Pharmacotherapy of Alzheimer’s Disease: Seeking Clarity in a Time of Uncertainty

Alzheimer’s disease (AD) is recognized as a major health hazard that mostly affects people older than 60 years. AD is one of the biggest medical, economic, and social concerns to patients and their caregivers. AD was ranked as the 5th leading cause of global deaths in 2016 by the World Health Organization (WHO). Many drugs targeting the production, aggregation, and clearance of Aβ plaques failed to give any conclusive clinical outcomes. This mainly stems from the fact that AD is not a disease attributed to a single-gene mutation. Two hallmarks of AD, Aβ plaques and neurofibrillary tangles (NFTs), can simultaneously induce other AD etiologies where every pathway is a loop of consequential events. Therefore, the focus of recent AD research has shifted to exploring other etiologies, such as neuroinflammation and central hyperexcitability. Neuroinflammation results from the hyperactivation of microglia and astrocytes that release pro-inflammatory cytokines due to the neurological insults caused by Aβ plaques and NFTs, eventually leading to synaptic dysfunction and neuronal death. This review will report the failures and side effects of many anti-Aβ drugs. In addition, emerging treatments targeting neuroinflammation in AD, such as nonsteroidal anti-inflammatory drugs (NSAIDs) and receptor-interacting serine/threonine protein kinase 1 (RIPK1), that restore calcium dyshomeostasis and microglia physiological function in clearing Aβ plaques, respectively, will be deliberately discussed. Other novel pharmacotherapy strategies in treating AD, including disease-modifying agents (DMTs), repurposing of medications used to treat non-AD illnesses, and multi target-directed ligands (MTDLs) are also reviewed. These approaches open new doors to the development of AD therapy, especially combination therapy that can cater for several targets simultaneously, hence effectively slowing or stopping AD.


INTRODUCTION
Alzheimer's disease (AD) is a chronic neurodegenerative disease that usually affects people older than 60 years. The etiologies of many neurodegenerative diseases are not limited to a single gene or pathway, but are rather an intricate network of causatives, including neuroinflammation, oxidative stress, mitochondrial dysfunction, protein misfolding, and aggregation that can lead to cell death Kamil et al., 2019). Various mechanisms were associated with the sporadic form of AD, which accounts for most AD cases, whereas mutations of three genes, including amyloid precursor protein (APP), presenilin 1 (PS1), and presenilin 2 (PS2) are heavily linked with familial AD cases (Newman et al., 2017). The irreversible symptoms of AD, such as progressive deterioration of intellect, memory, and attentiveness, are noticeable when both NFTs and Ab plaques have disseminated through the limbic system (Canter et al., 2016).
Ab plaques-induced brain atrophy begins with the loss of synapses and enlargement of the ventricles (Ramos-Rodriguez et al., 2017), whereas NFTs are often associated with grey matter loss (Bejanin et al., 2017). Progressive loss of cortical interneurons and specific neurotransmitter pathways such as acetylcholine (ACh), noradrenaline (NA), and serotonin (5-HT) are also reported in AD (Ribeiro et al., 2017). In the early stages of AD, typically around age 70, cells in the hippocampus start to degenerate, causing mild forgetfulness of recent events and familiar names, and also difficulty in solving simple mathematical problems (Selkoe, 2011;Bie et al., 2018). After 10 years, atrophy of the cerebral cortex occurs in moderate AD stage, resulting in a decline in language, emotional outbursts, impaired ability in conducting simple tasks such as combing hair and buttoning shirts, and an inability to think clearly. In advanced stages of AD, where more nerve cells have undergone degeneration, patients are often agitated, wandering, and unable to recognize faces and communicate (Braak and Tredici, 2018).
AD-related deaths have markedly increased over the past two decades, but a cure remains elusive (Alzheimer's Association, 2019). Treatment options for AD that are approved by the Food and Drug Administration (FDA) do not have a curative effect or are able to slow down the progression of the disease (Folch et al., 2018a). The most commonly prescribed drugs are acetylcholinesterase inhibitors (AChEIs), such as tacrine, donepezil, rivastigmine, and galantamine, and N-methyl-D-aspartate receptor (NMDAr) antagonists, such as memantine (Morsy and Trippier, 2018; Table 1). Tacrine was eventually discontinued due to its hepatotoxicity, and tacrine hybrids are now being studied (Sameem et al., 2017). AChEIs delay the metabolism of ACh by inhibiting acetylcholinesterase (AChE) as AD patients have a deficiency of ACh (Kumar et al., 2015). Meanwhile, memantine prevents excitotoxicity by blocking NMDAr's activation (Marttinen et al., 2018).
