st,

diah, Ch

Alzheimers diseaseAmyloidDementiaNeuropharmacologyNeurobrillary tangles

s thy ofthe

its management. These drugs provide symptomatic improvement alone but do less to modify the disease

mosting evea slow

age; others being a positive family history, head trauma, femalegender, previous depression, diabetes mellitus, hyperlipidemia andvascular factors (Kivipelto et al., 2001). The understanding of thepathophysiology of AD is constantly changing; for instance thetangles, a well known pathological hallmark of AD, earlier thought

mmation, mitoticomponents play anisms are diverse,lting in AD.

2012), four cholinesterase inhibitors and memantine are the onlydrugs approved by the US Food and Drug Administration for itstreatment. These drugs provide symptomatic treatment but do notalter the course of the disease. Hence the modern therapeutic op-tions that target the disease modication part are on a rise. Themultiple mechanisms involved in the pathogenesis of AD createconsiderable difculty in producing an effective treatment (Fig. 1).This current reviewattempts to summarize the existing therapeutic

* Corresponding author. Tel.: 91 416 2284267.

Contents lists availab

rm

ev

Neuropharmacology 76 (2014) 27e50E-mail addresses: grifndoc@gmail.com, anandr@cmcvellore.ac.in (R. Anand).components playing a role. The greatest risk factor is advancing Although AD is known for about a century (Ramirez-Bermudez,imposes tremendous emotional and nancial burden to the pa-tients family and community through the provision of care and lossof wages. The disease maybe classied based on the age of onsetinto early-onset AD and late-onset AD. Early onset AD accounts forapproximately 1%e6% of all cases and manifests roughly between30 and 60 years. Late onset form accounting for around 90% of caseshas an age at onset later than 60 years. Etiology of AD is multifac-torial with genetic, environmental, behavioral and developmental

tangles accumulation, hormone imbalance, inadysfunction, calcium mishandling, and genetic crole in the disease process. Although the mechaneuronal death, the inevitable event occurs resu

2. Therapeutics in AD2005). One person for every 85 individuals can be expected tosuffer from AD by the year 2050 (Brookmeyer et al., 2007). AD also

them, various deranged mechanisms such as chronic oxidativestress, mitochondrial dysfunction, Ab production, neurobrillary1. Introduction

Alzheimers disease (AD) is thementia, and with a new case occurrally, the disease itself is becoming0028-3908/$ e see front matter 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.neuropharm.2013.07.004process. The extensive insight into the molecular and cellular pathomechanism in AD over the past fewdecades has provided us signicant progress in the understanding of the disease. A number of novelstrategies that seek to modify the disease process have been developed. The major developments in thisdirection are the amyloid and tau based therapeutics, which could hold the key to treatment of AD in thenear future. Several putative drugs have been thoroughly investigated in preclinical studies, but many ofthem have failed to produce results in the clinical scenario; therefore it is only prudent that lessons belearnt from the past mistakes. The current rationales and targets evaluated for therapeutic benet in ADare reviewed in this article.

This article is part of the Special Issue entitled The Synaptic Basis of Neurodegenerative Disorders. 2013 Elsevier Ltd. All rights reserved.

common cause of de-ry seven seconds glob-pandemic (Ferri et al.,

to be responsible for the disease now rather seem to reect thedamage which the neurons have endured over a long time. Thenotion that amyloid beta peptide (Ab) and phosphorylated tau arepathologic molecules is slowly changing, and it seems that theyrepresent a cellular adaptive strategy to oxidative stress. Apart fromKeywords:26 June 2013Accepted 2 July 2013

emotional and nancial burden to the patients family and community. The disease has been studied overa century, but acetylcholinesterase inhibitors and memantine are the only drugs currently approved forReview

Therapeutics of Alzheimers disease: Pa

R. Anand a,*, Kiran Dip Gill b, Abbas Ali Mahdi c

aDepartment of Biochemistry, Christian Medical College, Vellore 632002, Tamilnadu, InbDepartment of Biochemistry, Postgraduate Institute of Medical Education and ResearccDepartment of Biochemistry, King Georges Medical University, Lucknow, UP, India

a r t i c l e i n f o

Article history:Received 29 December 2012Received in revised form

a b s t r a c t

Alzheimers disease (AD) itorial, and pathophysiologcases of AD, emphasizing

Neuropha

journal homepage: www.elsAll rights reserved.andigarh, India

e most common cause of dementia worldwide. The etiology is multifac-the disease is complex. Data indicate an exponential rise in the number ofneed for developing an effective treatment. AD also imposes tremendouspresent and future

le at ScienceDirect

acology

ier .com/locate/neuropharm

armaR. Anand et al. / Neuroph28strategies till date with correlation to the pathophysiologic mech-anisms for AD (Table 1).

2.1. Modulating neurotransmission

The cholinergic group of neurons is the main neurotransmittersystem involved in AD and basal forebrain cholinergic loss is a wellrecognized pathology. These neurons maintain cortical activity,cerebral blood ow, modulate cognition, learning, task and mem-ory related activities, development of the cerebral cortex andregulation of sleepewake cycle (Berger-Sweeney, 2003; Schliebsand Arendt, 2006). Considering the many functions of the cholin-ergic neurons, the symptom complex in AD can at least be partiallyunderstood.

The dysfunction of the cholinergic system in AD occurs atvarious levels including a decreased choline acetyltransferase ac-tivity, reduced choline uptake, a decrease in acetylcholine synthesis(Slotkin et al., 1990) and altered levels of acetylcholine receptors(AChRs) (Xu et al., 2012). Glutamate is the primary excitatoryneurotransmitter in the hippocampal and neocortical regions of thebrain, and they do play a signicant role in cognition, learning and

Fig. 1. Pathogenesis of Alzheimers disease eWeb of causation. The gure shows the variouare in bold boxes. *Indicates the mechanisms against which therapeutic strategies are targcology 76 (2014) 27e50memory process. The post-synaptic membrane has high density ofone of its receptors- the N-methyl-D-aspartate (NMDA) receptor.Studies have shown there is an extracellular glutamate excess inAD, contributed both by an increased presynaptic glutamate releaseand decreased re-uptake which, in turn lead to a tonic activation ofNMDA receptors (Revett et al., 2013). The excitotoxicity is slow incontrast to the acute or rapid form that occurs with stroke or epi-lepsy. Impaired insulin signaling along with mitochondrialdysfunction and receptor abnormalities (Beal, 1992) can predisposeto this process when glutamate is excitotoxic even at physiologicalconcentrations (Novelli et al., 1988). Dysfunction in other neuro-transmitter systems such as g-aminobutyric acid (GABA), hista-mine, and serotonin systems of neurons also lead to AD. Themodulation of neurotransmission with drugs continue to remainthe best approach to providing symptomatic improvement in pa-tients; of late, mechanistic insights into their disease modifyingaspects have also been highlighted.

2.1.1. Cholinesterase systemThe four acetylcholinesterase inhibitors (AChEI) approved by

the U.S. Food and Drug Administration for the treatment of AD are

s mechanisms involved in pathogenesis of Alzheimers disease. The major mechanismseted.

armaTable 1Therapeutic strategies in Alzheimers disease.

1. Modulating neurotransmission

Cholinesterase inhibitors N-methyl D-aspartate receptor antagonism GABAergic modulation Serotonin receptor modulation Histaminergic modulation Adenosine receptor modulation

2. Tau based therapies

Tau phosphorylation inhibition Microtubule stabilization Blocking Tau oligomerization Enhancing Tau degradation Tau based immunotherapy

3. Amyloid based strategies

Secretase enzymes modulation Amyloid transport Preventing amyloid aggregation Promoting amyloid clearance Amyloid based immunotherapy

4. Modulating intracellular signaling cascades5. Oxidative stress reduction

Exogenous antioxidant supplementation Augmenting endogenous defense

6. Mitochondrial targeted therapy7. Modulation of cellular calcium homeostasis8. Anti-inammatory therapy9. Others

Gonadotropin supplementation

R. Anand et al. / Neurophtacrine, donepezil, rivastigmine, and galantamine, however, tacrineis now rarely used because of its hepatotoxicity (Watkins et al.,1994). AChEI enhance cholinergic neurotransmission through in-hibition of acetylcholinesterase (AChE), thus decreasing thebreakdown of acetylcholine. It is clear that, when cholinergictransmission occurs, or their receptors become activated there is anincrease in the long term potentiation. The cholinergic AChRs areexpressed on principal and inhibitory interneurons, both pre- aswell as post-synaptically in most regions of hippocampus; thus,boosting acetylcholine levels in the synaptic cleft can have bidi-rectional inuences (Drever et al., 2011). Galantamine also pos-sesses agonist activity at the nicotinic a4b2 receptor subtype and itsclinical benets are probably due to both the mechanisms (Coyleand Kershaw, 2001). More recently it is known that the musca-rinic M1 AChRs are present in intracellular locations, especially inthe hippocampal regions (Anisuzzaman et al., 2013). The cell sur-face M1 AChRs activate the phosphatidylinositol cascade, whereasintracellular M1 AChRs activate the extracellular regulated kinases1/2 (Anisuzzaman et al., 2013); both the pathways regulate longterm potentiation and synaptic plasticity. Cholinergic transmissionalso plays a role in modulating the mechanisms involved in adultneurogenesis (Bruel-Jungerman et al., 2011) and studies do suggestthat AChEIs alleviates oxidative stress in animal studies andhumans (Klugman et al., 2012). Various short term trials with AChEImonotherapy have shown clinically apparent and encouragingimprovement in cognitive function, slowed the pace of functionaldecline or clinical worsening compared with placebo and reducedbehavioral symptoms in mild-to-moderate and moderate-to-

Lipid modiers e Statins Growth factor supplementation Metal chelation Epigenetic modiers Caspase inhibitors Nitric oxide synthase modulation Nucleic acid drugs Multi-target directed ligandssevere AD patients; data from meta-analyses also attest to thesame fact (Birks, 2006). Efforts to deliver standard drugs efcientlyare another area of development; trans-dermal delivery systems forall the three drugs are available but not currently approved. Theoral dosing of AChEIs increases the plasma drug levels in a shorttime interval, which probably accounts for the observed gastroin-testinal side effects. The trans-dermal patches can ensure peak lessand prolonged delivery with minimal uctuations in the plasmadrug concentrations (Lefevre et al., 2008). Rivastigmine patches(9.5 mg/24 h) produced plasma drug concentrations and resultsthat were comparable with oral capsules (12 mg/day); safety andtolerability prole of the patches were better. The subjects alsoexperienced signicantly decreased discomfort with patches(Winblad et al., 2007).

