Butyrylcholinesterase K and Apolipoprotein E-ɛ4 Reduce the Age of Onset of Alzheimer’s Disease,Accelerate Cognitive Decline,and Modulate Donepezil Response in Mild Cognitively Impaired Subjects

INTRODUCTION

Alzheimer’s disease (AD) is an adult-onset chronic neurodegenerative disorder that occurs predominantly later in life. It is the most common cause of dementia and represents the fourth most common cause of death in the developed world [1]. It is estimated that more than 35 million people worldwide are afflicted by the disease [2] among whom 95% of cases are sporadic, with the majority of the remaining cases showing an autosomal dominant inheritance [3]. For identical twins, the probability that the co-twin of a demented twin will also have dementia was shown to vary from 60–72% [4]. Recently, subjects diagnosed with mild cognitive impairment (MCI) [5], also referred to as the prodromal AD [6], received particular attention from the scientific community. Previous research indicated that nearly 80% of amnestic MCI subjects (aMCI), the dominant MCI subtype with a primary memory component [7], will progress to AD within the course of a six year trial [8], at an annual rate of 10–15% [5].

The disease is also characterized by selective hippocampal and cortical neuron degenerations. In the cholinergic system, alterations observed in AD are characterized by cortical deafferentation from cholinergic nuclei and by cholinergic synaptic modifications [9]. These modifications are associated with a significant reduction (10–20%) in acetylcholinesterase (AChE) activity and a similar increase of butyrylcholinesterase (BuChE) activity (about 120%), and are especially found in the hippocampus; suggesting a relationship with the first symptoms of AD (loss of episodic memory). This is in agreement with the reported observation that increased levels of BuChE in temporal cortex are linked to cognitive decline of demented patients [10]. However, the increased levels of BuChE in AD brain and greater cognitive decline are in contrast with more recent data on CSF BuChE levels [11].

Apart from apolipoprotein E4 (APOE-ɛ4) genetic variant, for which the association with sporadic AD [12, 13] has been replicated countless times byindependent studies worldwide, the butyrylcholinesterase gene (BCHE) has become one of the few AD susceptibility gene candidates supported by meta-analyses of association studies [14] and amyloid pathology-related genome wide association studies [15]. The BCHE-K^* variant is the most frequently reported [16] among more than 40 polymorphisms in the coding region of the BCHE gene [17]. Interestingly, this coding variant is associated with a 30% decreased enzymatic activity relative to the wild type alleles. BCHE-K^* shows reduced ability to inhibit Aβ fibril formation in vitro [18, 19], while homozygosity for BCHE-K^* was found to increase the risk for development of neurofibrillary pathology but not amyloid deposits in cognitively-intact individuals under the age of 45 years [20]. The combined action of the APOE-ɛ4 and BCHE-K^* variants was shown to markedly affect cortical thickness in AD [21]. Studies on the association between BCHE-K^* and the risk of developing AD have, however, yielded conflicting results. On one hand, several studies showed positive association between BCHE-K^* and AD risk [22–26] and pathology in late-onset AD [20, 28], particularly so in APOE-ɛ4 carriers [29], whereas some reports could not verify this association [30].

AD pathology is also known to significantly alter basal forebrain cholinergic projections. In the mammalian brain, BuChE and AChE enzymes co-regulate levels of synaptic and extracellular acetylcholine [31]. In MCI and early AD, damage to cholinergic systems are mostly attributed to cholinergic synaptic modifications [32], while extensive deafferentation from cholinergic nuclei characterize more severe stages of the disease [9]. Cholinergic synapses are particularly vulnerable to extracellular amyloid-beta peptide (Aβ) [33]. Particularly in individuals <75 years of age, APOE-ɛ4 means more rapid loss of cholinergic neurotransmission, cognitive decline, hippocampal atrophy, and accelerated loss of myelin in the medial and temporal lobes prior to the appearance of clinical dementia [34, 35]. Elevated BuChE activity found in AD pathology [36] was shown to lessen amyloid fibril formation [37]. Reduced BuChE enzymatic activity in BCHE-K^* carriers was proposed to accelerate AD pathology [38] possibly through reduced anti-Aβ aggregating activity [37].

The 3-4 years randomized, placebo-controlled InDDEx study, that included more than 1,000 aMCI subjects, did not find that the BCHE allele status alone predicted progression to AD, but in retrospective analyses, APOE-ɛ4 and BCHE-K^* were seen to significantly interact such that carriers of both alleles had the highest rate of progression to dementia [39]. This differential progression rate to AD amongst genotype groups was further increased when the patient population <75 years was evaluated with 39% of carriers of both alleles progressing to AD over 3-4 years versus just 4% in aMCI subjects without either allele [40]. InDDEx data showed that carrying both APOE-ɛ4 and BCHE-K^* enhanced cognitive decline and hippocampal atrophy in a subgroup of 231 placebo-treated MCI subjects [39]. In contrast, BCHE-K^* carriers who did not carry the APOE-ɛ4 allele exhibited the least cognitive decline and progression to AD, along with the smallest reductions in hippocampal volume [39]. The authors suggested that BCHE-K^* may exert a potentiated effect when combined to the APOE-ɛ4 in MCI subjects, presumably due to accelerate damage to basal forebrain cholinergic pathways. Moreover, this study in aMCI subjects, similar to patients with mild AD and mild to moderate Parkinson’s disease dementia, showed that only MCI subjects who carried both APOE-ɛ4 and BCHE-K^* alleles exhibited greater cognitive responses to rivastigmine than other genotype groups, suggesting accelerated cholinergic neurotransmission deficits in carriers of both of these alleles [38].