Heterogeneity in AD pathogenesis impedes the development of curative strategies. Three neuropathological mechanisms postulated in AD are: i) formation of extracellular Ab plaques by insoluble amyloid proteins' aggregate, ii) formation of NFTs (disorganized bundles of filaments in the neuronal cytoplasm) by hyperphosphorylated tau proteins, and iii) neuronal loss as the aftermath of the Ab plaques and NFTs (Revett et al., 2013). Thus, many drugs failed to improve cognition in mild-to-moderate AD patients as the drugs target a single pathology without acknowledging other neurological insults (Jobke et al., 2018). Hence, a multi-target approach is being highly investigated in clinical trials to synergistically target distinct pathways and ameliorate AD (Cummings et al., 2019).

AMYLOID PLAQUES
Overproduction and reduced clearance of Ab 42 monomers cause the deposition of Ab plaques that eventually leads to alteration in downstream neurobiological events (Takahashi et al., 2017). Ab plaques cause catastrophic damage to cellular membranes' integrity through the formation of the membrane's pore and the reduction of the membrane's fluidity, hence, leading to neuronal death (Yasumoto et al., 2019). Ab plaques also trigger the activation of microglia and astrocyte as an inflammatory response, alter the neuronal calcium homeostasis causing oxidative injury, and disrupt the protein kinase and phosphatase-related pathways, resulting in hyperphosphorylation of tau and formation of NFTs (Reiman, 2016). Furthermore, self-propagation of Ab 42 plaques and hyperphosphorylated tau, via a prion-like mechanism, may exaggerate the synaptic dysfunction, neurotransmitter deficits, and neuronal loss in the brain (Goedert, 2015).
Although Ab plaques alone may not be adequate in causing the transmission of pathological tau, amyloid cascade hypothesis suggests that deposition Ab plaques is the triggering factor for the cognitive deteriorations in AD . Hence, future drug development should seek to determine whether a single-target therapy targeting Ab is sufficient to treat AD or whether a combination therapy between anti-Ab and anti-tau is needed .

NEUROFIBRILLARY TANGLES
Intracellular NFTs are the deposits of insoluble proteins in neuronal cell bodies (Vanden Dries et al., 2017). Tau is a cytoskeletal microtubule-associated protein (MAP) that is phosphorylated at three sites -serine (S), threonine (T), and at residues adjacent to proline -and binds at the microtubules (MTs) to sustain the MTs' stability and integrity (Pradeepkiran et al., 2019). The toxicity of tau can impair neuronal function depending on its post-translational modifications. The most potent phosphorylations of tau take place at T231, S235, and S262, which results in the loss of tau's ability to bind to MTs, leading to tau self-assembly into paired helical filaments (PHF) (Iqbal et al., 2018).
Phosphorylation of tau detaches it from MTs to allow the intracellular transportation of subcellular organelles such as mitochondria and lysosomes from the nerve terminals to the cells' soma through secretory vesicles (Pradeepkiran et al., 2019). Hyperphosphorylation of tau sequesters the normal tau in which it may excessively impair tau binding and destabilize MTs, thus, impairing the axonal transport causing neurodegeneration through synaptic starvation, neurite outgrowth, and neuronal death (Minjarez et al., 2013). Hyperphosphorylated tau tends to misfold and forms PHF which eventually aggregates to form NFTs as a defense mechanism in the cell soma (Gandini et al., 2018).
In contrast to Ab pathology, which causes hyperactivity of neurons, tau silences the neurons (Busche et al., 2019). This provokes the question on how the coexistence of Ab and tau pathologies causes neurodegeneration in AD. From the fully eradicated neuronal hyperactivity and drastic decline of cortical activity in rats with both Ab and tau pathologies, it can be concluded that deposition of Ab plaques may be the triggering factor that sparks other AD etiologies, but tau pathology is the one dominating the aftermath effects of this dual proteinopathies in AD. It is tau pathology that determines the cognitive status in AD compared to Ab pathology, which is another solid reason for the constant failures of Ab drugs. The combination of antiamyloid and anti-tau is crucial, as suppressing gene expression of tau is less effective in restoring the neuronal impairments in the presence of Ab plaques (DeVos et al., 2018).

