Artificial neural networks allow the use of simultaneous measurements of Alzheimer disease markers for early detection of the disease

β-Amyloid promotes platelet activation and activated platelets act as bridge between risk factors and Alzheimer's disease

Alzheimer's disease (AD) is an evolving challenge that places an enormous burden on families and society. The presence of obvious brain β-amyloid (Aβ) deposition is a premise to diagnose AD, which induces the subsequent tau hyperphosphorylation and neurodegeneration. Platelets are the primary source of circulating amyloid precursor protein (APP). Upon activation, they can secrete significant amounts of Aβ into the blood, which can be actively transported to the brain across the blood-brain barrier and promote amyloid deposition. In this review, we summarized the changes in the platelet APP metabolic pathway in patients with AD and further comprehensively explored the targets and downstream events of Aβ-activated platelets. In addition, we attempted to clarify whether patients with AD are in a state of general platelet activation, with inconsistent results. Considering the increasingly evident bidirectional relationship between AD and vascular events, we speculate that the AD pathology alone seems to be insufficient to induce the general activation of platelets; however, the intervention of third-party factors, such as atherosclerosis, exposes the extracellular matrix and leads to platelet activation, further promoting AD progression. Therefore, we proposed a framework in which the relationship between platelets and AD is indirect and mediated by vascular factors. Therapies targeting platelets and interventions for vascular risk factors are likely to contribute to the prevention and treatment of AD.

Can platelet activation result in increased plasma Aβ levels and contribute to the pathogenesis of Alzheimer's disease?

One of the central lesions in the brain of subjects with Alzheimer's disease (AD) is represented by aggregates of β-amyloid (Aβ), a peptide of 40-42 amino acids derived from the amyloid precursor protein (APP). The reasons why Aβ accumulates in the brain of individuals with sporadic forms of AD are unknown. Platelets are the primary source of circulating APP and, upon activation, can secrete significant amounts of Aβ into the blood which can be actively transported to the brain across the blood-brain barrier and promote amyloid deposition. Increased platelet activity can stimulate platelet adhesion to endothelial cells, trigger the recruitment of leukocytes into the vascular wall and cause perivascular inflammation, which can spread inflammation in the brain. Neuroinflammation is fueled by activated microglial cells and reactive astrocytes that release neurotoxic cytokines and chemokines. Platelet activation is also associated with the progression of carotid artery disease resulting in an increased risk of cerebral hypoperfusion which may also contribute to the AD neurodegenerative process. Platelet activation may thus be a pathophysiological mechanism of AD and for the strong link between AD and cerebrovascular diseases. Interfering with platelet activation may represent a promising potential adjunct therapeutic approach for AD.

Platelet APP Processing: Is It a Tool to Explore the Pathophysiology of Alzheimer's Disease? A Systematic Review

The processing of the amyloid precursor protein (APP) is a critical event in the formation of amyloid plaques. Platelets contain most of the enzymatic machinery required for APP processing and correlates of intracerebral abnormalities have been demonstrated in platelets of patients with AD. The goal of the present paper was to analyze studies exploring platelet APP metabolism in Alzheimer's disease patients trying to assess potential reliable peripheral biomarkers, to offer new therapeutic solutions and to understand the pathophysiology of the AD. According to the PRISMA guidelines, we performed a systematic review through the PubMed database up to June 2020 with the search terms: "((((((APP) OR Amyloid Precursor Protein) OR AbetaPP) OR Beta Amyloid) OR Amyloid Beta) OR APP-processing) AND platelet". Thirty-two studies were included in this systematic review. The papers included are analytic observational studies, namely twenty-nine cross sectional studies and three longitudinal studies, specifically prospective cohort study. The studies converge in an almost unitary way in affirming that subjects with AD show changes in APP processing compared to healthy age-matched controls. However, the problem of the specificity and sensitivity of these biomarkers is still at issue and would deserve to be deepened in future studies.