Although carriers of BCHE-K^* and/or APOE-ɛ4 may have the most rapid cognitive decline and in the MCI stage of AD, they have the slowest progression by the time that they reach the moderate stage of AD [41, 42]. Therefore, the impact of these alleles on disease progression and response to cholinomimetic treatment appears to be AD stage-dependent. A 2-year retrospective, drug-efficacy analysis that contrasted rivastigmine, a sustained dual inhibitor of AChE and BuChE, and donepezil, a monospecific AChE inhibitor, in subjects with moderate to moderately severe AD demonstrated slightly better responses to rivastigmine in younger subjects [43]. The efficacy of these two drugs, however, did not differ in older subjects [44]. A more recent study contrasted the efficacy of these cholinergic drugs according to BCHE genotypes (i.e., wild-type BCHE and BCHE-K^*) in AD subjects who were under the age of 75. It was found that while wild-type BCHE carriers generally showed greater treatment responses to rivastigmine (an inhibitor of AChE and BuChE) over two years when compared to subjects treated with donepezil (a selective inhibitor of AChE), no statistically significant treatment differences were observed in BCHE-K^* carriers on any efficacy measure [45].

Here, 367 patients of an eastern Canadian population isolate of AD subjects were genotyped for K^* variant of BCHE and ɛ4 polymorphism of APOE. We then analyzed the age of onset of AD dementia as a function of the genotypes. Independently, aMCI subjects enrolled in the original three-year follow-up, double-blind, placebo-controlled vitamin E/donepezil Alzheimer’s Disease Cooperative Study (ADCS) of Petersen and colleagues [46] were investigated for cognitive decline (ADAS-Cog score) and progression to AD as a function of BCHE and APOE genotypes. In addition, responses to donepezil on these parameters were investigated. Finally, a cohort of autopsy-confirmed AD and control subjects obtained from the Douglas Institute Bell-Brain Bank underwent APOE and BCHE genotyping and their respective hippocampal (n = 22) and cortical (n = 37) choline acetyltransferase activities analyzed and stratified according to their genotypes.

MATERIALS AND METHODS

RESULTS

DISCUSSION

In the QFP cohort, a positive correlation was observed between APOE-ɛ4 carriers and an earlier age of onset of the disease; consistent with previous studies [54–57] whereas subjects carrying both APOE-ɛ4 and BCHE-K^* variants display a much earlier onset than their ɛ4- and K^*- negative counterparts (72.5 versus 75.9 years old). While initial studies examining associations between BCHE-K^* and AD risk led to inconsistent results [22–25, 58–60], a recent meta-analysis [59] of all datasets shows significant association between BCHE-K^* variant and AD risk, especially in Asians subjects. Consistent with our results, the authors also showed that BCHE-K^* variants was significantly associated with late-onset AD and that APOE-ɛ4/BCHE-K^* carriers had higher AD risk than K^*-negative carriers [59]. Recently, greater cognitive decline was reported in APOE-ɛ4/BCHE-K^* carriers when compared to mutant negative subjects [61]. BCHE-K^* and APOE-ɛ4 alleles were shown to interact synergistically in mediating cognitive decline, hippocampal atrophy as well as progression from MCI to AD [39], consistent with our results indicating earlier conversion in dual ɛ4/K^* carriers (–3 years).

Using genotype profiling of aMCI subjects from the Vitamin E/Donepezil ADCS study [46], analyses reveal that aMCI subjects with BCHE-K^* allele showed significantly greater cognitive response to a daily dose of donepezil three years after MCI diagnosis relative to K^*-positive carrier subjects who were administered a placebo. This beneficial pharmacogenetic response was disproportionately compelling in women who, on average, exhibited ADAS-Cog score improvements at 36 months relative to baseline. Interestingly, this contrasts with a numerically more compelling response in male relative to female aMCI carriers of K^* and ɛ4 alleles over 3-4 years in the InDDex study [42]. Caution is warranted on the potential to overinterpret these gender effects in such retrospective analyses. On the other hand, it should be emphasized that rivastigmine acts via inhibition of both BuChE and AchE activities whereas donepezil is specific for AchE. Both studies suggest that a cognitive response to cholinesterase inhibitors (ChE-Is) in aMCI is only seen in carriers of both K^* and ɛ4.

