Metabolomics is the comprehensive study of the metabolome, the repertoire of biochemicals (or small molecules) present in cells, tissues, and body fluids. The study of metabolism at the global or "-omics" level is a rapidly growing field that has the potential to have a profound impact upon medical practice. At the center of metabolomics, is the concept that a person's metabolic state provides a close representation of that individual's overall health status. This metabolic state reflects what has been encoded by the genome, and modified by diet, environmental factors, and the gut microbiome. The metabolic profile provides a quantifiable readout of biochemical state from normal physiology to diverse pathophysiologies in a manner that is often not obvious from gene expression analyses. Today, clinicians capture only a very small part of the information contained in the metabolome, as they routinely measure only a narrow set of blood chemistry analytes to assess health and disease states. Examples include measuring glucose to monitor diabetes, measuring cholesterol and high density lipoprotein/low density lipoprotein ratio to assess cardiovascular health, BUN and creatinine for renal disorders, and measuring a panel of metabolites to diagnose potential inborn errors of metabolism in neonates. We anticipate that the narrow range of chemical analyses in current use by the medical community today will be replaced in the future by analyses that reveal a far more comprehensive metabolic signature. This signature is expected to describe global biochemical aberrations that reflect patterns of variance in states of wellness, more accurately describe specific diseases and their progression, and greatly aid in differential diagnosis. Such future metabolic signatures will: (1) provide predictive, prognostic, diagnostic, and surrogate markers of diverse disease states; (2) inform on underlying molecular mechanisms of diseases; (3) allow for sub-classification of diseases, and stratification of patients based on metabolic pathways impacted; (4) reveal biomarkers for drug response phenotypes, providing an effective means to predict variation in a subject's response to treatment (pharmacometabolomics); (5) define a metabotype for each specific genotype, offering a functional read-out for genetic variants: (6) provide a means to monitor response and recurrence of diseases, such as cancers: (7) describe the molecular landscape in human performance applications and extreme environments. Importantly, sophisticated metabolomic analytical platforms and informatics tools have recently been developed that make it possible to measure thousands of metabolites in blood, other body fluids, and tissues. Such tools also enable more robust analysis of response to treatment. New insights have been gained about mechanisms of diseases, including neuropsychiatric disorders, cardiovascular disease, cancers, diabetes and a range of pathologies. A series of ground breaking studies supported by National Institute of Health (NIH) through the Pharmacometabolomics Research Network and its partnership with the Pharmacogenomics Research Network illustrate how a patient's metabotype at baseline, prior to treatment, during treatment, and post-treatment, can inform about treatment outcomes and variations in responsiveness to drugs (e.g., statins, antidepressants, antihypertensives and antiplatelet therapies). These studies along with several others also exemplify how metabolomics data can complement and inform genetic data in defining ethnic, sex, and gender basis for variation in responses to treatment, which illustrates how pharmacometabolomics and pharmacogenomics are complementary and powerful tools for precision medicine. Our metabolomics community believes that inclusion of metabolomics data in precision medicine initiatives is timely and will provide an extremely valuable layer of data that compliments and informs other data obtained by these important initiatives. Our Metabolomics Society, through its "Precision Medicine and Pharmacometabolomics Task Group", with input from our metabolomics community at large, has developed this White Paper where we discuss the value and approaches for including metabolomics data in large precision medicine initiatives. This White Paper offers recommendations for the selection of state of-the-art metabolomics platforms and approaches that offer the widest biochemical coverage, considers critical sample collection and preservation, as well as standardization of measurements, among other important topics. We anticipate that our metabolomics community will have representation in large precision medicine initiatives to provide input with regard to sample acquisition/preservation, selection of optimal omics technologies, and key issues regarding data collection, interpretation, and dissemination. We strongly recommend the collection and biobanking of samples for precision medicine initiatives that will take into consideration needs for large-scale metabolic phenotyping studies.