Some novel AChEI molecules have also been developed. Mem-ogain (GLN-1062; Galantos Pharma) the benzoyl ester of galant-amine is a pro-drug that is available as intranasal formulation. Thedrug has shown excellent efcacy and central nervous system(CNS) bioavailability in animal studies (Maelicke et al., 2010).Huperizine A is a natural alkaloid isolated from the Chinese mossshrub (Huperzia serrata) and possesses AChE inhibiting actionwithmodest effects on amyloid precursor protein (APP) metabolism andneuroprotection (Zhang et al., 2008). The drug showed promisingsafety prole in both phase I and phase II trials. At a dose of 400 mgtwice daily, the drugwas able to improve cognitive outcome by 2.27points in patients with mild to moderate AD (Rai et al., 2011). Apro-drug of huperizine, ZT-1 has shown admirable pharmacoki-netic prole in a recent phase I study (Jia et al., 2013). Meth-anesulfonyl uoride (SNX-001) is an irreversible inhibitor of AChErst reported in 1999 for its therapeutic value in AD (Moss et al.,1999). Preclinical studies demonstrating its benet in cognitionhave revived interest in the molecule recently. A phase I trial hasstudied the extent of AChE inhibition in healthy subjects and hasfound promising results (Moss et al., 2013). The dual acting AChEIcompounds are to be discussed in a different section.

Direct modulation of the cholinergic AChRs is also underconsiderable evaluation. In AD, the levels of presynaptic M2 AChRsdecrease, but those of postsynaptic M1 AChRs remain unchanged. Anumber of M1 partial agonists like AF102B, AF150(S), AF267B andAF292 and allosteric agonists such as 77-LH-28-1, LY-593093 andLu AE51090 are available; ML 169 is a recently reportedM1 positiveallosteric modulator (Reid et al., 2011). The strong side of the M1agonists seems to be their role in APP processing and thus indi-rectly on other processes such as tau phosphorylation; studies haveshown ablation of M1 AChRs lead to increased amyloid b (Ab)generation (Medeiros et al., 2011). AF102B, an M1 partial agonist,signicantly lowered CSF Ab levels in AD patients (Nitsch et al.,2000), where AChEIs showed no effect (Parnetti et al., 2002).AF150(S) and AF267B have also shown promising results in thepreclinical setup (Fisher, 2008; Fisher et al., 2002). Another mixedmuscarinic/s1 agonist- ANAVEX 2-73 is currently in phase I/IIatrials. This compound has partial agonistic activity at both musca-rinic AChR and s1 protein (chaperone protein in endoplasmic re-ticulum activated by unfolded protein response) (Collina et al.,2013). In animal studies ANAVEX 2-73 attenuates Ab inducedmemory decits and toxicity, decreases seeding of Ab, blocks theactivation of glycogen synthase kinase-3 (GSK3b) and in turn thehyperphosphorylation of tau (Lahmy et al., 2013).

While we slowly begin to understand the therapeutic potentialof muscarinic AChRs, the role of nicotinic AChRs in AD is debatableat best. Different neuronal systems express these receptors, andthey play diverse functional roles in cognition, memory processes,trophism and neuroprotection (Wallace and Bertrand, 2013).Recent evidences have also uncovered their pathological side and

cology 76 (2014) 27e50 29the possible mechanisms by which nicotinic AChRs may contribute

armato the pathophysiology of AD (Hernandez and Dineley, 2012; Parriand Dineley, 2010). Thus, their modulation is governed by a subtlebalance. The levels of nicotinic receptors may remain unchanged oreven upregulated, with progression of the disease process in AD(Ikonomovic et al., 2009).

Evidence show that nicotinic AChR agonists produce bothbenecial and damaging effects on neuronal function; hence, thenet effect is not clear (Fisher, 2012). Studies have shown that a7nicotinic agonists attenuate Ab mediated toxicity (Kihara et al.,2001) but, on the other hand, modify the reactivity and increasephosphorylation state of tau protein (Hellstrom-Lindahl et al.,2000). How modulating the same receptor decreases one pathol-ogy, but increases the other is not clear. More surprisingly, antag-onists of a7 nicotinic AChRs also cause similar effects (Mousavi andHellstrom-Lindahl, 2009). A few recent studies highlight thecognitive enhancing potential of cotinine, a metabolite of nicotine;the compound is a positive allosteric modulator of a7 nicotinicAChRs (Echeverria and Zeitlin, 2012). The properties which makecotinine a better ligand than nicotine would be its low toxicityprole, non addictive nature and good blood brain barrier clearance(Echeverria and Zeitlin, 2012). Apart from its receptor modulationthe drug also inhibits Ab aggregation (Echeverria et al., 2011). Thepharmacokinetic and safety prole of cotinine have already beeninvestigated in humans (Benowitz et al., 1983; Bowman and Mc,1962) but no documentation is available regarding its role in AD.Many of novel ligands for a7 nicotinic AChR are also currently indevelopment (Toyohara and Hashimoto, 2010). EVP-6124 is a novelselective a7 partial agonist that improves memory performance inanimals (Prickaerts et al., 2012); the compound has successfullycompleted phase II trials (NCT01073228). MT-4666 is anothernicotinic agonist that is currently in phase II trials (NCT01764243).Amongst other novel compounds, ABT-384 (NCT01137526) hascompleted phase II trials while a trial with ABT-126 is ongoing(NCT01527916).

To summarize, the advantage of M1 agonists over AChEIs couldbe their potential role as disease modifying agent along with itssymptomatic benets. The role of nicotinic modulators, on theother hand, is not clear. From this group, only AChEIs are currentlyused in the clinical setting; the direct acting ligands describedwould require more convincing evidence. Since neuronaldysfunction starts early in the course of disease, the utility ofAChEIs is to provide symptomatic relief in that transitional periodsustaining the functionwith the available neurons; with increasingneuronal damage, the therapeutic effectiveness of AChEI slowlydiminishes. As to how long the drugs remain valid or how long thepatients should be treated with AChEI, the answers vary; reportsindicate that benet may last up to four years (Rogers et al., 2000).The benet of AChEIs in the behavioral symptoms of AD and theirsynergistic role in combination therapy with memantine is to bedealt in the following section.

2.1.2. N-methyl D-aspartate antagonismNeuronal pathology in AD also extends to the glutamatergic

system but at a later stage of the disease. Glutamatergic neuronsregulate synaptic plasticity, neuronal growth and differentiation,cognition, learning and memory (Buttereld and Pocernich, 2003).A glutamate cycle occurs between the pre- and post-synapticneurons, and astrocytes that determines the synaptic concentra-tion of glutamate available for the receptors (Revett et al., 2013).Cycle defects occur at different levels in AD leading to a state ofextracellular glutamate accumulation, increased NMDA receptoractivation and excitotoxicity (Revett et al., 2013).

A large body of evidence shows mutual interaction betweenNMDA receptors and Ab peptides. Studies suggest that NMDA re-

R. Anand et al. / Neuroph30ceptors activation lead to Ab production and vice versa-Aboligomers binding and activating NMDA receptors, further sub-stantiating the importance of glutamatergic system in AD(Dinamarca et al., 2012; Revett et al., 2013). Memantine is an un-competitive NMDA antagonist, has voltage dependency, rapidblocking kinetics, moderate afnity and blocks the channel bytrapping it in open conformation (Gilling et al., 2009). Mg2 ionsblock the NMDA channel under resting conditions; when glutamateis available the blockade is relieved, and the NMDA channel is nowopen for Ca2 inow. In pathological states such as AD, there is alow and persistent state of NMDA activation even at resting pe-riods; in such states, Mg2 ions are excluded from the channel,thereby, allowing continuous Ca2 ow across the membrane. Themoderate afnity and voltage dependency property of memantineallows it to block the persistent NMDA activation and is thus,benecial in AD. Evidences also indicate memantine mediatedblockade is relieved by high glutamate concentrations in the syn-aptic cleft. Hence, when a physiological impulse arrives the gluta-mate over-rides the memantine blockade, and physiologicaltransmission can continue without interference (Parsons et al.,2013). Experimental evidences show that memantine treatmentimproves spatial learning in animal models of AD, protects neuronsfrom Ab induced toxicity, decreases apoptosis, free radical medi-ated damage and restored synaptic degeneration (Miguel-Hidalgoet al., 2012). Reportedly memantine also seems to have its antag-onizing effect on other receptors such as a7 and a4b2 nicotinicAChRs (Buisson and Bertrand, 1998; Maskell et al., 2003), 5-HT3(Rammes et al., 2001), a3b2 (Lee et al., 2012), 5-HT2A, dopamine D2receptors and histaminergic neurons (Motawaj et al., 2011; Nakayaet al., 2011). Hence, the therapeutic benets of memantine may notbe strictly due to its effect on NMDA alone, but except at a7 nico-tinic AChRs there are no solid evidences to show that the effects onother receptors occur at the therapeutically administered concen-trations in AD (Rammes et al., 2008).

Currently memantine is the only drug approved for clinical usein moderate to severe AD in USA and Europe; studies showconvincing evidence of memantines value (Hellweg et al., 2012;McShane et al., 2006). Although the effect of memantine isevident in late stages, its role in early AD is unclear. The three mainstudies that have seen the role of memantine in mild to moderateAD show there are some benecial effects on cognitive and globalfunctioning status, but it does not impede the progression of dis-ease (Bakchine and Loft, 2008; Peskind et al., 2006; Porsteinssonet al., 2008). A recent metanalysis also indicates the same(Schneider et al., 2011). Memantines lack of benet in the earlystages is not well understood yet. The involvement of cholinergicneurons probably occurs early in the disease but, damage to glu-tamatergic system and excitotoxic degeneration occurs late in thecourse of disease (Ni et al., 2013). The effect of memantine on otherreceptor channels might as well play a role here. Memantine alsoblocks a7 nicotinic AChRs more potently at therapeutic concen-trations (Aracava et al., 2005); this blockade could affect neuro-transmission during the early stages of disease when functioningcholinergic neurons are still available. Hence as of now, use ofmemantine is restricted to the later stages of the disease.