The Effect of TCM-Induced HAMP on Key Enzymes in the Hydrolysis of AD Model Cells

Alzheimer's disease (AD) process is characterized classically by two hallmark pathologies: β-amyloid (Aβ) plaque deposition and neurofibrillary tangles of hyperphosphorylated tau. Aβ peptides play an important role in AD, but despite much effort the molecular mechanisms of how Aβ contributes to AD remain unclear. The present study evaluated the effects of the active components of Epimedium, Astragalus and Radix Puerariae induced HAMP on key enzymes in the hydrolysis of APP in HT22 cells. The active components of Epimedium, Astragalus and Radix Puerariae could effectively up-regulate the expression of HAMP, alleviate the iron overload in the brain tissues of mice, significantly improve the learning and memory ability of AD, down-regulate the expression of Aβ and reduce the deposition of SP in an APPswe/PS1_ΔE9 transgenic mouse model of AD. HAMP and Aβ_25-35 induced HT22 cells are used as AD cell models in this study to investigate the effect of the compound consisting of the effective components of Epimedium, Astragalus and Pueraria on the key enzymes in the hydrolysis of APP. After the administration of traditional Chinese medicine (TCM), the expression levels of ADAM10 and ADAM17 were increased while the expression level of BACE1 decreased. This indicates that TCM can promote the expression level of ADAM10 and ADAM17, inhibit the expression level of BACE1, thus further inhibiting the production of amyloid protein and reducing the production of Aβ and SP. Compared with RNAi group, the expression level of ADAM10 and ADAM17 in Aβ + RNAi group was decreased while the expression level of BACE1 increased. Compared with the Aβ + RNAi group the expression level of ADAM10 and ADAM17 in the Aβ + RNAi + TCM group was increased while the expression level of BACE1 was decreased. The present study indicated the effects of the active components of Epimedium, Astragalus and Radix Puerariae may alleviate AD by up-regulating the expression of HAMP, thus reducing brain iron overload, promoting the expression of ADAM10 and ADAM17, inhibiting the expression of BACE1, and reducing the deposition of Aβ.

Mathematical Models to Shed Light on Amyloid-Beta and Tau Protein Dependent Pathologies in Alzheimer's Disease

The number of patients suffering from dementia due to Alzheimer's disease (AD) is constantly rising worldwide. This has accordingly resulted in huge burdens on the health systems and involved families. Lack of profound understanding of neural networking in normal brain and their interruption in AD makes the treatment of this neurodegenerative multifaceted disease a challenging issue. In recent years, mathematical and computational methods have paved the way towards a better understanding of the brain functional connectivity. Thus, much attention has been paid to this matter from both basic science researchers and clinicians with an interdisciplinary approach to determine what is not functioning properly in AD patients and how this malfunctioning can be addressed. In this review, a number of AD-related articles and well-studied pathophysiologic topics (e.g., amyloid-beta, neurofibrillary tangles, Ca²⁺ dysregulation, and synaptic plasticity alterations) has been literally surveyed from a computational and systems biology point of view. The neural networks were discussed from biological and mathematical point of views and their alterations in recent findings were further highlighted. Application of the graph theoretical analysis in the brain imaging was reviewed, depicting the relations between brain structure and function, without diving into mathematical details. Moreover, differential rate equations were briefly articulated, emphasizing the potential use of these equations in simplifying complex processes in relevance to pathologies of AD. Comprehensive insights were given into the AD progression from neural networks perspective, which may lead us towards potential strategies for early diagnosis and effective treatment of AD.

Deep learning-based classification of multi-categorical Alzheimer's disease data

It is urgent to find the appropriate technology for the early detection of Alzheimer's disease (AD) due to the unknown AD etiopathologies that bring about serious social problems. Early detection of mild cognitive impairment (MCI) has pivotal importance in delaying or preventing the AD onset. Herein, we utilize deep learning (DL) techniques for the purpose of multiclass classification between normal control, MCI, and AD subjects. We used multi-categorical data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) including brain imaging measurements, cognitive test results, cerebrospinal fluid measures, ApoE4 status, and age. We achieved an overall accuracy of 87.197% for our artificial neural network classifier and a similar overall accuracy of 88.275% for our 1D convolutional neural network classifier. We conclude that DL-based techniques are powerful tools in analyzing ADNI data although further method refinements are needed.

Protein Expression of BACE1 is Downregulated by Donepezil in Alzheimer's Disease Platelets

BACKGROUND

Abnormal amyloid-β protein precursor (AβPP) metabolism is a key feature of Alzheimer's disease (AD). Platelets contain most of the enzymatic machinery required for AβPP processing, and correlates of intracerebral abnormalities have been demonstrated in platelets of patients with AD. Thus, AβPP-related molecules in platelets may be regarded as peripheral markers of AD.

OBJECTIVE

We sought to determine the protein expression of the AβPP secretases (ADAM10, BACE1, and PSEN1) and AβPP ratio in platelets of patients with mild or moderate AD compared to healthy controls. We further determined whether the protein expression of these markers might be modified by chronic treatment with donepezil.