CURRENT DRUGS TARGETING Ab -FAILURES
According to the updated AD drug development pipeline in 2018, although more than 50% of drugs in Phase III trials are targeting Ab, there is still a steep 40% decline from year 2017 to 2018 in anti-Ab drugs in Phase I and II trials, which manifests the shift in AD research following the repetitive failures of anti-Ab drugs (Mullane and Williams, 2018) (Table 2). Reducing the generation of Ab 42 , inhibiting the aggregation of Ab plaques, or increasing the rate of Ab clearance from the cerebrospinal fluid (CSF) and brain are the common approaches of anti-Ab drugs (Scheltens et al., 2016). At present, the complexity of AD's pathogenesis is vaguely understood, which may involve numerous other proteins beside Ab and various biological pathways (Doig et al., 2017). This multifactorial AD pathogenesis is most probably the main reason for the repetitive failures of anti-amyloid drugs because a single target treatment may not be able to cater for all the altered pathways involved in the neurodegenerative events (Selkoe, 2019).
Initially, these anti-Ab agents show potential curative effects through effective Ab clearance in CSF and the brain during the early stage of drug development. However, the Ab level in mildto-moderate AD patients or even prodromal AD patients may have reached the threshold of irreversible neurotoxicity, which is the main reason for the failure of many anti-Ab drugs once proceeded to phase III trials (Dobrowolska Zakaria and Vassar, 2018). This is probably due to lack of AD biomarkers in the past to ensure early detection and recruitment of potential AD patients for clinical trials (Folch et al., 2018b). Plus, Ab plaques accumulate at a slow rate which may provide a large time window for a potential intervention that could either enhance the clearance or hinder the accumulation of Ab insoluble proteins before brain atrophy and memory impairment commences (Villemagne et al., 2013). According to PET images taken by Vlassenko and colleagues, significant changes in CSF Ab level and Ab deposition in the brain of AD patients were reported, respectively 25 and 15 years prior to the clinical representation of AD symptoms (Vlassenko et al., 2012). This finding suggested an optimal time for early intervention of AD; hence, early treatment can be given to the presymptomatic AD patients who should have been recruited for clinical trials rather than symptomatic AD patients. This is mainly because symptomatic AD patients usually have irreversible synaptic loss and neuronal deaths. Therefore, better recruitment of patients with earlier stage of neurodegeneration and more consistent pathology underlying their AD to participate in clinical trials may generate more beneficial clinical outcomes (Briggs et al., 2016). The targeted protein to minimize the generation of Ab42 is b-secretase 1 (BACE1), which is involved in the first proteolytic cleavage of the APP protein and also g-secretase which plays a role in the second cleavage in order to produce the Ab42 protein.Verubecestat, a BACE1 inhibitor, was discontinued from the phase III trial due to its lack of efficacy and inability to establish a positive risk/benefit ratio towards mild-to-moderate and prodromal AD patients, although significant reduction of Ab in the patients' CSF and brains were achieved during the trial (Egan et al., 2018). This finding emphasized that solely targeting amyloid may not be an appropriate strategy in treating AD. Lanabecestat is a selective BACE1 inhibitor with satisfying blood-brain barrier (BBB) penetration, high potency and permeability, and slow off-rate that is critical for its efficacy (Eketjall et al., 2016). However, lanabecestat was discontinued from the trials due to its unlikeliness to meet the primary end points in mild-to-moderate and prodromal AD patients based on a recommendation from the data monitoring committee (Panza et al., 2019). Meanwhile, a phase II/III trial of a non-selective BACE1, atabecestat, was halted as the benefit/risk ratio was no longer favorable due to chronic elevations of liver enzymes observed during the trial (Taylor, 2018).