A number of investigators have looked at the possible advantageof its combining memantine with AChEI in AD (Atri et al., 2013;Dantoine et al., 2006; Howard et al., 2012; Lopez et al., 2009;Porsteinsson et al., 2008; Riepe et al., 2007; Tariot et al., 2004).Results of most studies indicate that addition of memantine toAChEI may add to the therapeutic value and improve clinicaloutcome in subjects. However two recent systematic reviews haveconcluded that theremay be few signicant favorable changes fromthe combination therapy, but it is not currently recommended(Farrimond et al., 2012; Muayqil and Camicioli, 2012). A current

cology 76 (2014) 27e50trial of memantine and donepezil combination in moderate to

armasevere stages of AD is ongoing (NCT00866060). A once daily xeddose combination of memantine and donepezil has also beendeveloped (ADS-8704; Adamas pharmaceuticals); the above com-pound is currently in phase III trials.

2.1.3. GABAergic modulationGABA is one of the main inhibitory neurotransmitters. Amongst

the hippocampal neurons, earliest to be affected are the cholinergicgroup followed by the glutamatergic neurons; for unknown rea-sons there is relative sparing of GABAergic neurons (Rissman et al.,2007). Studies also indicate when chronic growth factor depriva-tion occurs, the GABA transmission changes from inhibitory toexcitatory stimulus (Lagostena et al., 2010). However, it is not clearas to which function of GABA is more harming to the cells.GABAergic drugs are currently another class of compoundscurrently tried for their cognitive enhancing potential (Limon et al.,2011). SGS742 is a GABAB antagonist that showed promising resultsin preclinical and phase I studies. The compound reached phase IItrial stage but not beyond. Etazolate, a pharmacological modulatorof GABAA receptor is also neuroprotective (Marcade et al., 2008).Apart from being a GABAA modulator the drug also activates a-secretase (Marcade et al., 2008) and inhibits phosphodiesterase(PDE)-4 activities (Wang et al., 1997). Etazolate was safe and welltolerated in a recent trial, but the effectiveness and long termbenets are to be determined (Vellas et al., 2011). Inverse agonistsfor GABAA receptor with a5 subunit specically also seem toimprove cognition in animal models, but studies in humans haveyielded conicting results. One such compound alpha5IA was ableto restore the alcohol induced impairment in healthy controls, butthere was worsening in learning performance in the elderly sub-jects (Atack, 2010). Hence the role of GABA modulators in AD is notclearly known.

2.1.4. Serotonin receptor modulationThe role of serotonin receptors in AD gained prominence when

observation suggested that levels of receptor and the density of 5-HT positive neurons signicantly decline in AD brains (Reynoldset al., 1995). The areas of the brain concerned with learning andmemory show high concentrations of 5-HT1A, 5-HT4, 5-HT6 and 5-HT7 receptors (Cifariello et al., 2008; King et al., 2008). Majoranatomical distribution of the individual receptor subtypes differ;the hippocampal formation, entorhinal cortex and raphe nucleiexpress 5-HT1A receptors in (Chalmers and Watson, 1991) at bothpre- and post-synaptic locations (Rodriguez et al., 2012). The post-synaptic 5-HT1A stimulation inhibits cholinergic transmission, but,on the other hand, the pre-synaptic 5-HT1A autoreceptors exert anegative feedback on the serotonergic transmission. The basalganglia and hippocampus also have a high density of 5-HT4 (Vilaroet al., 2005) and 5-HT6 receptors (Marazziti et al., 2012). Theinteraction of the serotonin system in the nervous system getsfurther complicated because the 5-HT receptors co-localize onglutamatergic, cholinergic and GABAergic neurons, thus, indicatingserotonin system is capable of regulating of a variety of otherneurotransmitter systems (King et al., 2008).

A number of serotonomimetic compounds (monoamine oxidaseinhibitors and selective serotonin reuptake inhibitors) already inclinical use are under consideration in AD as monotherapy or alongwith AChEI for their cognitive enhancing capacities (Rodriguezet al., 2012). Many novel ligands with agonistic or antagonisticproperties targeting different 5-HT receptors are available (5-HT1A,5-HT4 and 5-HT6) (King et al., 2008). A 5-HT1A antagonist lecozotanproved to be safe and effective in phase I (Patat et al., 2009) andphase II trials (NCT00151333), but the drug has not been studiedbeyond that. A number of 5-HT4 agonistic compounds like PRX-

R. Anand et al. / Neuroph03140, velusetrag, TD-8954, RQ-00000009, SUVN-D1003019 andSUVN-1004028 have shown cognitive benets in preclinical studieswith potential effects on amyloid processing (Shen et al., 2011). Thesafety and pharmacokinetic properties of velusetrag are known,and the drug is under use currently for gastrointestinal disorders(Long et al., 2012). SB-742457, a novel 5-HT6 agonist has also shownpositive results in phase II trials as a monotherapy and combinationwith donepezil (Maher-Edwards et al., 2011, 2010). Interestingly,few 5-HT6 antagonists like Ro-4368554, SB-258585 and SB-399885have also shown cognition enhancing properties in animal exper-iments (Gravius et al., 2011; Hirst et al., 2006); hence, the clinicalutility of 5-HT6 ligands is yet to be understood. Serotonergicmodulation also appears to relieve some behavioral manifestationsof AD. Inverse agonists of 5-HT2A receptors maybe of some help inimproving cognitive and non-cognitive processes. Pimavanserin isa novel inverse agonist that successfully reversed psychosis likefeatures in animal models (Price et al., 2012) and in patients withParkinsons disease (Meltzer et al., 2010). 5-HT7 is a recently char-acterized receptor protein and its role in learning, and memoryprocesses are under consideration (Cifariello et al., 2008).

2.1.5. Histaminergic modulationHistamine receptor H3 expression is high in several brain re-

gions, including those involved in sleepwake regulation andcognitive functions (Motawaj et al., 2010). Activating the receptorinhibits histamine release in the brain but its selective antagonismenhances the release of various neurotransmitters includingacetylcholine, GABA, dopamine and nor adrenaline (Chazot, 2010).Intriguingly, the expression of H3 receptor is unaltered despiteprogress in AD suggesting its modulation could be of therapeuticbenet (Medhurst et al., 2009). Preclinical studies show cognitionenhancing properties for novel H3 antagonists, including BF2.649,PF-03654746, GSK189254, MK-0249, JNJ-17216498, and ABT-288(Brioni et al., 2011). However, results of the phase II trial of MK-0249 show no benet in subjects with mild to moderate AD(Egan et al., 2012). PF-03654746 has completed a phase I trial, butthe results are not available (NCT01028911). ABT-288 was safe andwell tolerated in healthy adults (Othman et al., 2013); the com-pound recently completed its phase II trial (NCT01018875). A smallpilot trial with selective H3 antagonist GSK239512 also showedexcellent safety prole with positive effects on attention andmemory (Nathan et al., 2013). The therapeutic role of these com-pounds is not clear yet.

2.1.6. Adenosine receptor modulationThe neuromodulation role of adenosine has recently come into

light, and its role in neurodegenerative disorders is under investi-gation. Adenosine receptors, especially adenosine2A play pivotalroles in modulation of neuronal function, linking the system to ADrelated cognitive decits (Canas et al., 2009). In vivo studies haveshown neuroprotective value for SCH58261, an adenosine2Ablocker (DallIgna et al., 2007). Investigation of the utility of cil-ostazol, an antiplatelet drug and PDE-3 inhibitor for its benet indementia is ongoing (NCT01409564).

2.1.7. Tackling the neuropsychiatric aspects of ADThe other key aspect of the clinical syndrome of AD is the non

cognitive manifestations for which no approved treatments ormanagement guidelines exist. Some treatment related factors thatneed attention in this regard include the age of the subject, theirability to tolerate psychotropic drugs and its side effects and thepossible drug interactions. The common neuropsychiatric symp-toms in patients with AD are lack of cooperation/concentration,tremors, irritability, apathy, depression, negativism, sleep distur-bances, disinhibition, delusions or hallucinations (Mohs, 2005).

cology 76 (2014) 27e50 31Selective serotonin reuptake inhibitors are commonly used as

armaadjunct drugs to treat the psychotic, apathy and depressive featuresassociated with AD. However a recent report has concluded thatsertraline and mirtazapine are not benecial in AD (Banerjee et al.,2013). Mood stabilizers also do not seem to offer any treatmentrelated benets in AD (Xiao et al., 2010). The overlap between theneural circuitry involved in dementia and neuropsychiatric aspectsallows the drugs modulating one neurotransmission to exertbenet in different domains. There is substantial evidence that thecommonly used AChEI drugs-donepezil, rivastigmine and galant-amine improve psychiatric and behavioral manifestation alongwith cognitive enhancement (Pinto et al., 2011). AChEIs as mono-therapy as well as in combination with antidepressant or antipsy-chotics provide a synergistic effect in AD.

Studies also demonstrate the potential value of memantine inimproving the behavioral manifestations. At the dose commonlyused (20 mg/kg body weight), memantine improves overall func-tioning status, improves cognitive abilities and activities of dailyliving, slows down functional decline and decreases worsening ofmood and behavior (Hellweg et al., 2012; McShane et al., 2006). Apost marketing surveillance study has observed that memantinemonotherapy for minimum 6 months improved behavioraloutcome in the subjects (Clerici et al., 2012). Memantine also hadan added benet in subjects with agitation, aggression and psy-chosis (Wilcock et al., 2008). Other drugs that have shown somebenet include methylphenidate and melatonin. Methylphenidatehas shown improvement of apathy and depression in AD; however,the ndings need to be conrmed from large studies (Padala et al.,2010). Melatonin also has shown its effectiveness in decreasing thesundowning symptoms and sleep disturbances in AD (Cardinaliet al., 2010). Currently there is a lack of controlled studies thatdemonstrate the value of drugs regarding their psychotropic effectsin AD; the future merits appropriately designed studies to deter-mine the same.

2.2. Tau-based therapies

Neuronal cells commonly express tau protein,where, its purposeis to stabilize themicrotubules. By regulatingmicrotubule assembly,tau modulates the functional organization of the neurons, particu-larly axonal morphology, growth, and polarity (Buee et al., 2000).The protein has several phosphorylation sites, and the microtubulebinding property of tau is dependent on the phosphorylation state.The phosphorylated tau binds microtubules with a lesser afnityleading to microtubule instability. In AD, hyperphosphorylated tauaccumulates and aggregates into paired helical laments, loses itsmicrotubule binding and stabilizing role, contributing to neuronaldegeneration (Garcia and Cleveland, 2001).