Abstract Background Apolipoprotein E (APOE) plays a critical metabolic role by facilitating the binding of lipid complexes to cell surface receptors, providing tissues with energy substrates. The APOE‐ε4 ( APOE4) polymorphism is the predominant genetic risk factor for late‐onset Alzheimer’s Disease (LOAD), while APOE3 is considered risk‐neutral. Female biological sex also nearly doubles LOAD risk. This study utilizes the JAX humanized APOE knock‐in mouse model to investigate the contributions of chromosomal sex and APOE genotype to plasma metabolic profiles, along with potential correlates to cognitive performance in a novel objection recognition (NOR) task. Method Male and female mice aged 23‐25 months old (37M/32F) with humanized APOE3/3, APOE3/4 , and APOE4/4 genotypes underwent blood plasma extraction and subsequent metabolic profiling via mass spectrometry using the Biocrates MxP® Quant 500 platform. A subset of animals (34M/28F) additionally performed an NOR paradigm prior to metabolic profiling. Two‐way ANOVA and post‐hoc t‐tests were utilized to determine data significance. Result Metabolomic analyses indicated that male mice exhibited relatively higher levels of glucogenic amino acids, including glycine, asparagine, and histidine. In contrast, female mice had higher levels of TMAO and ADMA along with lower levels of phosphatidylcholines, cholesteryl esters (CE), and triglycerides. In males, APOE3/4 and APOE4/4 genotypes were associated with lowered CE levels. APOE3/4 female mice had the lowest CE plasma concentrations. Female APOE4/4 mice had increased concentrations of several diglyceride and triglyceride species, relative to APOE3/3 and APOE3/4 female mice. Behavioral correlates indicated that triglyceride levels negatively correlated with locomotion throughout the NOR paradigm. Memory performance, as determined by the discrimination index, positively correlated with plasma concentrations of glycine and negatively correlated with hydroxy acylcarnitine levels. Conclusion In mice aged to resemble a ∼70‐year‐old human population, these results suggest that biological sex affects peripheral metabolic profiles greater than APOE genotype. Male metabolism was characterized by elevated glucogenic amino acid levels; female metabolic profiles demonstrated increased TMAO and lower phosphatidylcholines and triglycerides, suggesting differences in lipid metabolism. APOE3/4 genotype was associated with lowered CE in both sexes while APOE4/4 females had higher triglycerides, indicating potential sex‐genotype interactions upon metabolism that may confer elevated LOAD risk and influence behavior.
Importance Increasing evidence suggests an important role of liver function in the pathophysiology of Alzheimer disease (AD). The liver is a major metabolic hub; therefore, investigating the association of liver function with AD, cognition, neuroimaging, and CSF biomarkers would improve the understanding of the role of metabolic dysfunction in AD.
Objective To examine whether liver function markers are associated with cognitive dysfunction and the “A/T/N” (amyloid, tau, and neurodegeneration) biomarkers for AD.
Design, Setting, and Participants In this cohort study, serum-based liver function markers were measured from September 1, 2005, to August 31, 2013, in 1581 AD Neuroimaging Initiative participants along with cognitive measures, cerebrospinal fluid (CSF) biomarkers, brain atrophy, brain glucose metabolism, and amyloid-β accumulation. Associations of liver function markers with AD-associated clinical and A/T/N biomarkers were assessed using generalized linear models adjusted for confounding variables and multiple comparisons. Statistical analysis was performed from November 1, 2017, to February 28, 2019.
Exposures Five serum-based liver function markers (total bilirubin, albumin, alkaline phosphatase, alanine aminotransferase, and aspartate aminotransferase) from AD Neuroimaging Initiative participants were used as exposure variables.
Main Outcomes and Measures Primary outcomes included diagnosis of AD, composite scores for executive functioning and memory, CSF biomarkers, atrophy measured by magnetic resonance imaging, brain glucose metabolism measured by fludeoxyglucose F 18 (¹⁸F) positron emission tomography, and amyloid-β accumulation measured by [¹⁸F]florbetapir positron emission tomography.
Results Participants in the AD Neuroimaging Initiative (n = 1581; 697 women and 884 men; mean [SD] age, 73.4 [7.2] years) included 407 cognitively normal older adults, 20 with significant memory concern, 298 with early mild cognitive impairment, 544 with late mild cognitive impairment, and 312 with AD. An elevated aspartate aminotransferase (AST) to alanine aminotransferase (ALT) ratio and lower levels of ALT were associated with AD diagnosis (AST to ALT ratio: odds ratio, 7.932 [95% CI, 1.673-37.617]; P = .03; ALT: odds ratio, 0.133 [95% CI, 0.042-0.422]; P = .004) and poor cognitive performance (AST to ALT ratio: β [SE], −0.465 [0.180]; P = .02 for memory composite score; β [SE], −0.679 [0.215]; P = .006 for executive function composite score; ALT: β [SE], 0.397 [0.128]; P = .006 for memory composite score; β [SE], 0.637 [0.152]; P < .001 for executive function composite score). Increased AST to ALT ratio values were associated with lower CSF amyloid-β 1-42 levels (β [SE], −0.170 [0.061]; P = .04) and increased amyloid-β deposition (amyloid biomarkers), higher CSF phosphorylated tau181 (β [SE], 0.175 [0.055]; P = .02) (tau biomarkers) and higher CSF total tau levels (β [SE], 0.160 [0.049]; P = .02) and reduced brain glucose metabolism (β [SE], −0.123 [0.042]; P = .03) (neurodegeneration biomarkers). Lower levels of ALT were associated with increased amyloid-β deposition (amyloid biomarkers), and reduced brain glucose metabolism (β [SE], 0.096 [0.030]; P = .02) and greater atrophy (neurodegeneration biomarkers).
