WO2020219957A1 - Methods of detecting and measuring molecular aging in the brain - Google Patents

Abstract

Provided herein are methods for analyzing the transcriptome to probe for biomarkers in sample from a human being to determine the molecular age to the brain to predict the propensity of the human being to show increased cognitive decline and other neurological disorders. Also provided are methods for screening transcriptome based SNP biomarkers, as well as methods for treating a human being having a high risk for cognitive decline and other neurological disorders.

Description

METHODS OF DETECTING AND MEASURING MOLECULAR AGING IN THE

BRAIN

RELATED APPLICATION

This application claims priority to U.S. Provisional Application No. 62/838,008, filed on April 24, 2019, incorporated by reference herein in its entirety.

INCORPORATION BY REFERENCE

Table 6, filed herewith as“Appendix A - Table 6” is incorporated herein by reference in its entirety.

BACKGROUND

Quality of life in old age is often compromised by dementia, mild cognitive impairment, and declining mobility. Dementia is common in the elderly with a prevalence that rises with age from about 5% in people 71 -79 years old to nearly 40% in people over 90 years old. Alzheimer’s dementia is the most common form of dementia (about 70% of cases), with vascular dementia, Lewy body dementia, frontotemporal dementia and Parkinson’s disease making up the majority of other cases. Many people (20-30%) meet criteria for more than one type of dementia (mixed dementia). Alzheimer’s dementia is characterized clinically by progressive memory impairment, declining judgment, and increased mood symptoms, leading to eventual loss of most cognitive function and death. Pathological features of Alzheimer’s dementia include irreversible neuronal loss, particularly in the hippocampus and temporal cortex, extracellular b amyloid plaques and neurofibrillary tangles. The

overwhelming majority of Alzheimer’s dementia cases are late onset (LOAD) (~93%) and non-familial (99%). The two biggest risk factors for LOAD are advanced age and the presence of e4 alleles of the APOE gene. How these risk factors relate to each other is an open question.

Studies have characterized age-related differences in transcription in the human brain. These studies show decreases in expression of neuronal synaptic-related genes, calcium signaling, and DNA damage-related genes and increases in glial inflammation-related genes with age. These transcriptional changes have been used as a gauge to assign a molecular age to any brain sample, and can identify brains that deviate from the expected based on chronological age; i. e. brains showing unusually slow or fast aging compared to the average. A few studies have further interrogated the intersection of normal brain aging and Alzheimer’s disease and have generally supported an overlap between normal age-related transcriptional differences in the brain and differences between Alzheimer’s dementia and control. Two studies showed aging-acceleration in Alzheimer’s dementia subjects vs.

controls. These studies are based on small cohorts (n<50 subjects) and do not relate brain aging to known genetic risk factors for Alzheimer’s dementia.

Besides advanced age, the biggest risk factor for LOAD is an allelic variant of the APOE gene. More than 15 genome wide association studies (GWAS) have implicated AROEe4 in Alzheimer’s dementia, making it by far the most consistent genetic risk factor. APOE encodes apolipoprotein E, a constituent of the low-density lipoprotein particle involved in clearance of cholesterol and a component of amyloid plaques. There are three human variants of the gene (AROEe2 (cysl 12, cysl58), AROEe3 (cysl 12, argl58), and AROEe4 (argl 12, argl58). AROEe3 is considered the wild-type allele and is the most common genotype with an allele frequency about 76%. AROEe4, with an allele frequency of about 14%, increases the lifetime risk of Alzheimer's disease by 2-4-fold. There also appears to be a dose effect, in that disease-free survival was shown to be lower in homozygotes compared to heterozygous. Consistent with these findings, AROEe4 alleles shift the age at onset earlier in the presence of one allele and earlier still in the presence of two alleles. AROEe4 may also be a risk factor for other dementias including dementia with Lewy bodies and perhaps PD, although the results of these studies are mixed with some showing significant risk and others showing none.

SUMMARY

The disclosure provides a method for determining the Aage of a subject comprising: measuring a transcriptome from a neuronal sample of the subject; determining gene expressions of one or more age-related genes from the isolated transcriptome; comparing the gene expressions from the subject to corresponding gene expressions in a pre-determined standard; using the comparison to model a molecular age of the subject; and calculating the difference between the molecular age of the subject and a standard molecular age at the chronological age of the subject to obtain Aage of the subject.