2.2.1. Targeting tau phosphorylationTau phosphorylation is a key event in AD contributing to

microtubule instability; hence, inhibition of kinases to prevent theprocess is a valid rationale. The main focus has been on glycogensynthase kinase 3 (GSK3), one of the primary enzymes involved intau phosphorylation. Lithium and valproate have inhibitory actionson GSK3 and when administered they reduce tau pathology intransgenic mice (Engel et al., 2006; Noble et al., 2005). Small scaletrials have shown some benecial effects, but larger, morecontrolled studies have failed to prove the benet for both lithium(Hampel et al., 2009) and valproate (Tariot et al., 2009). In additionto an inhibitory effect of caffeine on PDE enzyme, recent studiessuggest it also inhibits GSK3b; caffeine administration in Ab trans-genic animalmodels has showndecreasedAbproduction (Arendashet al., 2009). Various epidemiologic studies also report a decreasedincidence of AD in heavy caffeine users (Eskelinen et al., 2009).

R. Anand et al. / Neuroph32Tideglusib (NP031112) is an irreversible inhibitor of GSK3b currentlystudied for its value in AD and progressive supranuclear palsy. Apilot study inmild tomoderate ADwith different escalating doses ofthe drug showed an increasing trend in mini mental state exami-nation (MMSE) and cognitive scales; the drug has recentlycompleted phase IIb trials (NCT01350362). Several small moleculeinhibitors of GSK3 like SB216763, CHIR-98014 (Selenica et al., 2007)and SRN-003-556 (Hampel et al., 2009) are currently in pre-clinicalstudies. SB216763 was able to decrease the amount of phosphory-lated tau, but serious adverse effects occurred in control animals (Huet al., 2009), thus raising caution regarding its use. This result alsoemphasizes the necessity of a compound that does not affect thebasal activity of GSK3. Bezabrate, a pan-peroxisome proliferatoractivated receptor agonist has shown its effectiveness in decreasingtau phosphorylation improving behavioral features in P301S micebut the clinical implications are not clear (Dumont et al., 2012).Evidences also suggest a protective role for insulin in AD; preclinicalstudies show that when administered via intranasal route insulindecreases GSK3 b activation and in turn reduces tau phosphoryla-tion (Yang et al., 2013). Clinical evidences demonstrate the potentialof intranasal insulin therapy in improving cognitive function(Shemesh et al., 2012) but this effect seems to be stronglydependenton the ApoE 34 carrier status and sex of the subject (Claxton et al.,2013; Reger et al., 2008). Numerous clinical trials evaluating thevalue of insulin in AD are currently ongoing (NCT01767909,NCT01636596, NCT01436045, NCT01595646).

Activation of protein phosphatases to dephosphorylate tau isanother strategy under evaluation. The main dephosphorylatingenzyme is protein phosphatase 2A; hence activators of the enzymemay hold benet in AD. Sodium selenite (Ve-015) is a proteinphosphatase 2A activator currently in a phase IIa trial in Australia(www.anzctr.org.au; ACTRN12611001200976). Studies show thepost-translational glycosylation of tau protein with b-N- acetyl-glucosamine occurs at the same threonine and serine residues,pathologically phosphorylated in AD. Another recent researchdemonstrates a parallel GSK3 activation with inhibition of N-ace-tylglucosaminidase (Yu et al., 2012). Hence, maintaining glycosyl-ation is one of the indirect strategies to prevent tauphosphorylation. Thiamet-G is an inhibitor of N-acetylglucosami-nidase, tried with some success animal studies (Liu et al., 2004).

2.2.2. Microtubule stabilizationVarious compounds with microtubule stabilizing effects are in

development. The administration of microtubule stabilizer pacli-taxel to tau-transgenic mice improves fast axonal transport,microtubule density and motor function (Zhang et al., 2005).However being a potent anticancer drug, paclitaxel raises safetyconcerns for its use in non-malignant conditions. Epothilone D isanother microtubule stabilizing compound known for its bloodbrain barrier clearance (Andrieux et al., 2006). Low dose chronicepothilone D administration was able to demonstrate signicantamelioration in microtubule pathology (Brunden et al., 2010). Twoneuropeptides NAP (NAPVSIPQ) and D-SAL (SALLRSIPA) are avail-able that boast microtubule stabilization effects (Gozes, 2011;Gozes et al., 2008). NAP preferentially interacts with neuronaland glial tubulin thus inuences microtubule assembly and dy-namics with some inhibitory effect on tau phosphorylation at lowconcentrations (Matsuoka et al., 2008). Intra nasal administrationof the peptide has also been tried, and preliminary studies haveshown positive results in AD animal models (Matsuoka et al., 2007).BMS-241027 is a small molecule microtubule stabilizer currently inphase I trials in subjects with mild AD (NCT01492374).

2.2.3. Preventing tau oligomerizationA compound that prevents tau interaction and neurobrillary

cology 76 (2014) 27e50tangle accumulation could be extremely helpful in the treatment of

armaAD. Drugs such as astemizole, lansoprazole (both benzimidazole de-rivatives) show a strong afnity to binding tau protein, therefore,indirectly reduce tauetau interaction (Rojo et al., 2010). Althoughlansoprazole decreases tau pathology it has a differential impact onamyloid metabolism; lansoprazole treatment in cell and animalmodels, demonstrate increased amyloid deposition and aggregationprobablydue to itsg-secretasemodulatingeffect (Badiola et al., 2013).The dye methylene blue (methylthioninium chloride) has a range ofpharmacologic effects, one of which is its property to prevent tauinteractions (Congdon et al., 2012). In addition, methylene blue alsohas roles on inhibiting amyloid aggregation (Necula et al., 2007),improving electron transport, decreasing oxidative stress, preventmitochondrial damage (Atamna et al., 2008), regulate autophagy(Congdon et al., 2012) and inhibition of AChE (Pfaffendorf et al., 1997)and hsp70 activity (Jinwal et al., 2009). Rember (TauRx therapeu-tics) was the rst generation proprietarymolecule of methylene bluetried in AD; the drug successfully stabilized the progression of ADover 50 weeks in human studies (Wischik et al., 2008). The encour-aging led to the development of next generation version of thecompound, leuco-methylthioninium (LMTX; TauRx therapeutics).LMTXhas currently been advanced into phase III clinical trials (2012).Recently three natural phenolic compounds obtained fromolives andderived food products hydroxytyrosol, oleuropein and oleuropeinaglycone have also shown effectivity in preventing tau brillizationin vitro (Daccache et al., 2011).

2.2.4. Enhancing tau degradationAnother appropriate strategy is to increase the breakdown of

polymerized tau, thus decreasing its toxicity. Heat shock protein 90(Hsp 90), a chaperone involved in folding the denatured proteins,seems to play a role in preventing tau degradation (Dickey et al.,2007). Curcumin has a wide range of action, one of which is toinhibit Hsp 90 (Giommarelli et al., 2010). Curcumin treatment al-leviates tau pathology in tau transgenic mice by suppressing tangleformation as well as promoting dissolution of already formedtangles (Ma et al., 2013). Various specic inhibitors of Hsp 90 areavailable currently (reviewed by (Zhao et al., 2012), a few of themare clinical trials as an anti-cancer compound. The therapeuticpotential of Hsp 90 inhibitors is under consideration in tauopathicand AD animal models. A brain penetrant Hsp 90 inhibitor, EC102reduced the amount of tau aggregates in the brains of transgenicmice signicantly (Luo et al., 2007). The issue of targeting chap-erone molecules is the potential interference with their basal ac-tivity, which could lead to adverse effects. Although theirtherapeutic value in malignancies is promising, the future role forthese compounds in the eld of neurodegeneration is unclear.

2.2.5. Tau immunotherapyRecently interest in the approaches to promote immunological

clearance of tau tangles has increased tremendously (Rosenmann,2013). Evidence for this principle arose from preliminary studieswhere animals immunized with wild tau protein epitopes devel-oped CNS inltrates and encephalitic response (Rosenmann et al.,2006). Later, investigators modied the approach using the path-ologically phosphorylated epitopes as immunogens. A study using acocktail of different pathological epitopes showed a strong reduc-tion in tau pathology in two different animal models without sig-nicant adverse effects (Boimel et al., 2010). Passive immunizationapproach with monoclonal antibodies against phosphorylated taumolecules also show benet in tau transgenic animal models(Boutajangout et al., 2011; Chai et al., 2011). The treated animalsdisplayed less motor impairment, decrease in phosphorylation oftau and its insoluble aggregates. Monoclonal antibodies against tauoligomers have also been tried with some success in animals

R. Anand et al. / Neuroph(Lasagna-Reeves et al., 2011). The rst translational vaccine trialwill use a conjugated vaccine (tau peptide-KLH-conjugate; AAD-vac1, Axon neuroscience) (NCT01850238). Although the conceptand preliminary results have been encouraging, considering thepresent situation of Ab vaccination trials in humans, the value oftau immunotherapy in patients will only be evident in the future.

2.3. Amyloid targeted strategies

Though Ab bers are one of the pathological hallmarks of AD,evidences suggest that the peptide has several physiological roles(Atwood et al., 2003). The conditions that turn Ab into a patho-logical molecule are not clearly understood but maybe dependenton the concentration of the peptide. When the production of Abexceeds the capacity for its clearance they begin to accumulate,increasing the concentration towards toxic levels. A dynamicequilibrium exists between Ab brils-oligomers-monomers, andoligomeric species are more harmful than others (Jan et al., 2011).When Ab is in abundance, the oligomeric molecules maybe readilyavailable for producing damage. The amyloid based therapeuticstarget various aspects of APP metabolism (Schenk et al., 2012).

2.3.1. Decreasing Ab production- secretase modiersAb peptides originate from the transmembrane protein- APP

following secretase processing. a-Secretase is the principal enzymeacting on APP under physiological conditions followed by g-secre-tase. APP undergoes amyloidogenic processing when acted upon bythe alternate enzyme b-secretase instead of a-Secretase. The ratio-nale ofmodifying secretases stems fromthe idea that augmentinga-secretase, converts APP into nontoxic byproducts, whereas inhibit-ing b and g- secretases, decreases amyloidogenic APP processing.Ligands binding at the cell surface receptors (commonlymuscarinic/GABA agonists) and activation of signaling cascades like proteinkinase C regulate the activity of a-secretase strongly. One of themechanisms of action of the muscarinic M1 receptor agonists andetazolate is to activate a-secretase. Recent investigations have sug-gested that epigallocatechin-gallate, a polyphenol compound fromgreen tea also induces a-secretase activity and thus non-amyloidogenic APP processing (Smith et al., 2010). The antioxidantbenets of the same compound are also worth mentioning. A phaseII trial investigating the drugs benet in early stages of AD iscurrently underway (NCT00951834). Bryostatin 1 (BlanchetteRockefeller Neurosciences Institute) is a potent protein kinase Cactivator and an investigational anti-cancer agent. Results fromdifferent invitro and animal studies suggest the therapeutic poten-tial of bryostatin; the drug is currently in phase II clinical trials.