Conclusions and Relevance Consistent associations of serum-based liver function markers with cognitive performance and A/T/N biomarkers for AD highlight the involvement of metabolic disturbances in the pathophysiology of AD. Further studies are needed to determine if these associations represent a causative or secondary role. Liver enzyme involvement in AD opens avenues for novel diagnostics and therapeutics.
Increasing evidence supports that Alzheimer disease (AD) is a metabolic disease with high diabetes co-morbidity and a wide range of metabolic perturbations occurring early in the disease process. While genetic risk clearly plays a role in AD, the gut microbiome, exposome and diet also exert influences on brain metabolic health. Advances in analytical chemistry enable simultaneous measurement of 1000's of metabolites that can be integrated into tissue-specific metabolic networks enabling a new kind of metabolic systems analysis of AD and where peripheral and central metabolic changes can be connected. We will highlight developments in the metabolomics field and large initiatives within Accelerated Medicine Partnership (AMP-AD) and MOVE AD where metabolomics data is being generated and informed by genomics and imaging data. Fasting baseline serum samples from Alzheimer's Disease Neuroimaging Initiative ADNI1 (199 CN, 356 LMCI and 175 AD participants) were analyzed using targeted (AbsoluteIDQ-p180 and Biocrates Bile Acid Kit) and non-targeted metabolomics and lipidomics platforms (UPLC-TOF MS; UPLC-QTOF MS). ADNI GO/2 (150 CN, 100 SMC, 100 EMCI, 150 LMCI, and 150 mild AD) and Rotterdam studies were used for validation of findings. Partial correlation networks revealed changes in sphingomyelins and ether-containing phosphatidylcholines in preclinical biomarker-defined AD stages providing mechanistic insights about abeta and tau pathology and pointing to changes in membrane structure and function. Subsequent changes in acylcarnitines and several amines, including the branched chain amino acid valine and α-aminoadipic acid correlated with imaging changes and cognitive decline pointing to a role of mitochondrial energetics. Several metabolic changes related to liver enzymatic functions and gut microbiome activity correlated with imaging and cognitive changes highlighting a possible role for gut liver brain axis in disease pathogenesis. Meta-analysis across metabolomics studies reveals common biochemical findings when evaluated within a metabolic network context. Metabolomics identified key disease-related metabolic changes and disease-progression-related changes some related to liver and gut microbiome activity. Defining metabolic changes during AD disease trajectory and its relationship to clinical phenotypes provides a powerful roadmap for drug and biomarker discovery. Genomics and imaging data inform metabolomics data and together they provide a powerful systems approach for the study of AD.
Late-onset Alzheimer's disease (AD) can, in part, be considered a metabolic disease. Besides age, female sex and APOE ε4 genotype represent strong risk factors for AD that also give rise to large metabolic differences. We systematically investigated group-specific metabolic alterations by conducting stratified association analyses of 139 serum metabolites in 1,517 individuals from the AD Neuroimaging Initiative with AD biomarkers. We observed substantial sex differences in effects of 15 metabolites with partially overlapping differences for APOE ε4 status groups. Several group-specific metabolic alterations were not observed in unstratified analyses using sex and APOE ε4 as covariates. Combined stratification revealed further subgroup-specific metabolic effects limited to APOE ε4+ females. The observed metabolic alterations suggest that females experience greater impairment of mitochondrial energy production than males. Dissecting metabolic heterogeneity in AD pathogenesis can therefore enable grading the biomedical relevance for specific pathways within specific subgroups, guiding the way to personalized medicine.