In another embodiment this disclosure also provides a method of determining the polygenic risk score (PRS) of a subject comprising: performing meta-analysis of genomewide association to identify one or more single nucleotide polymorphisms (SNPs) in a biological sample that is associated with the calculated Aage; and using machine learning to create an algorithm that weights these SNPs to create the PRS. In another embodiment the subject defined, is a mammal. In another embodiment, the mammal is a human.

In another embodiment, the biological sample is blood, serum, or epithelial cells. In another embodiment, the one or more age-related genes are selected from the group shown in Table 1. In one embodiment, the pre-determined standard is generated from a cohort selected from a Common Mind cohort.

In another embodiment, tire modeling of the molecular age comprises the steps of using elastic net regression of prefrontal cortex age-related transcripts controlling for potential sources of noise. In another embodiment, the one or more SNPs are selected from the group shown in Appendix A - Table 6, that has been incorporated by reference, herein, in its entirety.

The disclosure also provides a method of determining the propensity of a human subject to be afflicted with neurodegenerative disease and/or increased cognitive decline comprising, measuring a transcriptome from a neuronal sample of the subject; determining gene expressions of one or more age-related genes from the isolated transcriptome; comparing the gene expressions from the subject to corresponding gene expressions in a pre-determined standard; using the comparison to model a molecular age of the subject; and calculating the difference between the molecular age of the subject and a standard molecular age at the chronological age of the subject to obtain Aage of the subject, wherein if the molecular Aage is +5 or more, the probability of neurodegenerative disease and/or increased cognitive decline is greater than average. In another embodiment, the neurodegenerative diseases comprise Alzheimer’s Disease, Parkinson’s Disease, Frontotemporal Dementia, Depression,

Huntington’s disease, or amyotrophic lateral sclerosis. In another embodiment, the neurodegenerative disease is Alzheimer’s disease. In certain embodiments, the increased cognitive decline or neurodegenerative diseases are more prevalent in old age. In certain embodiments, old age is greater than 50, 55, 60, 65, 70, 75, or 80 years of age.

In one embodiment the method also includes the steps of determining the presence of an APOE allelic variant in a sample from the subject wherein: if the molecular Aage is +5 and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurodegenerative disease and/or increased cognitive decline; and if the molecular Aage is about 0 and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurodegenerative disease and/or increased cognitive decline. In another embodiment, the neurodegenerative diseases comprise Alzheimer’s Disease, Parkinson’s Disease, Frontotemporal Dementia, Depression,

Huntington’s disease, or amyotrophic lateral sclerosis. In another embodiment, the neurodegenerative disease is Alzheimer’s disease.

In another embodiment, the subject is a human. In another embodiment the subject defined, is a mammal. In another embodiment, the mammal is a human. In one embodiment, the biological sample is any human tissue containing cells. In another embodiment, the human tissue containing cells comprises blood, serum, or epithelial cells.

The disclosure also provides a method of treating a symptom associated with brain aging in a human subject in need thereof comprising administering to the subject a therapeutically effective amount of a sirtuin activator. In one embodiment the sirtuin activator activates SIRT1 or NAD.

The disclosure also provides a method of determining the polygenic risk score (PRS) of a subject comprising: identifying one or more SNPs in a biological sample from the subject that associate with Aage; and calculating a polygenic risk score (PRS) for the subject based on the presence of absence of the one or more SNPs. In one embodiment the subject defined, is a mammal. In one embodiment, the mammal is a human. In another embodiment, the biological sample is any human tissue containing cells. In another embodiment, the human tissue containing cells comprises blood, serum, or epithelial cells. In another embodiment, the one or more SNPs are selected from the group consisting of Appendix A - Table 6.

The disclosure also provides a method of determining the propensity of a subject to be afflicted with neurological disease comprising, identifying one or more SNPs in a biological sample from the subject that associate with Aage; and calculating a PRS for the subject based on the presence of absence of the one or more SNPs wherein if the molecular PRS is correlated with positive Aage, the probability of neurodegenerative disease and/or increased cognitive decline is greater than average. In another embodiment, the neurological diseases comprise Alzheimer’s Disease, Parkinson’s Disease, Frontotemporal Dementia, Depression, Huntington’s disease, or amyotrophic lateral sclerosis. In another embodiment, the neurodegenerative disease is Alzheimer’s disease. In certain embodiments, the increased cognitive decline or neurodegenerative diseases are more prevalent in old age. In certain embodiments, old age is greater than 50, 55, 60, 65, 70, 75, or 80 years of age.