The g-secretase enzyme performs the processing of multipleclass proteins at the basal level besides APP. One crucial molecule isthe Notch protein that regulates cell proliferation, development,differentiation, growth, cell communication and cell survival status.Twomain classes of drugs that act on the enzyme exist: g-secretaseinhibitors and modulators. The use of inhibitors totally blocks theenzyme and affects its processing of other proteins while g-secre-tase modulators have a Notch sparing effect; despite the differenceneither class of compounds has shown signicant clinical successyet. Some molecules of the nonsteroidal anti-inammatory drugs(NSAIDs) group have g-secretase modulating activity. Tarenurbil,the enantiomer of urbiprofen modulates the g-secretase activityand thereby reduce Ab levels (Eriksen et al., 2003). The phase II trialof the drug showed genuine promise, whereas the phase III trialsproved to be a considerable disappointment and so, is no longerused. Subsequent analyses of the study results demonstrated thereasons for tarenurbils failure. In the study, CSF levels of taren-urbil attained weremuch lower than the predicted concentrationsfrom the preclinical evidence; in addition there was no change in

cology 76 (2014) 27e50 33CSF Ab levels (Galasko et al., 2007). The apparent benet of

armatarenurbil in phase II trials was, in fact, due to an acceleratedcognitive worsening in the placebo arm. The anti-inammatoryactions of tarenurbil could also be a contributory reason for theobserved failure (Imbimbo, 2009).

Semagacestat is another g-secretase inhibitor that reachedclinical setting, but in a large phase III trial the drug produceddisappointing results and so abandoned before its completion(Imbimbo et al., 2011). Another specic g-secretase inhibitorELND006 showed therapeutic benet in animal studies and waslead to clinical evaluation, but similar to semagacestat, the drugproduced signicant adverse effects and the clinical trials met withfailure (Hopkins, 2011). The analysis of failure of two specic g-secretase inhibitors has yielded valuable lessons. It became awarethat the g-secretase enzyme system plays a pivotal role in manysignaling pathways mediated by proteins such as p75NTR, Notch,CD46 and about 50 additional substrates. Some of the proteinsregulate cell survival, neurogenesis, cell communication, choles-terol metabolism and angiogenesis (Beel and Sanders, 2008). Hencethe implications of targeting g-secretase are not fully understood.The anti-amyloid effects obtained by inhibiting the enzyme have tobe weighed against its affected physiological roles; this fact couldbe one of the major reasons for the failure of the g-secretase in-hibitors. The exact role of these drugs on Ab peptide dynamics isalso not clear. A recent model based metanalysis has estimated thatusing the g-secretase inhibitors the Ab levels drop at the time ofdosage but increase during the dosing intervals. The net effect isthat Ab levels increase over a 24 h period, which could provide apartial explanation for the failure of these drugs (Niva et al., 2013).

BMS-708163 (Avagacestat; Bristol-Myers Squibb) is a g-secre-tase modulatorwith Notch-sparing effect; the drug had a favorablesafety prole in a small phase II trial (Dockens et al., 2012). Twolarge studies in early as well as mild tomoderate AD for a treatmentperiod of 104 weeks are on-going (NCT00810147, NCT00890890).GSI-953 (Begacestat; Wyeth) is another Notch-sparing moleculethat targets g-secretase (Martone et al., 2009). The results from itsphase II trial are not available (NCT00547560, NCT00959881). Fewg-secretase modulators like NIC5-15 have completed phase II trialwhile other drugs-E2212, GSI-1 are in different stages ofdevelopment.

A variety of molecules targeting b-secretase are available, butonly a few have entered clinical trials to date (Ghosh et al., 2012).CTS-21166 (CoMentis, USA) was one of the rst molecules tocomplete phase I trials successfully, but it did not undergo subse-quent testing (Hsu, 2010). MK-8931 (Merck, USA) has completedthe phase I trial (NCT01496170) now in a phase II/III trial(NCT01739348). LY2886721 (Eli Lilly and company) is the other b-secretase inhibitor that is in phase I/II trial (NCT01561430). GRL-8234 is an experimental b-secretase inhibitor that has shownpositive results in preclinical studies; when administered for a longterm it demonstrates a reduction in Ab load in transgenic micewithout adverse effects (Chang et al., 2011).

Posiphen (QR Pharma Inc.) is a () enantiomer of phenserine,that decreases the levels of APP directly. By interacting with the 50-untranslated region of APP mRNA, posiphen inhibits ribosomalaccess and blocks translation; it is thus effectively an APP synthesisinhibitor (Shaw et al., 2001). Posiphen and its principal metabolitesalso possess sufcient inhibitory effect on interleukin-1b, synthesisof a-synuclein and AChE activity (Yu et al., 2013). In a small pilotstudy, posiphen showed favorable reduction in the levels of CSFbiomarkers (Maccecchini et al., 2012); the compound is currently ina phase I trial (NCT01072812).

2.3.2. Modulating Ab transportApolipoproteins play prominent roles in Ab metabolism and

R. Anand et al. / Neuroph34transport (Fan et al., 2009); though they do not cross the bloodbrain barrier, they regulate the movement of Ab between thecentral nervous system and periphery (Ladu et al., 2000). Apoli-poprotein E 34 (ApoE 34) increases passage of Ab from blood to brain(Martel et al., 1997). This transport is receptor mediated, and thelow-density lipoprotein receptor-related protein (LRP) plays acentral role in the process (Zlokovic, 2004). With age, LRP expres-sion decreases, impairing Ab efux contributing to prolonged Abstay in the brain (Shibata et al., 2000). Antibodies against LRPreduce Ab efux from the brain (Shibata et al., 2000), and periph-eral administration of soluble LRP to increase Ab efux from brainhas also been proposed as a potential treatment strategy in AD(Sagare et al., 2007).

Receptor for advanced glycation end products (RAGE) is a multi-ligand receptor that binds Ab with high afnity and facilitates theentry of Ab into CNS; Ab binds to RAGE at the bloodebrain barrierand contribute to increased CNS entry, inammation and neuronaldeath (Chen et al., 2007). The expression of RAGE increases in AD(Lue et al., 2001). PF-04494700 was the rst oral small moleculeantagonist of RAGE tried in humans; the drug had an acceptablesafety prole in phase I trials (Sabbagh et al., 2011), but the phase IItrial was a failure (NCT00566397). Creating a soluble RAGE receptoranalog that would serve as a decoy receptor thus, reducing ligandbinding is a valid approach. The infusion of soluble receptor hasshown signicant benets in transgenic animals (Arancio et al.,2004). One such soluble RAGE receptor molecule TTP4000 (Trans-tech Pharma) is currently in phase I trials (NCT01548430). FPS-ZM1is a novel multimodal and specic RAGE receptor developedrecently. The studies with the compound has demonstrated sig-nicant blood brain barrier clearance, reduced amyloid deposition,and improved cognitive and cerebrovascular parameters in trans-genic mice (Deane et al., 2012).

2.3.3. Decreasing Ab aggregationTramiprosate is a glycosaminoglycan that binds to monomeric

Ab, preventing its oligomerization and aggregation (Wright, 2006).The drug demonstrated promising prole in phase II studies. Alarge phase III trial conducted subsequently had intriguing obser-vations. The primary endpoints of the study did not show signi-cant treatment related benets. However, subsequent secondaryanalyses of data suggested the probable disease modifying benetsof the drug. Tramiprosate administration produced domain speciccognitive improvement in various aspects of memory, praxis skillsand language (Saumier et al., 2009). Treated subjects showed astatistically insignicant but slowing down pattern of cognitivedecline when assessed with Alzheimer Disease Assessment Scaleecognitive subscale but not with Clinical Dementia Rating e Sum ofBoxes scoring (Aisen et al., 2011). Volumetric magnetic resonanceimaging showed a signicant reduction in hippocampal atrophy indrug treated subjects (Gauthier et al., 2009). Although the effect oftramiprosate on Ab appears benecial, its effect on tau metabolismis unclear. An experimental study has shown that tramiprosateadministration leads to unexpected tau aggregation, raisinganother potential warning statement (Santa-Maria et al., 2007).

ELND005 (Scyllo-inositol) is another compound known for itsanti-oligomerization properties. The compound effectivelydecreased insoluble Ab and reversed cognitive decline in transgenicmice (DaSilva et al., 2009). A phase II trial evaluated different dosesof the compound in mild to moderate AD patients, but only the lowdose group completed the trial. Few phase II trials with ELND005including a 12 week extension study are currently ongoing(NCT01766336, NCT01735630). Colostrinin or proline rich peptidecomplex was isolated rst from ovine colostrum. The proteincomplex has strong immunoregulatory properties besides which itaffects learning, memory and cognitive functioning (Janusz and

cology 76 (2014) 27e50Zablocka, 2010). Experimental data indicate colostrinin can

armaprominently inhibit the aggregation of Ab peptides and dissolvepre-formed brils (Janusz et al., 2009). One of the early pilot trialsstudied the effect of oral colostrinin for 3 weeks and observedpositive outcome in subjects with AD (Leszek et al., 2002). A sub-sequent open label double blinded trial examined the dose effectfor 30 weeks where colostrinin was able to increase the cognitiveand daily activity of subjects with minimal adverse effects(Bilikiewicz and Gaus, 2004). The compound has not been testedsubsequently. A recently characterized inhibitor of Ab aggregationD737 has shown signicant benets in Drosophila model (McKoyet al., 2012). Some novel peptoid compounds with anti-aggregation properties have also been developed (Luo et al., 2013).

The hormone melatonin, discussed under antioxidants alsoseems to possess anti-Ab aggregating properties. Administration ofmelatonin to Ab-overexpressing transgenic mice decreased senileplaque accumulation (Olcese et al., 2009); however, whether it is adirect effect of melatonin or secondary to its antioxidant and anti-inammatory actions is not clear. Gelsolin an actin-binding proteinis a key regulator of actin lament assembly and disassembly.Intracellularly gelsolin is present in the cytosol and mitochondriaand extracellularly in plasma and cerebrospinal uid (Kwiatkowskiet al., 1988). Plasma gelsolin levels decrease in AD and the levelspositively correlates with rate of decline in patients (Guntert et al.,2010). Experimental evidence suggests that gelsolin binds Ab, in-hibits its brillization, dissolves the preformed Ab brils and ac-celerates its removal; therefore, it is another possible therapeuticcandidate in AD (Chauhan et al., 2008).