The clinical trial of semagacestat was rushed without a strong foundation of knowledge on the compound's physiological, structural, and functional properties (De Strooper, 2014). It was even more perplexing when the trial of semagacestat continued to the next phase without any significant results from the previous trial and was terminated before the completion of phase III trial. The drug was not even as effective as the placebo, aggravated cognitive deterioration at higher doses, and demonstrated high incidence of skin cancer and infections in the study group (Doody et al., 2013). The drug's adverse effects may potentially be linked to the altered Notch signaling pathway, where Notch was one of the alternate substrates for g-secretase, important for cell differentiation. Blocking g-secretase through semagacestat may have blocked the differentiation of cells vital for the immune system, pigmentation, and gastrointestinal functions such as B and T lymphocytes, melanocyte stem cells, and gastrointestinal epithelial cells (Henley et al., 2014). The loss of Notch signaling by semagacestat may trigger mutations responsible for the development of skin cancer in patients' groups receiving semagacestat (Nowell and Radtke, 2017). Apart from that, tramiprosate was an anti-glycosaminoglycan compound that targeted the inhibition of Ab aggregation studied until the phase II trial (Folch et al., 2018b). Ab binds with glycosaminoglycan on the cell surface for cellular uptake and internalization depending on the electrostatic interaction between the positively charged Ab and negatively charged sulfate residue on the glycoaminoglycan (Stopschinski et al., 2018). Tramiprosate was also withdrawn from the trial due to a lack of consistent cognitive improvement as measured in Alzheimer's Disease Assessment Scale-cognitive subscale (ADAS-Cog) (Aisen et al., 2011). Nevertheless, tramiprosate was found to have higher efficacy to apolipoprotein E4 homozygotes in AD patients. Therefore, a thorough molecular analysis elucidated the mechanism of action of tramiprosate in aggregate Ab 42 that may be advantageous for further development of tramiprosate as an AD therapeutic regimen (Sabbagh, 2017). Concisely, tramiprosate prevents the misfolding of self-assembled Ab 42 monomers by enveloping them, thereby hindering the aggregation of neurotoxic Ab 42 plaques (Kocis et al., 2017).
Besides, extensive development of Ab humanized IgGI monoclonal antibodies agents, such as solanezumab and bapineuzumab, that targeted the central epitopes of soluble Ab monomers and the N-terminus of Ab 42 were also halted due to a lack of efficacy in mild-to-moderate AD patients Castellani et al., 2019). After repetitive failures in phase III clinical trials in mild-to-moderate AD patients, EXPEDITION 1, EXPEDITION 2 and EXPEDITION 3, solanezumab are currently being tested in asymptomatic and mildly symptomatic patients with biomarker evidence of Ab plaques deposition in brains as a preventive strategy towards AD (Willis et al., 2018). A combined therapy of solanezumab and gantenerumab was also terminated due to their lack of clinical advantages and apparent side effects when combined with BACE1 inhibitor (Cummings et al., 2019). The combination therapy was initiated to enhance the immune response towards Ab plaques, hence, promoting Ab clearance while BACE1 inhibits the generation of new Ab (Folch et al., 2018b). Nonetheless, there are two emerging anti-Ab oligomers monoclonal antibodies with promising efficacy, aducanumab and BAN2401, that bind to insoluble fibrils and soluble Ab protofibrils, thus, relieving the brains' Ab burden with positive impact on cognition (Panza et al., 2019). Aducanumab is an Abtargeting monoclonal antibody that is currently showing significant dose-dependent reduction of Ab plaques' size (Haeberlein et al., 2018). Besides binding to both forms of Ab, soluble oligomers and insoluble fibrils, aducanumab also alleviates calcium dyshomeostasis in affected neurons, since neuronal calcium was found altered in AD brains (Gamage and Kumar, 2017). Further studies should investigate whether the alleviation of elevated intracellular calcium by aducanumab plays a role in restoring the cognitive functions in AD. In a phase III trial, aducanumab failed to slow down cognitive deterioration in mild-to-moderate patients due to reasons such as the patients were irreversibly symptomatic, targeting Ab was not sufficient as it may have already caused irreversible synapse and microglia toxicity, and multifactorial AD may require combination therapy (Selkoe, 2019).
BAN2401 is a humanized monoclonal antibody with encouraging therapeutic effects in treating AD. The drug is highly selective to Ab protofibrils and recedes the formation of Ab plaques, causing a 30% delay to cognitive impairment in mild-to-moderate AD patients within 18 months and a 47% delay by the highest dose in a phase II trial (Swanson et al., 2018). However, future studies should explore the potential of this drug in a larger group (Mendes and Palmer, 2018;Panza et al., 2019).