METHODS

Platelet samples were obtained from patients and controls at baseline and after 3 and 6 months of continuous treatment with therapeutic doses of donepezil. The protein expression of platelet markers was determined by western blotting.

RESULTS

AD patients had a significant decrease in AβPP ratio, ADAM10, and PSEN1 compared to controls at baseline, but these differences were not modified by the treatment. Nonetheless, a significant reduction in the protein expression of BACE1 was observed in patients treated with donepezil for 6 months.

CONCLUSION

Our results corroborate previous findings from our group and others of decreased AβPP ratio and protein expression of ADAM10 in AD. We further show that PSEN1 is decreased in AD platelets, and that the protein expression of BACE1 is downregulated by chronic treatment with donepezil. This effect may be interpreted as evidence of disease modification.

On the early diagnosis of Alzheimer's Disease from multimodal signals: A survey

INTRODUCTION

The number of Alzheimer's Disease (AD) patients is increasing with increased life expectancy and 115.4 million people are expected to be affected in 2050. Unfortunately, AD is commonly diagnosed too late, when irreversible damages have been caused in the patient.

OBJECTIVE

An automatic, continuous and unobtrusive early AD detection method would be required to improve patients' life quality and avoid big healthcare costs. Thus, the objective of this survey is to review the multimodal signals that could be used in the development of such a system, emphasizing on the accuracy that they have shown up to date for AD detection. Some useful tools and specific issues towards this goal will also have to be reviewed.

METHODS

An extensive literature review was performed following a specific search strategy, inclusion criteria, data extraction and quality assessment in the Inspec, Compendex and PubMed databases.

RESULTS

This work reviews the extensive list of psychological, physiological, behavioural and cognitive measurements that could be used for AD detection. The most promising measurements seem to be magnetic resonance imaging (MRI) for AD vs control (CTL) discrimination with an 98.95% accuracy, while electroencephalogram (EEG) shows the best results for mild cognitive impairment (MCI) vs CTL (97.88%) and MCI vs AD distinction (94.05%). Available physiological and behavioural AD datasets are listed, as well as medical imaging analysis steps and neuroimaging processing toolboxes. Some issues such as "label noise" and multi-site data are discussed.

CONCLUSIONS

The development of an unobtrusive and transparent AD detection system should be based on a multimodal system in order to take full advantage of all kinds of symptoms, detect even the smallest changes and combine them, so as to detect AD as early as possible. Such a multimodal system might probably be based on physiological monitoring of MRI or EEG, as well as behavioural measurements like the ones proposed along the article. The mentioned AD datasets and image processing toolboxes are available for their use towards this goal. Issues like "label noise" and multi-site neuroimaging incompatibilities may also have to be overcome, but methods for this purpose are already available.

Platelet biomarkers in Alzheimer’s disease

The search for diagnostic and prognostic markers in Alzheimer’s disease (AD) has been an area of active research in the last decades. Biochemical markers are correlates of intracerebral changes that can be identified in biological fluids, namely: peripheral blood (total blood, red and white blood cells, platelets, plasma and serum), saliva, urine and cerebrospinal fluid. An important feature of a biomarker is that it can be measured objectively and evaluated as (1) an indicator of disease mechanisms (markers of core pathogenic processes or the expression of downstream effects of these processes), or (2) biochemical responses to pharmacological or therapeutic intervention, which can be indicative of disease modification. Platelets have been used in neuropharmacological models since the mid-fifties, as they share several homeostatic functions with neurons, such as accumulation and release of neurotransmitters, responsiveness to variations in calcium concentration, and expression of membrane-bound compounds. Recent studies have shown that platelets also express several components related to the pathogenesis of AD, in particular to the amyloid cascade and the regulation of oxidative stress: thus they can be used in the search for biomarkers of the disease process. For instance, platelets are the most important source of circulating forms of the amyloid precursor protein and other important proteins such as Tau and glycogen synthase kinase-3B. Moreover, platelets express enzymes involved in membrane homeostasis (e.g., phospholipase A₂), and markers of the inflammatory process and oxidative stress. In this review we summarize the available literature and discuss evidence concerning the potential use of platelet markers in AD.