2.3.4. Increasing Ab clearanceSome proteases that degrade the Ab plaques are plasmin,

neprilysin, insulin degrading enzyme, endothelin-convertingenzyme, angiotensin-converting enzyme and metalloproteinase9; a few other proteases also play a minor role (Nalivaeva et al.,2012). The levels of Ab degrading enzymes decline in AD whichmay contribute to Ab accumulation (Yasojima et al., 2001). Tissueplasminogen activator activates plasmin, but plasminogen activatorinhibitor 1 blocks this action. Experimental evidence suggests thatinhibitors of plasminogen activator inhibitor 1 decrease the plasmaand brain Ab levels in transgenic animals (Jacobsen et al., 2008).Increasing neprilysin levels through viral vector-delivered geneexpression has been successful in animalmodels (Marr et al., 2003).The peptide hormone somatostatin regulates Ab clearance throughactivation of neprilysin (Saito et al., 2005). The expression of so-matostatin in the brain decreases with age (Gahete et al., 2010);therefore targeting neprilysinwith somatostatin or its analogs is anoption in AD. Recently small molecule activators of insulindegrading enzyme have also been identied. Although many pro-tease inhibitors are available, the therapeutic benet of the prote-ase activators in AD is yet to be evaluated (Cabrol et al., 2009). Beinga non-specic approach, use of protease inhibitors will requirefurther evaluation. More controlled or targeted delivery approachesmay play a valuable role in AD.

2.3.5. Amyloid targeted immunotherapyResults from animal experiments have shown the benecial

effect of anti-amyloid immunization approaches. AN1792 was therst active vaccine tried in humans. The vaccine used a full lengthaggregated amyloid peptide (Ab1-42). During the trial about 6% ofsubjects developed meningoencephalitis hence, was discontinued(Orgogozo et al., 2003). The initial results were discouraging, andCSF biomarkers or their cognitive performances did not show sig-nicant differences. Later it became clear that about 60% of patientsresponded with antibody production and neurophysiological testbattery indicated favorable performance in them (Gilman et al.,

R. Anand et al. / Neuroph2005). Volumetric imaging in the vaccine responders at 10e11months showed an accelerated brain and hippocampal volume loss,but there was no reduction in cognitive scores (Fox et al., 2005). Atthe end of 4 years, there was a slowing down of functional declinein vaccine responders compared with placebo, but neither grouphad differences in cortical volume loss pattern (Vellas et al., 2009).Neuropathological examination in 8 vaccine treated subjectsshowed signicant clearance of amyloid plaques (Holmes et al.,2008). Although following vaccination, the subjects showed aslowed cognitive decline and superior plaque clearance it did notstop their progression to severe stages (Holmes et al., 2008).

To avoid the non-specic immune response that might arise dueto full length peptides, investigators designed the next generationvaccines against small epitopes. CAD106 has rst 6 amino acids asthe immunogenic sequence (Ab1-6). The vaccine treatment did notcause adverse effects in a 52 weeks trial; adequate antibodyresponse occurred in more than 75% of subjects with two differentdosages (Winblad et al., 2012). The vaccine has recently completedphase II trials (NCT00956410, NCT01097096). ACC-001 is anothersecond generation vaccine currently in phase II trials(NCT01284387, NCT01227564, and NCT00479557). A few novelanti-amyloid vaccines like MER5101, Lu AF20513 have been suc-cessful in the pre-clinical studies and some candidate vaccines likeACI-24, Aftope AD-02/AD-03, UB-311 and V-950 are in differentphases of development (Davtyan et al., 2013; Galimberti et al.,2013; Liu et al., 2013). More recently hybrid vaccines have comeup; an orally administered Ab-RAGE complex vaccine has shownbetter results in transgenic mice than Ab administration alone(Webster et al., 2012).

Bapineuzumab is a humanized anti-Ab monoclonal antibodydirected against its N-terminus. It was the rst passive immuni-zation therapeutic tried in AD, but the results from its phase II trialwere inconclusive. APOE 34 carriers did not show treatment differ-ences, but in non-carriers there was a signicant improvement incognitive and functional end points (Salloway et al., 2009). Theinvestigators predicted that observation could be due to theincreased amyloid plaque burden in APOE 34 carriers. The phase IIItrials did not show signicant differences in the primary endpointirrespective of the APOE 34 genotype, but they had a high incidenceof amyloid related imaging abnormalities (Sperling et al., 2012). Afew investigators initiated trials especially in APOE 34 carriers, butcurrently no trials on bapineuzumab are underway (Galimbertiet al., 2013). In a pooled analysis of the results of CSF biomarkersfrom two trials, bapineuzumab arm showed signicantly decreasedlevels of phosphorylated tau compared with placebo. CSF Ab levelswere unchanged, but the total tau levels decreased signicantlyfrom baseline in the treated group (Blennow et al., 2012). Follow upneuropathological examination in 3 subjects administered withbapineuzumab has shown some intriguing ndings. There was nodifference in plaque densities or distribution between immunizedand non-immunized subjects, but there was an increase in solubleAb deposits with a low Ab42: Ab40 ratio suggesting that bapineu-zumab had an impact on Ab dynamics (Roher et al., 2013).

Solanezumab was the second anti-Ab antibody that enteredtrials in AD. The phase II trial with solanezumab did not show anytreatment related benets, but it produced a dose dependent in-crease in plasma and CSF Ab42 concentrations (Siemers et al., 2010).Encouraged by the results solanezumab entered phase III trials(EXPEDITION 1 and EXPEDITION 2). Results of co-primary end-points did not reach signicance in both studies, but, pooled dataanalyses showed a signicant reduction in cognitive decline inimmunized patients (Tayeb et al., 2013); solanezumab is about toenter another set of phase III trials including the AsymptomaticAlzheimers disease (A4) trial (Corbyn, 2013). Gantenerumab(Hoffmann-LaRoche) is another monoclonal antibody that has

cology 76 (2014) 27e50 35shown promising results in preclinical studies. Gantenerumab

armapreferentially interacts with aggregated Ab in the brain withoutaffecting plasma Ab levels suggesting its systemic specicity(Bohrmann et al., 2012); the effect is dose dependent (Ostrowitzkiet al., 2012). Gantenerumab is currently in a phase III trial(NCT01224106). Gantenerumab and solanezumab are together in aphase II/III trial that attempts to assess their potential in individualsat risk for dominantly inherited AD (DIAN TU; NCT01760005). Toreduce the risk of Fcg receptor-mediated overactivation of micro-glia a novel antibody of IgG4 type-crenezumab (MABT5102A;Genetech) is available (Adolfsson et al., 2012). A phase II trial ofcrenezumab in mild to moderate AD is ongoing (NCT01343966).Alzheimers Prevention Initiative trial will examine the benet ofcrenezumab exclusively in individuals carrying Presenilin1 muta-tions (Corbyn, 2013). PF-04360365 has completed phase II trials,and results are awaited. A few other passive immunotherapeuticsare in various stages of development include GSK933776A, NI-101,PF-05236812, RN6G, SAR-228810, BAN-2401 (Galimberti et al.,2013).

Even healthy volunteers have detectable levels of anti-Ab anti-bodies in their plasma, but the antibody titres decrease in AD(Weksler et al., 2002). A 6 month pilot study with 5 patients wasone of the rst reports of the benet of IVIG administration in AD.Therewas no change in MMSE scores but the trial saw a decrease inCSF Ab levels with a concurrent increase in serum total Ab levels(Dodel et al., 2004). A subsequent different phase II study carriedout IVIG infusion for 15 months with a 3 month interim discon-tinuation period. The levels of Ab in CSF decreased during the studyperiod signicant improvement in the MMSE score; however, Ablevels returned to baseline during the discontinuation (Relkin et al.,2009). In a dose nding phase II trial, IVIG (Octagam; OctapharmAG) produced a signicant reduction in area under the curve forplasma Ab for one of its doses; the greatest reduction was in thesecond week interval, contrasting it from the other studies (Dodelet al., 2013). Baxter Corporation announced the results of theirlarge phase III study recently. The preliminary analysis of data hasyielded disappointing results with no signicant treatment benetsin the treated group (http://www. Baxter.com). The subsequentdetailed analysis may provide valuable information.

2.4. Targeting intracellular signaling cascades

Ab bers activate various intracellular pathways, so drugs thatdisrupt the abnormal pathways could be useful in AD. cAMPsignaling and nitric oxide/cGMP/cGMP-dependent protein kinase/cAMP responsive element-binding protein cascade is prominentlylinked to Ab induced synaptic decits and memory loss associatedwith AD (Puzzo et al., 2009). In light of this process, pharmacologicinhibition of PDE provides signicant benet in experimentalmodels. Caffeine and etazolate both possess PDE inhibiting activitythat could play a role in their pharmacologic effects. Rolipram is aPDE-4 selective inhibitor that effectively reversed memory andcognitive decits in Ab treated mice (Cheng et al., 2010); sildenal,a PDE-5 inhibitor also produced similar results (Puzzo et al., 2009).Cilostazol inhibits PDE-3 activity and its administration protectedtransgenic mice from Ab mediated damage, decreased amyloidaccumulation and tau phosphorylation (Park et al., 2011); benet ofcilostazol is currently under evaluation in demented subjects(NCT01409564). Some PDE inhibitors that are in preclinical stagesof development include AVE-8112, BCA-909, GEBR-7b and THPP-1(Froestl et al., 2013).

Growing evidence also indicates disturbed lipid signalingpathways in AD; arachidonic acid and phospholipases seem to bekey mediators of Ab induced pathogenesis in animal models of AD.The levels of activated phospholipase A2 group IV isoform increase

R. Anand et al. / Neuroph36in the hippocampus of AD transgenic mice as well as patients withAD (Sanchez-Mejia et al., 2008). Evidences also show that ablationof phospholipase A2 protects rats from cognitive decits and de-creases the levels of total tau in brain (Sanchez-Mejia et al., 2008;Schaeffer et al., 2011). Rilapladib is an oral inhibitor of lipoproteinassociated phospholipase A2 that is currently in phase II trials(NCT01428453). Interestingly enough, activation of phospholipaseA2 is also suggested as a strategy for cognitive enhancement in AD(Schaeffer et al., 2009). Aberrant phospholipase C signaling alsocontributes to cytosolic Ca2 surge and excitotoxicity. But, the valueof its modulation is not clear because, phospholipase C appears topossess a differential role (Shimohama et al., 1998). Recently therole of phospholipase D2 has also been highlighted in AD (Oliveiraet al., 2010).