Passive immunization is the most predominant therapeutic approach in targeting Ab where exogenous monoclonal antibodies (mAbs) are administered to the patients. However, this approach has been weighted with repetitive failures that subsequently theorized several augmentations for the anti-Ab mAbs development (Piton et al., 2018). Firstly, mAbs targeting the N-terminus of Ab are highly potent in suppressing the aggregation of Ab and disaggregating the pre-existing Ab fibrils (van Dyck, 2018). Next, better penetration through the BBB and higher doses of mAbs may be tested in the future since mAbs have astonishing safety profiles. Finally, recruiting preclinical AD patients for the prevention of AD clinical trials is one of the initiatives taken to maximize the benefits of anti-Ab mAbs in treating AD at an early stage (Hampel et al., 2010).

EMERGING TREATMENTS: NEUROINFLAMMATION
Recently, the drive for new therapeutic strategies has focused on neuroinflammation interceded by microglia and astrocytes in AD pathogenesis, rather than the accustomed AD hypotheses such as Ab and tau pathologies. This has resulted in extensive investigations of anti-inflammatory and antioxidant agents (Heneka et al., 2015). In a healthy brain, microglia provide protection against exogenous insult, while astrocytes furnish nutritional and structural support for neurons (Van Eldik et al., 2016). At the early stages of AD, excessive deposition of extracellular Ab plaques and continuous activation of glial cells cause the release of inflammatory cytokines such as interleukin-1a (IL-1a), tumor necrosis factor-a (TNF-a), and complement protein (C1q) (Figure 1). The cytokines also increase astrocytes' expression of insulin-degrading enzyme (IDE) for Ab degradation and signal for neuronal apoptosis (Son et al., 2016). NFkB signaling cascades enhanced in glial cells of AD brain to produce various inflammatory and immune proteins, including compliment component 3 (C3) that initiates neural destruction through complement-mediated synapse pruning when it binds to C3a receptors on a neuron (Benarroch, 2018). Constant production of inflammatory cytokines by microglia leads to neuroinflammation and synaptic loss. Neuroinflammation suppresses the phagocytosis of Ab plaques which may aggravate neurodegeneration (Cianciulli et al., 2020). Despite the undefined mechanism of rilapladib, it is presumed that rilapladib reduces neuroinflammation through the reduction of proinflammatory cytokines and restoration of BBB integrity (Maher-Edwards et al., 2015;Huang et al., 2020).
Microglia are activated into disease-associated microglia (DAM) through triggering receptors expressed on myeloid cells 2 (TREM2) independent and dependent pathways during the microglia-Ab plaques interaction, which facilitate Ab plaques phagocytosis and suppresses overproduction of inflammatory cytokines (Keren-Shaul et al., 2017). Activation of TREM2 leads to the phosphorylation of immunoreceptor tyrosine-based activation motif (ITAM) that causes the spleen tyrosine kinase (SYK) to dock within the receptor complex and activates the phosphatidylinositol 3-kinase (PI3K) cascades (Zheng et al., 2018). The initial response following the activation of PI3K pathway was to produce pro-inflammatory cytokines as a neuroprotective feedback (Cianciulli et al., 2016). Several studies reported that mutations of TREM2 in AD reduced the phagocytic capacity of DAM, disrupted the downstream PI3K pathway, and impaired suppression of pro-inflammatory cytokines released by the glial cells (Jay et al., 2017;Achebe et al., 2018). Meanwhile, overexpression of TREM2 ameliorated neuroinflammation by inhibiting the pro-inflammatory responses initiated by microglia (Ren et al., 2018).
In addition, restoring the microglia function in Ab clearance may open new doors for the development of AD treatments through inhibiting receptor-interacting serine/threonine protein kinase 1 (RIPK1) that are highly expressed in microglia of AD brains (Mullard, 2018). Inhibition of RIPK1 reduces the overexpression of Cst7 that encodes for cystatin F, an endosomal/lysosomal cathepsin inhibitor, which reduces the phagocytic capacity of primary immune cells (Ofengeim et al., 2017). Therefore, RIPK1 inhibitor is suggested to encounter neuroinflammation caused by inflamed microglia that disrupts the phagocytosis of toxic dead cells by reducing the expression of Cst 7 and escalating Ab plaques clearance.