enlargethispage *2ptFailed attempts to find a positive response to ChE-Is treatments in unstratified MCI populations were recently attributed to the fact that there were too few MCI subjects with significant damage to cholinergic pathways to significantly benefit from ChE-Is treatment. It was suggested that progression of disease to either mild or moderate stages of AD, is accompanied with more marked cholinergic system deficits, and consequently enhanced cognitive response to ChE-Is treatment [62]. Moreover, consistent with findings from the InDDEx study [63], BCHE-K^* carriers, who are known to feature endogenous reductions of BCHE activity, more so in the presence of APOE-ɛ4 allele [64, 65], did not exhibit lower ADAS-Cog performance relative to wild-type BCHE carriers in our MCI subjects. However, findings from the present study suggests that endogenous reductions of BuChE hydrolytic activity in BCHE-K^* carriers [64] together with donepezil’s selective inhibition of AChE synergistically attenuated cognitive decline in MCI subjects tested at the 36 month time point. Alternatively, it is conceivable, based on the damage in cholinergic system observed in K^*/ɛ4 carriers in our postmortem study in AD, that the severe cholinergic deficit in this subpopulation of MCI subjects is actually driving the response to increases to synaptic acetylcholine. The pharmacogenetic effect most markedly benefited MCI women who showed greater ADAS-Cog score improvements after a three years treatment; a finding that dramatically contrasts with the 5-point decline in placebo-treated BCHE-K^* women. Although not systematically controlled for in this study, it appears plausible that lowered estrogen levels in post-menopausal women not under estrogen replacement therapy, which are known to negatively impact basal forebrain cholinergic neurons function [66], could further cholinergic systems deficits in BCHE-K^* MCI women, which would in turn predict improved cognitive response to donepezil treatment. Future studies should directly assess how estrogen levels modulate the underlying action mechanisms of the pharmacogenetic effect found herein.

Known to accelerate the progression of both familial and sporadic AD [55, 67], to increase the likelihood of cognitive impairments in clinically normal 50 + years old over time [68] and to precipitate progression to AD among MCI subjects [8], the APOE-ɛ4 allele was studied along with the BCHE-K^* variant to investigate response to cholinesterase inhibitors in both AD and MCI populations. In contrast with BCHE-K^*, the APOE-ɛ4 allele narrowly failed to significantly associate with cognitive response to donepezil in MCI subjects over the course of the 3-year trial. This is at variance with prior studies conducted with AD subjects who carried the APOE-ɛ4 allele that showed positive response to 12–16 months donepezil therapy [69] as well as after 48 weeks of treatment [70]. Whereas APOE-ɛ4 alone was reported to be a major predictor of progression to AD in MCI subjects from the ADCS study [46], further interaction between BCHE-K^* and APOE-ɛ4 computed in these same MCI subjects did not appear to synergistically improve cognitive response to donepezil. This is at odds with a previous MCI study that revealed positive BCHE-K^* and APOE-ɛ4 synergistic effects on cognitive response to rivastigmine treatment over the course of 3-4 years [39]. In sharp contrast, stratification for the APOE-ɛ4 allele among BCHE-K^* MCI subjects from this study revealed equivalent tendencies across APOE subgroups for a favorable cognitive response to donepezil relative to a placebo. In a series of analyses, we examined the effect of K^* on MCI progression to AD using Kaplan-Meier analyses. In contrast to the APOE-ɛ4 allele [71], the K^* variant does not affect the conversion rate in this MCI cohort, with or without ɛ4 stratification (not shown).

Consistent with previous findings [21], for AD subjects carrying the APOE-ɛ4 genotype a significant reduction of ChAT activity was observed. The observed genotype-dependent reduction in ChAT activity in autopsied AD subjects provides with a possible explanation for the early and marked response of MCI subjects toward cholinomimetics: carriers of APOE-ɛ4 and/or BCHE-K^* variants display a marked genotype-driven deterioration of the cholinergic system in the same subjects known to display early amyloid deposition (APOE-ɛ4 and BCHE-K^* carriers) and faster progression rate from MCI to AD [39].

These novel results reiterate the necessity to prospectively control for pharmacogenomic effects in, and suggest enrichment strategies for, secondary prevention trials involving prodromal AD subjects. The research focus in the AD field now is turning to prevention, and a number of studies are underway to test the effectiveness of various therapies in people without symptoms or who have only slight cognitive deficits. While the initial focus of these trials was to recruit symptom-free volunteers whose PET scans show abnormally high levels of Aβ deposition in the brain (the anti-amyloid treatment in asymptomatic Alzheimer’s disease trial (A4)) or, who are born carriers of gene mutations in genes causally associated with AD and amyloid metabolism (Alzheimer’s Prevention Initiative) and the dominantly-inherited Alzheimer trial unit (DIAN-TU)), new initiatives are now stratifying subjects on the basis of parental history (PREVENT-AD) or, APOE-ɛ4 heterozygosity (Tomorrow trial). In light of the above findings, it would make sense to include the BCHE gene variant as part of the statistical analysis plan either as a covariant, or better, as a stratifying variable which should be properly balanced in both the placebo and treatment arms of the trials.

In the event of the development of novel replace-ment cholinomimetic drugs to be tested in MCI subjects, we strongly believe that both APOE and BCHE genotypes stratification should be implemented to identify the potential responder sub-group(s) most likely to benefit from the experimental intervention.