2.5. Tackling oxidative stress

Building oxidative stress is a crucial pathogenetic process in AD(Zhu et al., 2007) hence it becomes necessary to devise strategiesthat reduce the oxidative burden in cells.

2.5.1. Exogenous antioxidantsEvidence for the protection offered by antioxidants including

vitamins (E, C, and carotenoids), phytochemicals and syntheticcompounds in AD is inconsistent. Administration of vitamin E intransgenic AD models at early ages reduced lipid peroxidation andplaque burden, but supplementation trials in humans have notshown convincing evidence of their benet (Farina et al., 2012).Combination of vitamin E with donepezil did not provide addi-tional benet in patients with AD or mild cognitive impairment(Petersen et al., 2005). A phase III trial combining vitamin E withmemantinewas completed recently, and the results will answer themany queries that surround the benet of vitamin E if any (Dyskenet al., 2013). A phase III trial of vitamin E and selenium is currentlyongoing (NCT00040378). Flavonoids and carotenoids, the othergroup of ubiquitous antioxidants have also shown neuroprotectiveeffect in experimental setups (Kelsey et al., 2010; Khalili andHamzeh, 2010). Rutin, a avonoid compound, protected rats fromstress induced damage and neuroinammation induced by strep-tozotocin (Javed et al., 2012). Lutein is a natural carotenoid withcytoprotective effect (Vijayapadma et al., 2012); when supple-mented in combination with docosahexaenoic acid, verbal uency,memory scores and rate of learning improved in elderly women(Johnson, 2012).

The spice curcumin has shown several benecial roles (antiox-idant, anti-inammatory, and amyloid-disaggregating properties)in experimental studies (Lim et al., 2001). Curcumin decreases Ab-induced inammation and modestly inhibits b-secretase and AChE(Yang et al., 2005). Animal experiments have shown substantialbenets, but human studies report no signicant differences incognitive function between placebo and curcumin groups (Baumet al., 2008; Ringman et al., 2008). An 18 month study of curcu-min in age related cognitive impairment is currently underway(NCT01383161). Other natural antioxidants such as blueberry(Joseph et al., 2003) red grape (Ho et al., 2009) have also beenstudied in transgenic mice, for their resveratrol content, and theresults show reduced plaque burden and improvement in behav-ioral decits (Karuppagounder et al., 2009). Epidemiologic studiesin humans also show that moderate to mild wine consumptionreduces the risk of dementia including AD (Orgogozo et al., 1997;Truelsen et al., 2002). Resveratrol is currently in a phase II trial(NCT01504854).

Melatonin is another potent antioxidant compound with addi-tional pleiotropic effects. Several mechanisms contribute to thedisease modifying potential, including inhibition of Ab generation,

cology 76 (2014) 27e50aggregation, formation of amyloid brils, attenuation of tau

armahyperphosphorylation, mitochondrial protection antioxidant andantiapoptotic effects (Wang andWang, 2006). Human studies showimproved cognitive and neuropsychiatric performance includingeven in a progressed state when treated with melatonin (Cardinaliet al., 2010). The utility of melatonin is not without controversy andreports indicate that the administration of melatonin after amyloidplaque deposition has occurred is not benecial (Quinn et al., 2005)but may offer neuroprotection in the early stages of disease(Gunasingh et al., 2008). Currently a phase II trial with melatonin inprolonged release form is ongoing (NCT00940589). Melatonin re-ceptor agonism has also been tried, but a short phase II trial with amelatonin agonist ramelteon was not able produce any signicanttreatment related parameters (NCT00325728). Neu-P11 is a novelmelatonin agonist that has attenuated neuronal loss and improvedmemory performance in rats (He et al., 2013).

To summarize the section, in experiments antioxidant com-pounds show promising results, but their translation to the clinicalsituation is less successful. With regards to the natural compounds,the epidemiological data consistently show a decreased incidenceof AD in the population. This has led to their large scale supple-mentation of trials, which have almost always yielded negativeresults. The failure of the natural compounds in human trials couldbe attributed to one of the many reasons. The other side of manyantioxidant molecules is that they behave as pro-oxidants,depending on their concentration and cellular environment(Jones, 2006). This double edged nature of antioxidant compoundsis one signicant reason for their failure. The heterogeneity in thepharmacokinetic response, in a population, is also to be borne inmind. A study by Lloret et al. (2009) examined the antioxidantredox status in patients treated with vitamin E. Interestingly thesubset of patients who failed to respond to vitamin E had lowerglutathione levels in blood, compared to responders. The high dosetherapy in that subset of non-responders led to rapid cognitivedecline. The mechanism behind the non-responsiveness is notclear. Bioavailability and tissue distribution of drugs is also a factorthat plays a role. Studies indicate that the bioavailability of vitaminE is different between smokers and nonsmokers so smoking couldbe a signicant confounding factor in determining drug responses(Lodge, 2005). Curcumins bioavailability is also dependent onvarious factors such as intestinal metabolism and absorption. Oraldosing fails to raise the serum concentrations of curcumin or itsmetabolites to detectable levels. The issue of their CNS clearance is,by far, the biggest issue that needs consideration while using themin AD. Having said that, the blood brain barrier clearance and CNSbioavailability following an oral dose of curcumin is a critical factor.Similar could be the case with resveratrol; due to an extensive rstpass metabolism, the bioavailable concentration of resveratrol isonly about 1% (Walle, 2011). Beyond tissue distribution, the abilityof the compounds to penetrate into the mitochondria, the heart ofcellular ROS production could be another crucial factor. Epidemi-ologically what we perceive as a benecial relationship could bereective of their cumulative protective role of these compounds.Results of different trials and experiments also point towards an-tioxidants value in the early stages of the disease than when suf-cient damage has already been done. Considering the long periodof latency in AD, appropriate timing of the treatment could also be asignicant determinant of a drugs success.

2.5.2. Facilitating endogenous antioxidant defenseThe primary endogenous antioxidant pathway is the nuclear

receptor factor 2 (Nrf2)/antioxidant response element (ARE)cascade. Nrf2 on activation gets translocated from the cytosol intothe nucleus where it regulates the expression of several genes. Thistranslocation process is blocked in AD; hence, different compounds

R. Anand et al. / Neurophthat can reestablish the pathway are under consideration (Ramseyet al., 2007). Facilitation of Nrf2 expression by tertbutylhy-droquinone or adenovirus-mediated gene delivery techniquesprotects cells against Ab induced toxicity (Kanninen et al., 2008).Similar studies have also been conducted with other inducercompounds like triterpenoid CDDO-methylamide, and the effectshave been encouraging (reduced plaque burden, Ab42 levels,inammation, and oxidative stress) (Dumont et al., 2009). Hencedrugs inducing the Nrf2/ARE pathway could be a key approach forthe treatment of AD. Peroxisome proliferator-activated receptor-gco-activator 1 alpha (PGC-1a) is another protein that plays multipleroles in mitochondrial biogenesis, energy metabolism and mito-chondrial antioxidants expression. In the human brain tissues, theexpression of PGC-1a decreases with progression of dementia (Qinet al., 2009). PGC-1a is a crucial transcription cofactor interactingwith the nuclear receptor peroxisome proliferator-activated re-ceptor-g (PPAR-g) thus regulating the downstream genes. PPAR-gagonists have also been studied as potential therapeutics for ADtreatment, and observations have suggested that pioglitazone,activator of PPAR-g improves mitochondrial oxidative metabolism(Skov et al., 2008). In animal models pioglitazone modies variousindices of aging but does not slow down the cognitive decline(Blalock et al., 2010). Pilot studies demonstrate the value of pio-glitazone in diabetic AD patients but not in nondiabetics(Geldmacher et al., 2011; Hanyu et al., 2009). The current evidenceregarding the benet of modulating the above pathways is still notsufcient.

2.6. Mitochondria specic therapy

Though the traditional antioxidants achieve their way tacklingthe produced reactive oxygen species (ROS), it is more relevant tocontrol the source of ROS production; hence, mitochondrial drugtargeting is tried for various disorders. Coenzyme Q10 (CoQ10) alsoknown as ubiquinone, is a protein shuttling electrons from complexI and complex II in the electron transport chain (ETC). CoQ10 sup-plementation has potential neuroprotective effects including sup-pression of ROS production, minimized ROS injury and stabilizationof mitochondrial function (Lee et al., 2009). The effect of CoQ10 inpreclinical studies has been encouraging, but its effect, neither asmonotherapy nor combination has been helpful in human studies.In one of the rst phase III trials, CoQ10 did not show therapeuticbenet in Parkinsons disease; therefore, the study was terminated(NCT00740714). Idebenone, a water-soluble analog of ubiquinonehas also been tried in humans; although an earlier study demon-strated dose dependent benecial effects on cognition and diseaseprogression for up to 2 years (Gutzmann and Hadler, 1998), asubsequent study found no signicant effect (Thal et al., 2003). Thelack of efcacy of the drugs could be due to the fact that bothidebenone and CoQ10 require an intact and fully functional ETC andsince mitochondria sustains defects in ETC in AD they may be ofquestionable value (Parker et al., 1994). Methylene blue (discussedearlier) also seems to serve as an alternative electron carrier,bypassing complex I/III blockage thus offering a role in neuro-protection (Wen et al., 2011). An imbalance in mitochondrial dy-namics with altered levels of ssion and fusion proteins also occurswith AD pathology (Wang et al., 2009); hence an agent that pre-serves mitochondrial dynamics may play a protective role in AD.

Other mitochondrial antioxidants that are under considerationinclude acetyl-L-carnitine and R-a-lipoic acid, both of which havedemonstrated a reduction in oxidative stress and mitochondrialabnormalities in animals (Siedlak et al., 2009). Lipoic acid alsoseems to increase acetylcholine production, chelate transitionmetals, scavenge free radicals, and down-regulate the expression ofpro-inammatory proteins (Maczurek et al., 2008). Lipoic acid in

cology 76 (2014) 27e50 37combination with vitamin E and C was compared with CoQ10 or

armaplacebo in a small phase I trial. It was observed that there was alowering of CSF isoprostane levels in the rst group indicating areduction in oxidative stress, but other groups did not showchanges (Galasko et al., 2012). However this benecial effect onoxidative stress did not reect in the CSF Ab, total or phosphory-lated tau levels and contrastingly a faster cognitive decline occurredin subjects (Galasko et al., 2012). Lipoic acid and omega-3-fattyacids combination therapy are currently in two phase I/II trials(NCT01780974, NCT01058941).