Besides, phenolic compounds with antioxidant properties such as oleuropein and epigallocatechin gallate (EGCG) are also of interest for early intervention in AD management (Kamil et al., 2018). This is mainly because a significant deficiency in brain antioxidant levels was identified as the oxidative stress marker in AD (Sharman et al., 2019). For instance, EGCG was said to suppress the Ab plaque-induced upregulation of proinflammatory cytokines in microglia, and also upregulate the expression of endogenous antioxidants such as nuclear arythroid-2 related factor 2 (Nrf2) and heme oxygenase-1 (HO-1). Hence, EGCG elicits a protective effect against oxidative stress and neuroinflammation (Cheng-Chung Wei et al., 2016).

NOVEL PHARMACOLOGY: MOLECULAR TARGETS
Development of novel AD pharmacotherapy is becoming profoundly important as the complexity of AD pathogenesis becomes better understood in recent times, resulting in the exploration of multitude targets in AD therapeutic strategies ( Table 3). The pipeline of AD treatments is augmented with compounds that either modify the underlying AD pathophysiology, target several molecular targets synergistically, or repurposed as an anti-Alzheimer's drug (Bachurin etal., 2017).
Other potential treatments for AD include those based on the inhibition of glycogen synthase kinase-3b (GSK-3b) and mammalian target of rapamycin (mTOR) to attenuate neuroinflammation by increasing Ab clearance and decreasing tau phosphorylation (Modrego and Lobo, 2019). Normally, extracellular ligands bind to a receptor tyrosine kinase (RTKs), which activate PI3K signaling cascades and lead to the activation of PDK-1 and Akt. PDK1 also indirectly activates the mTORC2 complex which activates Akt through the phosphorylation of the kinase at S473 and S450. Active Akt phosphorylates and activates tuberous sclerosis complex (TSC) 1 and 2, a negative regulator of mTORC1 (Hermida et al., 2017). mTOR inhibition activates the ubiquitin proteasome system and autophagy. In AD brain, where PI3K/Akt signaling pathway is downregulated due to Ab plaques-induced neurotoxicity, Akt is suppressed and mTORC1 activity is increased, which disrupts cell autophagy, leading to neuroinflammation (Shal et al., 2018). Wang and colleagues reported rapamycin's (mTORC1 inhibitor) ability to inhibit elevated activation of mTORC1 and pro-inflammatory cytokines in the hippocampus of AD rats . Suppressed Akt also promotes activation of GSK-3b to cause hyperphosphorylation of tau that aggregates to form NFTs once detached from the microtubules. The loss of microtubules' integrity induces neuroinflammation and increases the risk of neuronal death (Mancinelli et al., 2017).
Methylene blue (MB) is one of the anti-tau disease modifying agents (DMTs) that combats tau pathology through two molecular targets: GSK-3b and tau aggregation (Gandini et al., 2018). Hyperphosphorylation of tau is associated with the loss of counterbalance between the kinases and phosphatases involved in tau phosphorylation, especially if the phosphorylation takes place at the sites of kinases as this family of enzymes regulate most of the protein function (Pradeepkiran et al., 2019). Kinases involved in tau phosphorylation are mitogen-activated protein kinases (MAPKs), cyclin-dependent kinases (Cdks), GSK-3b, and protein kinase A (PKA) (Kheiri et al., 2018;Li et al., 2018).
Initially, MB was well-known for its inhibiting activity on tau aggregation in AD clinical trials (Gureev et al., 2019). In spite of the advance in knowledge on tau pathology, partial inhibition on tau aggregation by MB is not adequate to halt AD as the underlying event that causes tau-mediated neurotoxicity is the binding of granular tau oligomers during NFT formation (Soeda et al., 2019).