Platelets and Alzheimer’s disease: Potential of APP as a biomarker

Platelets are the first peripheral source of amyloid precursor protein (APP). They possess the proteolytic machinery to produce Aβ and fragments similar to those produced in neurons, and thus offer an ex-vivo model to study APP processing and changes associated with Alzheimer’s disease (AD). Platelet process APP mostly through the α-secretase pathway to release soluble APP (sAPP). They produce small amounts of Aβ, predominantly Aβ₄₀ over Aβ₄₂. sAPP and Aβ are stored in α-granules and are released upon platelet activation by thrombin and collagen, and agents inducing platelet degranulation. A small proportion of full-length APP is present at the platelet surface and this increases by 3-fold upon platelet activation. Immunoblotting of platelet lysates detects APP as isoforms of 130 kDa and 106-110 kDa. The ratio of these of APP isoforms is significantly lower in patients with AD and mild cognitive impairment (MCI) than in healthy controls. This ratio follows a decrease that parallels cognitive decline and can predict conversion from MCI to AD. Alterations in the levels of α-secretase ADAM10 and in the enzymatic activities of α- and β-secretase observed in platelets of patients with AD are consistent with increased processing through the amyloidogenic pathway. β-APP cleaving enzyme activity is increased by 24% in platelet membranes of patients with MCI and by 17% in those with AD. Reports of changes in platelet APP expression with MCI and AD have been promising so far and merit further investigation as the search for blood biomarkers in AD, in particular at the prodromal stage, remains a priority and a challenge.

Reduced platelet amyloid precursor protein ratio (APP ratio) predicts conversion from mild cognitive impairment to Alzheimer's disease

Studies have shown that platelet APP ratio (representing the percentage of 120-130 kDa to 110 kDa isoforms of the amyloid precursor protein) is reduced in patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD). In the present study, we sought to determine if baseline APP ratio predicts the conversion from MCI to AD dementia after 4 years of longitudinal assessment. Fifty-five older adults with varying degrees of cognitive impairment (34 with MCI and 21 with AD) were assessed at baseline and after 4 years. MCI patients were re-classified according to the conversion status upon follow-up: 25 individuals retained the diagnostic status of MCI and were considered as stable cases (MCI-MCI); conversely, in nine cases the diagnosis of dementia due to AD was ascertained. The APP ratio (APPr) was determined by the Western blot method in samples of platelets collected at baseline. We found a significant reduction of APPr in MCI patients who converted to dementia upon follow-up. These individuals had baseline APPr values similar to those of demented AD patients. The overall accuracy of APPr to identify subjects with MCI who will progress to AD was 0.74 ± 0.10, p = 0.05. The cut-off of 1.12 yielded a sensitivity of 75 % and a specificity of 75 %. Platelet APPr may be a surrogate marker of the disease process in AD, with potential implications for the assessment of abnormalities in the APP metabolism in patients with and at risk for dementia. However, diagnostic accuracy was relatively low. Therefore, studies in larger samples are needed to determine whether APPr may warrant its use as a biomarker to support the early diagnosis of AD.

Risk modeling strategies for pharmacogenetic studies

Pharmacogenetic risk models offer great promise as treatment decision tools; however, their uptake in routine clinical practice is so far disappointing, not least due to the lack of evidence of their benefit in randomized controlled trials and other types of studies. Prior to conducting such a study, it is imperative that the model’s predictive capability is first of all proven, and that it is shown to be superior to the most appropriate alternative model. When demonstrating predictive capability, clinical implications of applying the model should be a key consideration, and the Decision Curve Analysis method takes this into account for binary outcomes. Furthermore, when comparing a novel model to the best alternative, methods such as Net Reclassification Improvement or Integrated Discrimination Difference are recommended as they provide a more reliable comparison than other methods currently in common use. Where outcome is continuous, such as therapeutic dose, assessing a model’s performance is generally more intuitive and straightforward since the aim is to achieve a predicted dose as close as possible to the true therapeutic dose.

Machine learning methods for anticipating the psychological distress in patients with Alzheimer's disease