One promising compound for mitochondrial targeted treatmentis the triphenylphosphonium linked ubiquinone derivative, MitoQ(Kelso et al., 2001). MitoQ selectively accumulates in the organelle,continually recycled by mitochondrial enzymes, and it can functioneven in the absence of an intact ETC, which makes it a potentantioxidant compared to the untargeted ones (Murphy and Smith,2007). Szeto-Schiller peptide (SS-31) is also a novel mitochondrialtargeted ROS scavenger therapy; bothMitoQ and SS-31 have showngood results in cell lines (Manczak et al., 2010). However, MitoQtreatment produced disappointing results in a phase II trial inParkinsons disease patients (Snow et al., 2010). MitoVitE is a po-tential follow up compound of MitoQ. Selective mitochondrialpenetrating cations like SkQR1 and C12R1 have also been identiedrecently (Chernyak et al., 2013).

Opening of mitochondrial permeability transition pore (mPTP)is a key event in mitochondrial dysfunction, so compounds block-ing the process are evaluated in AD. Studies of the antihistaminedrug dimebon indicated its ability to block mPTP opening andprotect against cellular dysfunction and death (Bachurin et al.,2001). Phase II trial of the drug showed improvement in cogni-tive scores in the treated group, which led to phase III trials withinterest (Doody et al., 2008), but the trials did not show clinicalimprovement (Jones, 2010). The failure of dimebon was a tremen-dous disappointment. So what are the reasons for dimebons fail-ure? Dimebon is a drug with a complex pharmacology and severalpotential targets. At its working concentration dimebon is capableof inhibiting AChE and also interact with 5-HT6, dopamine 1, 5-HT4and 5-HT3 receptor amongst others (Okun et al., 2010). Hence thecognitive outcome would be dictated by the balance between thepro-cognitive (AChE inhibition, 5-HT4 activation) and anti-cognitive (5-HT6 stimulation on pyramidal cells and 5-HT3 stimu-lation on inhibitory interneurons) effects. The complex interactionof dimebon along with the alteration in the receptor levels thatoccur with disease progression can in part explain the lack ofbenet in the larger studies (Geerts et al., 2012). In addition,polymorphisms in COMT Val158Met gene modify the interaction ofdimebon with dopamine 1 receptor; hence, genotype variation inthe population could contribute modestly to the pharmacodynamicprole. Although some mitochondrial therapeutics exists, theirtranslation into the clinical setting is not successful. A key aspectthat would need attention in this regard would be to consider theimpact of drugs on the normal functioning tissue mitochondria.Compounds that will selectively targets diseased mitochondria willbe extremely helpful in the setting of neurodegenerative diseases.

2.7. Targeting cellular CA2 handling

Since perturbed Ca2 homeostasis is one of the major mecha-nisms in AD, it is prudent to evaluate drugs that target differentCa2 signaling pathways. Memantine produces modest decreasesin Ca2 inux thus reducing excitotoxicity (Lipton and Chen, 2004).Pre-clinical evidence suggests that antagonists of NMDA receptorwith GluN2B subunit (ifenprodil and Ro25e6981) and ligands ofmetabotropic mGluR5 receptors (MPEP) protect neurons from Abtoxicity (Rammes et al., 2011). EVT-101 is one of the few GluN2B

R. Anand et al. / Neuroph38antagonists to complete a phase I trial (NCT00526968). Novelstrategies modulating mitochondrial Ca2 handling have beenreviewed already (Hung et al., 2010). In animal studies both min-ocycline (Garcia-Martinez et al., 2010) and nonaspirin nonsteroidalanti-inammatory drugs (NSAIDS) (Sanz-Blasco et al., 2008) showmitochondrial membrane depolarization and reduce Ca2 intakeinside the mitochondria. Minocycline in addition, has an effect onvoltage dependent anion channels (Garcia-Martinez et al., 2010).KB-R7943 is a selective inhibitor of the Na/Ca2 exchanger whichcauses depolarization of isolated brain mitochondria and reducesmitochondrial Ca2 uptake (Storozhevykh et al., 2010). The Abinduced disturbances of Ca2 homeostasis also seem to be atten-uated by Ginkgo biloba extract (EGb761) (Shi et al., 2010), but longterm use of the compound did not provide protection from cogni-tive declinewhen compared to placebo (Vellas et al., 2012). Anotherinteresting nding was that triuorocarbonylcyanide phenyl-hydrazone (FCCP) a strong uncoupler inhibited mitochondrial Ca2

surge triggered by Ab1e42 oligomers (Sanz-Blasco et al., 2008).Although the ndings are encouraging, FCCP is a toxic drug. Thedirect effect of NSAIDs on mitochondrial membrane potential hasalso not been well established accordingly. A specic but low po-tency compound that modulates the uncoupling proteins may havetherapeutic benets. The effect of thyroid hormones, the physio-logical uncoupler on Ca2 homeostasis is also intriguing (Cole et al.,2012). Thyroid hormone supplementation also seem to lessencognitive decits in a mouse model of AD (Fu et al., 2010), but itsvalue in humans is yet to be studied. Although during supple-mentation, a concern of interfering with the thyroid hormone axisdoes exist, the approach warrants detailed investigation.

2.8. Anti-inammatory therapy

The abundant evidence for neuroinammation in the diseaseprocess had prompted various work groups to try NSAIDs in AD.Benet of NSAIDs may involve a variety of mechanisms apart fromtheir cyclooxygenase inhibition like maintaining Ca2 homeostasis,targeting g-secretase (Weggen et al., 2003), Rho-GTPases (Fu et al.,2007), and PPAR (Nicolakakis et al., 2008). Through their effect onRho-GTPases, NSAIDs manage various phenomena related to ADincluding axon growth (Fu et al., 2007), tau phosphorylation (Sayaset al., 1999), and astrocyte motility (Lichtenstein et al., 2010).Numerous epidemiological studies and clinical trials regarding thebenet of NSAIDs in AD are available (Imbimbo et al., 2010). Theepidemiological data point to a reduced incidence of AD in NSAIDusers (Cornelius et al., 2004; Lindsay et al., 2002), but data frommost clinical trials in AD andmild cognitive impairment have eithershown neutral or harmful effects (Aisen et al., 2003; Reines et al.,2004). A few studies show that NSAIDs are effective only in theApoE 34 carrier subpopulation (Hayden et al., 2007; Szekely et al.,2008). The disappointing results from clinical trials led to adecrease in pursuance of anti-inammatory therapy in AD. How-ever a recent study has rekindled the hope of NSAIDs in AD;administration of a COX-1 selective inhibitor, SC-560 in tripletransgenic mice has reduced inammation, neuropathology andimproved cognitive performance (Choi et al., 2013). The reasons forthe failure of the NSAIDs in AD may be related to the inhibitoryeffects on microglia. It is well known that ApoE 34 allele confers aninammatory state in the brainwhen compared with ApoE 33 (Viteket al., 2009). Physiologically, microglial cells seem to play promi-nent roles in the clearance of Ab plaques by the process of phago-cytosis and also by secreting various proteases (Lee and Landreth,2010). The evidence for the role of microglia and their inamma-tory mediators in modulating neurogenesis is also compelling(Ekdahl et al., 2009). The clearance function declines with agewherein accumulation of plaques continues in spite of increasing

cology 76 (2014) 27e50microglial numbers (Hickman et al., 2008). Considering the

armacomplex microglial interaction in brain functions, the result of anti-inammatory treatment is difcult to be predicted.

2.9. Other approaches in AD

2.9.1. GonadotropinsVarious studies have reported the relationship between AD and

the hormone dyshomoeostasis secondary to reproductive senes-cence. The hormones-testosterone, estrogen and progesterone, areneuroprotective, but their levels decrease with aging. Luteinizinghormone (LH), on the other hand, supports the disease process, butthe concentrations increase with aging (reviewed in (Barron et al.(2006)). Though the role of the hormones may not be direct orstrong, they could signicantly contribute to the pathogenesis. Therole of hormone replacement therapy (HRT) is still controversial;studies indicate that prior use of HRT decreases the risk of AD inwomen, but current use is not useful unless used more than 10years (Zandi et al., 2002). Another recent study shows that low doseestrogen therapy decreases the risk of AD (Yue et al., 2007) thebenet is more in non- 34 population (Burkhardt et al., 2004).Testosterone supplementation in males improves cognition andquality of life showing protective as well as therapeutic effects (Luet al., 2006). Since LH and follicle stimulating hormone levelscontinue to remain elevated in spite of cyclical estrogen and pro-gesterone therapy, modulation of gonadotropin (GnRH) levelsmight be more valuable than individual hormones themselves.GnRH agonist, leuprolide attenuated cognitive decline anddecreased Ab deposition in AD transgenic mice (Casadesus et al.,2006). In phase II trials of the compound, female patients demon-strated improvement in cognitive function (NCT00076440). A slowpellet release form of leuprolide underwent a phase III trial, but theresults are not available (NCT00231946). The role of GnRH antag-onists in AD is not known.

2.9.2. Lipid modifying therapyGiven the wealth of evidence linking hypercholesterolemia and

AD, statins have been extensively studied for their therapeuticbenets. Studies with statins highlight its pleiotropic effects anddose-dependent benecial effects on cognition, memory and neu-roprotection (Li et al., 2006). Statins modify the properties ofplasma membrane by diminishing cholesterol levels and modu-lating the secretase activities thus decreasing amyloidogenic APPprocessing (Buxbaum et al., 2002). They also seem to alter neuronalactivity by cholesterol independent effects such as modifying theprotein prenylation of different small GTPases altering their func-tion (Cordle et al., 2005). A possible role for the compounds incholinergic homeostasis is also suggested; simvastatin inhibitsAChE in rats (Cibickova et al., 2007) and prevent the blockade dueto AChE inhibitors at a7-nicotinic AChRs (Mozayan and Lee, 2007)thus enhancing cholinergic neurotransmission. Statins also protectprimary cortical neurons from glutamate toxicity (Bosel et al.,2005). It is seen that low dose statins prevent aberrant neuronalentry into mitosis (Sala et al., 2008), activate anti-apoptotic path-ways (Merla et al., 2007) and suppresses inammation (Cordl