Binding of Ab plaques or glutamate to synaptic receptors can initiate the production of cyclic adenosine monophosphate (cAMP) from adenylyl cyclase, activating the cAMP/PKA pathway. Under normal conditions, downstream cAMP/PKA pathway results in phosphorylation of transcription factors such as cAMP response element binding protein (CREB) at S133, which stimulates transcription and translation of genes involved in the formation of memory (Bartolotti et al., 2016;Gao et al., 2018). However, the level of p-PKA was significantly decreased in the hippocampus of AD mice while neuroinflammation was   Retinoid X receptors (RXR) agonist to reduce Ab in the brain Phase II clinical trial on mild-to-moderate AD patients Dosage: 75 mg of bexarotene b.i.d., 150 mg after 1 week Primary endpoint: 1. Change in brain amyloid burden measured by standard uptake unit regional (SUVr) according to genotypes involved in this trial Secondary endpoint: 1. MMSE 2. ADAS-cog 3. CDR 4. NPI 5. ADCS-ADL 6. Serum Ab 1-42 level 7. Ratio of Ab 42 to Ab 40 in non-ApoE4 carriers (Cummings et al., 2016) ELND005 (formerly known as AZD-103), scyllo-inositol NCT01735630  found to be increased (Cai et al., 2018). Decreased PKA subsequently decreased the phosphorylated CREB in the rats' hippocampus . High levels of BACE1 in AD brain inhibits adenylyl cyclase and impairs cAMP/PKA pathway, which interrupts the phosphorylation and eventually disrupts the transcription and translation of CREB-induced genes, leading to memory impairment in AD (Chen et al., 2012). CREB activation can restore memory impairment in AD as the CREB-induced genes, such as brain-derived neurotrophic factor (BDNF) and insulin-like growth factor 1, can enhance neuron morphological outgrowth and formation of long-term and short-term memories (Kubota et al., 2017).
Metformin is a first-line medication for type 2 diabetes, which was repurposed for AD treatment as it exhibits anti-inflammatory properties and neuroprotective features against cognitive deterioration in AD (Ou et al., 2018). Metformin hinders the neuronal apoptosis and promotes neurogenesis in the hippocampus through the activation of the AMP-activated protein kinase (AMPK) pathways, which leads to the improvement of memory formation. AMPK can be phosphorylated by 3 key kinases, such as the liver kinase B1 (LKB1) complex at T172, due to increased cytoplasmic level of AMP, increased cytosolic Ca 2+ , and mitogen-activated protein kinase 7 (MAP3K7), also known as transforming growth factor betaactivated kinase 1 (TAK1) . AMPK also activates TSC1/2 complex that inhibits mTOR . Activity of AMPK was decreased in the hippocampus of AD rats at age 4-5 months, while the activity of mTOR increased, causing disrupted cell autophagy and exacerbated AD (Du et al., 2015;Sun et al., 2019). Inhibition of mTOR by rapamycin restores normal cell autophagy and protein synthesis .
PKA was found to activate silent information regulator type 1 (SIRT1), a neuroprotective protein deacetylase that reduces tau acetylation and downregulates BACE1, hence, increasing degradation of tau and reducing Ab production (Zhang et al., 2019;. Resveratrol, through activation of SIRT1, reverses the drastic decrease in hippocampal expression of SIRT1 in AD (Turner et al., 2015;Hou et al., 2017). Resveratrol was also found to reduce neuronal apoptosis and eventually restore cognitive impairment in AD (Tian et al., 2016).
Current DMTs target either Ab pathology or tau pathology, which may be the reason for their lack of efficacy since both pathologies synergistically cause cognitive degeneration with the cholinergic deficit as a constant concern in AD. Multi targetdirected ligand (MTDLs) is a novel approach to surmount the multifactorial AD pathogenesis (Geldenhuys and Darvesh, 2015). RPEL was synthesized through the combination of N, N'-disubstituted piperazine anti-amyloid scaffold and tacrine, into one compound (Sergeant et al., 2019). It was found effective in preventing cognitive impairment as it minimizes the Ab plaques formation and tau phosphorylation in AD mice without any adverse effect besides maintaining the inhibitory activity on AChE. This approach accelerates the development of potential treatment for AD by minimizing the cost and time since the individual therapeutic effects of each compound is generally known (Hassan et al., 2019).

CONCLUSION
Despite decades of research, we are still encountering a lack of success in pharmacotherapy of AD, mostly due to the multifactorial etiologies of the disorder that can initiate neurodegeneration interdependently. At present, combination therapy targeting several factors simultaneously appears to be promising. Additionally, an increasing number of studies are also focusing on neuroprotection against neuroinflammation. The impact of neuroinflammation interceded by microglia and astrocytes in AD pathogenesis is of great interest as it opens new doors for novel therapeutic targets. In addition to pharmacotherapy, better prognosis through early detection of AD biomarkers or brain imaging will enable early intervention that could potentially prevent the deposition of Ab plaques and manifestations of various irreversible symptoms of AD.