In one embodiment, the method of determining the propensity of a subject to be afflicted with Alzheimer’s disease further comprises determining the presence of an APOE allelic variant in a sample from the subject wherein if the PRS is correlated with positive Aage and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurodegenerative disease and/or cognitive decline. In another embodiment, the neurodegenerative diseases comprise Alzheimer’s Disease, Parkinson’s Disease, Frontotemporal Dementia, or Depression. In another embodiment, the neurodegenerative disease is Alzheimer’s disease.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 shows a flow chart of the model development.

FIG. 2 shows validation of the predicted molecular age with the chronological age in all cohorts. FIGs. 2A, 2B and 2C show results using the Common Mind (CM) cohort. FIG. 2A shows data from all subjects, FIG. 2B shows data from younger subjects and FIG. 2C shows data from older subjects. FIGs. 2D and 2E show results using the PsychEncode (PE) cohort. FIG. 2D shows data from younger subjects and FIG. 2E shows older subjects. FIG. 2F shows results from the BrainCloud (BC) cohort. FIG. 2G shows results from the GTEx Consortium cohort (GTEx). FIG. 2H shows results from the ROS-MAP cohort, which is highly enriched in Alzheimer’s dementia

FIG. 3 shows calculation of molecular age and Aage using methylation data. FIG. 3 A shows correlation between molecular age calculated using methylation data and

chronological age for the BC cohort. FIG. 3B shows the same analysis for the ROS-MAP cohort. FIG. 3C shows correlation between Aage calculated using methylation data and transcriptional data in the BC cohort. FIG. 3D shows the same data with the ROS-MAP cohort.

FIG. 4A is a heat map showing 537 increasing and 834 decreasing transcripts from CM and PE cohorts.

FIG. 4B is a Venn diagram showing down regulated genes in PFC and neuronal cells.

FIG. 4C is a Venn diagram showing up-regulated in PFC and glial cells.

FIG. 5 shows relationships of Aage to clinical variables. Representative plots are shown using raw' data (a-i). P values w'erc determined by linear regression with relevant covariates. FIG. 5 A shows post-mortem final clinical diagnosis of Alzheimer’s. FIG. 5B shows mini mental exam score. FIG. 5C shows tangles density. FIG. 5D shows global PD score, a composite score for 4 signs: tremor, rigidity, bradykinesia, and gait. FIG. 5E shows rigidity score. FIG. 5F show's that PD pathology is present if Lewy bodies are present and that there was moderate to severe neuronal loss in the substantia nigra. FIG. 5G shows global cognition slope, a composite slope of the longitudinal changes overtime in five domains of cognition: working memory, visual-spatial ability, perceptual speed, episodic memory and semantic memory. FIG 5H shows change in episodic memory over time. FIG. 51 shows association of Aage & AROEe4.

FIG. 6A is a bar graph showing the odds ratio of Aage and APOEM for having Alzheimer’s dementia. FIG. 6B is a schematic showing synergistic effects of Aage and AROEe4 on Alzheimer’s dementia.

DETAILED DESCRIPTION

This disclosure relates to a reliable transcriptome based gauge of molecular brain aging, and use of this tool to determine how rapidly brains have aged compared to the average in a cohort. This method correlates well with the DNA methylation sites clock. The inferred aging rate is used to assess whether molecular brain aging is a risk factor for Alzheimer’s dementia and a variety of other late life maladies in a large naturalistic cohort of older subjects. Moreover, brain aging and AROEe4 status are related as risk factors for Alzheimer’s dementia, focusing most heavily on LOAD and cognitive aging. The findings presented herein suggest that brain aging is an important risk factor for Alzheimer’s dementia and acts synergistically with AROEe4 and may have important therapeutic implications for treating this and other brain diseases.

A detailed analysis of brain aging was carried out as described herein in order to ascertain how it interacts with other risk factors for neurodegenerative diseases.

Transcriptomic analysis was used to assign a molecular age to prefrontal cortex (PFC) samples from a series of cohorts from brain banks, and calculated the deviation of molecular age from chronological age (Aage) as a proxy for the rate of aging. These methods were developed by entraining the algorithm on one cohort of disease-free brains and applying them to four additional cohorts. This method showed a highly significant correlation between molecular and chronological age for all cohorts. Association of aging rates (Aage) was then examined in non-disease brains with Alzheimer’s dementia brains in the ROS-MAP cohort, which is highly enriched in Alzheimer’s dementia, and made several important findings.