Cholesterol has been implicated in the pathogenesis of Alzheimer's disease (AD). We seek out to encompass cholesterol contribution to AD in a metric suitable for application to human cohorts. To this end, we used the concept and assay for cholesterol efflux capacity of high-density lipoprotein (HDL) (CEC HDL), a metric that predicts atherosclerotic cardiovascular disease, as a paradigm to develop an analogous metric, CEC of cerebrospinal fluid (CEC CSF), for research and diagnostic purposes in AD. In the CEC HDL assay, radiocholesterol-labeled J774 macrophages are exposed to human HDL; radiocholesterol released to HDL is expressed as the percentage of the released and cell-retained radiocholesterol; efflux to sample HDL is divided by efflux to a standard HDL to derive CEC HDL values. Different cell types express cholesterol efflux transporters at varying levels. Macrophage is the most critical cell type in atherogenesis. Several cell types could be critical to AD. Neural SH-SY5Y, astrocyte A172, microglial N9 and J774 (control) cells were treated as in the CEC HDL assay and then exposed to CSF. CECs CSF for SH-SY5Y, A172, N9 and J774 cells were calculated as above. Commercially purchased and throwaway CSF samples (25 in total) were used in this study. SH-SY5Y, A172 and N9 cells expressed ABCA1, ABCG1, SR-B1 and ABCG4 and released cholesterol to apolipoprotein A-I (apo A-I) and HDL, albeit at lower levels compared to J774. CECs CSF varied 2-3X between the lowest and highest values. Remarkably, CECs CSF for N9 microglia and for J774 macrophages were nearly identical (R2=0.94; P<0.0001). CECs CSF for SH-SY5Y, A172 and N9 cells were correlated with one another less (R2=0.66-0.78; P<0.0001). CECs CSF for SH-SY5Y, A172 and N9 were strongly correlated with CSF apo A-I (R2=0.42-0.7) and apo J (R2=0.73-0.85) and weakly correlated with apo E (R2=0.28-0.39). After adjustment for apo J, the association between CECs CSF and apo A-I was no longer significant. CECs CSF for SH-SY5Y, A172 and N9 cells exhibit notable variability from individual to individual and from one and another and are determined mainly by CSF apo J levels.
Module-based analysis (MBA) aims to evaluate the effect of a group of biological elements sharing common features, such as SNPs in the same gene or metabolites in the same pathways, and has become an attractive alternative to traditional single bio-element approaches. Because bio-elements regulate and interact with each other as part of network, incorporating network structure information can more precisely model the biological effects, enhance the ability to detect true associations, and facilitate our understanding of the underlying biological mechanisms. However, most MBA methods ignore the network structure information, which depicts the interaction and regulation relationship among basic functional units in biology system. We construct the connectivity kernel and the topology kernel to capture the relationship among bio-elements in a module, and use a kernel machine framework to evaluate the joint effect of bio-elements. Our proposed kernel machine approach directly incorporates network structure so to enhance the study efficiency; it can assess interactions among modules, account covariates, and is computational efficient. Through simulation studies and real data application, we demonstrate that the proposed network-based methods can have markedly better power than the approaches ignoring network information under a range of scenarios.
Altered lipid metabolism has been implicated epidemiologically in risk for sporadic Alzheimer’s disease (AD); the underlying mechanisms remain uncertain. Although cholesterol and circulating lipoproteins have been the most well-studied lipids in AD, recent data implicate other classes such as glycerophospholipids (GPL). Plasmalogens (Pls), a GPL subclass, are integral membrane components with functions relevant to AD pathophysiology, including promoting vesicle fusion necessary for synaptic neurotransmitter release; modulation of membrane fluidity, raft dynamics, and antioxidant functions; and neuroprotection. Plasmalogen-rich membrane regions favor amyloid precursor protein (APP) breakdown by α-secretase to non-amyloidogenic products, over amyloidogenic β-secretase-mediated products that promote AD plaque formation. The initial steps of endogenous Pls synthesis require peroxisomes, notably in the liver. Peroxisomal function is decreased in older individuals; previous studies have shown decreased Pls in AD serum and postmortem brain. We measured Pls biosynthesis indices in ADNI-1 baseline serum vs. progression from MCI to AD. We measured 4 ethanolamine plasmalogens (PlsEtn) and 4 related phosphatidylethanolamines (PtdEtn) by flow-injection tandem mass spectrometry (FIA-MS/MS) in serum from 736 Alzheimer’s Disease Neuroimaging Initiative (ADNI-1) subjects that included 20 blinded technical replicates. We calculated 5 key ratios of PlsEtn species to each other and to corresponding PtdEtn reflecting Pls biosynthesis and peroxisomal beta-oxidation. Quality control analyses showed strong correlation between replicates (r2>0.95, p<0.001) for all ratios. We then calculated an index of plasmalogen biosynthesis from the mean of 2 ratios (PlsEtn 22:6/PtdEtn 22:6, PlsEtn 20:5/PtdEtn 22:6), termed “PL/PE.” We examined the relationship between the PL/PE ratio and progression to AD in 358 patients with MCI at baseline, adjusting for age, gender, APOE genotype, body mass index, and concomitant medications. Higher baseline PL/PE was associated with a significantly (HR = 0.80, p<0.01) reduced risk of progression to AD in patients with MCI at baseline (Figure).
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)