2015; de Morais Cardoso et al., 2017; Moraes et al., 2018).

Sorghum (Figure 5) is the fifth most-produced cereal crop in the world (Awika and Rooney, 2004). It is adaptable to grow in drought and hot climates. Thus, it is usually grown in warm semi-arid and arid areas across the globe (de Morais Cardoso et al., 2017). Sorghum grain has been mostly used as livestock feed and in the biofuel industry (de Morais Cardoso et al., 2017). Sorghum is gluten-free and low-fat while being high in protein and fiber. It has a high antioxidant and anti-inflammatory potential due to its bioactive compounds such as polyphenolics (Awika and Rooney, 2004; de Morais Cardoso et al., 2017).

Sorghum grain has been classified into three different groups based on extractable tannin content. Sorghum type I (low tannins extracted by 1% acidified methanol), type II (tannins extractable in 1% acidified methanol and not methanol alone), and type III (tannins extractable in both acidified methanol and methanol alone; Awika and Rooney, 2004).

Another common way of sorghum classification is based on its grain color and its total polyphenols. Sorghum has varieties with pigmented and non-pigmented precarps. White sorghum has no tannins or anthocyanins and has a very low level of total PP. Red sorghum (red pericarp) has a considerable amount of extractable PP without any tannins. Black sorghum (black pericarp) has a large amount of anthocyanins and finally, the brown sorghum (pigmented testa, different degrees of pericarp pigmentation) contains significant levels of tannins (Awika and Rooney, 2004). The concentration of flavonoid in sorghum is related to the pericarp color, pericarp thickness, and presence of testa (Taleon et al., 2012). However both environmental and genetic factors influence the phenolic level and profiles of sorghum grain (Awika and Rooney, 2004).

A limited numbers of studies such as Awika et al. (Awika and Rooney, 2004; Yang et al., 2009; Awika, 2017; Girard and Awika, 2018) and Cardoso (de Morais Cardoso et al., 2017) have investigated the potential benefit of sorghum on health and disease prevention. According to their findings, sorghum should be considered as a health-beneficial grain, not just a low-value cereal grain. Sorghum has shown a positive impact on glycemic control, colonic microbiota, cholesterol attenuation, cardiovascular disease, anti-mutagenicity, and anti-inflammatory activity (Stefoska-Needham et al., 2015; de Morais Cardoso et al., 2017) which are all risk factors of AD. Below, we summarize information on the effect of sorghum on some chronic disease and their relation to AD.

Some varieties of sorghum possess up to 6% (w/w dry basis) of phenolic compounds which is the highest level in any cereal grain (Su et al., 2017). Almost all classes of the phenolic compounds are present in sorghum (Awika and Rooney, 2004) including phenolic acids, flavonoids, tannins, and stilbenes (Tables s 1, 2; Vanamala et al., 2017). The bran fraction of sorghum has the highest concentration of PP thus processing to remove the bran (decortication), will notably decrease the potential health benefits of the grain and therefore un-decorticated sorghum (whole grain) is recommended for consumption (Girard and Awika, 2018; Ashley et al., 2019).

Based on the literature, almost all the polyphenolic compounds of the various sorghum genotypes have antioxidant activity which may be effective for the attenuation of AD pathological hallmarks (Awika and Rooney, 2004). Among all the PP of sorghum (Tables s 1, 2), caffeic acid, trans-resveratrol, quercetin, catechin, cinnamic acid, cyanidin, apigenin, and kaempferol have gained the most attention for AD prevention and treatment (Rossi et al., 2008; Jabir et al., 2018). Snow et al. (2019) showed that PP exert their anti-AD properties primarily through prevention of aggregation of Aβ fibrils and tau protein NFTs. The presence of hydroxyl groups adjacent to aromatic rings may enhance the inhibition of Aβ/tau aggregation (Snow et al., 2019) by reducing the secondary folding of β-sheet structures which are characteristic of Aβ plaques and NFTs. For example, this property is found in proanthocyanidins, which are highly effective in reducing plaques and tangles in the brain as well as in improving short-term memory. Concluding from this article, sorghum PP such as epicatechin, luteolin, quercetin, etc. with adjacent hydroxyl groups can provide Aβ/tau disaggregation (Figures 6, 7).

Collectively the functions of sorghum PP include a combination of antioxidant, anti-amyloid, anti-tau, anti-inflammatory, AchEI and BChEI activities (Rossi et al., 2008; Omar et al., 2017; Jabir et al., 2018).

In summary, sorghum has a diverse polyphenolic profile depending on its genotype. According to in vitro and in vivo studies, several PP of sorghum have the potential to act as an anti- AD agent through different pathways such as free-radical scavenging, inhibition of Aβ and tau aggregation. The neuroprotection potential of important single PP of sorghum is illustrated in the subsequent sections.