Recent studies proved that psychological distress is an accelerator of Alzheimer disease (AD). However, the factors that affect the psychological distress of AD patients are still unknown. The aim of this study was to predict the incidence and identify the risk factors of psychological distress in AD patients. Artificial neural networks and Machine learning models were used to predict the incidence of psychological distress in AD patients. Linear regression and decision tree models were used to identify the factors of psychological distress in AD patients. Among all models for predicting the incidence of psychological distress in AD patients, the artificial neural networks with 8 hidden neurons achieved the highest predictive accuracy of 81.92%. In the five machine learning models, the ADTree algorithm made the highest Predictive Accuracy of 77.94%. As for risk factor analysis, the Linear Regression and Decision Tree models reported similar sets of variables that affect the psychological distress of AD patients. Three variables were reported by Linear Regression to be in negative correlation with psychological distress: the use of professional care service, caregiver consuming cigarette, and caregiver consuming alcohol. The incidence of psychological distress in AD patients can be predicted by artificial neural networks with an accuracy of 81.92%. There are four main risk factors for psychological distress of AD patients: "Caregiver experiencing psychological distress", "Caregiver suffering from chronic disease or cancer", "Care recipient's health status being serious or getting worse", and "Lack of professional care service". These findings are potentially helpful for the prediction, prevention and intervention of psychological distress in AD patients.

Predicting Alzheimer dementia in mild cognitive impairment patients. Are biomarkers useful?

A correct clinical diagnosis in the early stage of Alzheimer disease is not only of importance given the current available treatment with acetylcholine esterase inhibitors, but would be the basis for disease-modifying therapy slowing down or arresting the degenerative process. Moreover, in the last years, several efforts have been made to determine if a patient with mild cognitive impairment has incipient Alzheimer disease, i.e. will progress to Alzheimer disease with dementia, or have a benign form of mild cognitive impairment. In this review, the recent published reports regarding progress in early and preclinical Alzheimer disease diagnosis are discussed and the role of peripheral and cerebrospinal fluid biomarkers highlighted. Approaches combining panels of different biomarkers show promise for discovering profiles that are characteristic of Alzheimer disease, even in the pre-symptomatic stage. More work is needed but available novel perspectives offered by recent introduced technologies shed some lights in identifying incipient Alzheimer disease in mild cognitive impairment subjects.

SNP selection at the NAT2 locus for an accurate prediction of the acetylation phenotype

PURPOSE

Genetic polymorphisms in the N-acetyltransferase 2 gene determine the individual acetylator status, which influences both the toxicity and efficacy profile of acetylated drugs. Determination of an individual's acetylation phenotype prior to initiation of therapy, through DNA-based tests, should permit to improve therapy response and reduce adverse events. However, due to extensive linkage disequilibrium between markers within NAT2, the genotyping of closely spaced markers yields highly redundant data: testing them all is expensive and often unnecessary. The objective of this study is to establish the optimal strategy to define, in the genetic context of a given ethnic group, the most informative set of single-nucleotide polymorphisms that best enables accurate prediction of acetylation phenotype.

METHODS

Three classification methods have been investigated (classification trees, artificial neural networks and multifactor dimensionality reduction method) in order to find the optimal set of single-nucleotide polymorphisms enabling the most efficient classification of individuals in rapid and slow acetylators.

RESULTS

Our results show that, in almost all population samples, only one or two single-nucleotide polymorphisms would be enough to obtain a good predictive capacity with no or only a modest reduction in power relative to direct assays of all common markers. In contrast, in Black African populations, where lower levels of linkage disequilibrium are observed at NAT2, a larger number of single-nucleotide polymorphisms are required to predict acetylation phenotype.

CONCLUSION

The results of this study will be helpful for the design of time- and cost-effective pharmacogenetic tests (adapted to specific populations) that could be used as routine tools in clinical practice.

Data-Mining Methods as Useful Tools for Predicting Individual Drug Response: Application to CYP2D6 Data

OBJECTIVES

Selecting a maximally informative subset of polymorphisms to predict a clinical outcome, such as drug response, requires appropriate search methods due to the increased dimensionality associated with looking at multiple genotypes. In this study, we investigated the ability of several pattern recognition methods to identify the most informative markers in the CYP2D6 gene for the prediction of CYP2D6 metabolizer status.

METHODS

Four data-mining tools were explored: decision trees, random forests, artificial neural networks, and the multifactor dimensionality reduction (MDR) method. Marker selection was performed separately in eight population samples of different ethnic origin to evaluate to what extent the most informative markers differ across ethnic groups.

RESULTS

Our results show that the number of polymorphisms required to predict CYP2D6 metabolic phenotype with a high accuracy can be dramatically reduced owing to the strong haplotype block structure observed at CYP2D6. MDR and neural networks provided nearly identical results and performed the best.

CONCLUSION

Data-mining methods, such as MDR and neural networks, appear as promising tools to improve the efficiency of genotyping tests in pharmacogenetics with the ultimate goal of pre-screening patients for individual therapy selection with minimum genotyping effort.