First, a positive Aage (rapid aging compared to average across all brains) associates with risk for Alzheimer’s dementia with a high significance. Second, AROEe4, the strongest genetic risk factor for sporadic AD, also strongly associates with Alzheimer’s dementia risk in the same cohort, as expected. Third, a rapid aging rate and AROEe4 are synergistic risk factors. Brains with the slowest aging are strongly protected against having the AROEe4 allele, and brains with the fastest aging have a greatly elevated Alzheimer’s dementia risk when combined with AROEe4. Our findings thus suggest that Alzheimer’s dementia can be induced by the simultaneous occurrence of two risk factors, and that interventions against either one might protect against the disease. It is intriguing that imaging studies report an effect of AROEe4 on brain structure as early as infancy, suggesting ihatAPOEe4 alleles are pathological and not simply drivers of premature aging. Our data shows a weak association between AROEe4 and brain aging. The DNA methylation“clock” (Horvath, 2013) correlated well with the transcriptome gauge but was not as broad in predicting risk of brain dysfunctional endpoints in aging.

Our analysis also provides several molecular insights into brain aging in the neurologic disease-free cohorts. First, aging is associated with changes in transcripts affecting rho GTPases, which are associated with synapse formation and actin cytoskeleton dynamics in axons. Other top categories of transcripts reduced in the aging brain encode GTPase inhibitors and other synaptic functions. Deficits in all of these transcripts are exacerbated in the much older ROS-MAP cohort. Thus, these findings are consistent with earlier findings of synaptic deficits in the aging brain, and suggest that the transcriptional deficits observed may trigger the defect in synapses. This is also consistent with a study showing decreases in synaptic genes in Alzheimer’s dementia subjects vs. control subjects that occur at the onset of neuropathology and are exacerbated in subjects that are AROEe4 positive. Second, there is a decrease in transcripts associated with neuronal signaling and mitochondrial/sirtuin function. In particular, a reduction in glutamate receptor signaling was found, which may partly explain the strong association observed between the rate of normal brain aging and cognitive decline. The mitochondrial/sirtuin category included many genes involved in electron transport and ATP synthesis. This novel finding suggests an interplay between the sirtuin/mitochondrial pathway, neuronal signaling, and synaptic function in aging. It is possible that intervention strategies, such as sirtuin activation, may be protective against neurodegenerative diseases. Third, there is an upregulation in inflammatory pathways, which is indicative of glial activation and consistent with earlier studies. Fourth, there is a coordination in the age-differences of all of the aging-sensitive transcripts, suggesting that segmental aging does not occur, at least in the PFC. Fifth, while the rate of brain aging strongly correlates with Alzheimer’s dementia and other brain maladies, there is no association with cardiac disease or cancer. This interesting finding raises the possibility that the aging of different tissues is not coordinated with brain aging in an individual. Indeed, our data do not directly address whether rates of aging are coordinated across different regions of the brain (although see discussion of PD below). Sixth, there is no association between brain aging and a history of smoking, which might have been expected.

The changes observed in PFC aging are unlikely to be simply due to neuronal loss or glial gain. Indeed, the transcriptional changes used to define brain aging begin early in adulthood, before any significant neuronal loss would occur. In addition, there is little change in the CM or PE cohorts in transcripts defined as neuronal-specific or glial-specific, suggesting there is not a significant loss of neurons in PFC, consistent with earlier data. Finally, assigning molecular by the transcriptome gauge correlated well with assignment by the DNA methylation clock, the latter of which is predictive of age for many tissue types. We thus conclude that aging-promoted expression changes within the brain are not explained by neuronal loss or glial gain.

In the ROS-MAP cohort, a significant association between the rate of brain aging and other neurological disorders was observed. This observed association with Parkinsonian sign score is perhaps more surprising than the Alzheimer’s dementia association, since the dopaminergic neurons affected in PD are in a distinct brain region and have a distinct function compared to cortical neurons used to entrain the aging algorithm.

A very strong association between rapid brain aging and cognitive decline was found in both control and disease ROS-MAP subjects. This association indicates that brain aging predisposes to loss of cognitive functions, and serves as a strong validation that our assignment of molecular ages is robust.

The question arises whether fast brain aging causes Alzheimer’s dementia or Alzheimer’s dementia somehow triggers the rapid aging. The fact that Ingenuity analysis associated many of the aging sensitive transcripts with neurodegenerative diseases underscores the complexity of determining cause and effect. Several factors lead us to favor a model that rapid brain aging is a cause of Alzheimer’s dementia and not a result. First, brain aging was defined in the CM cohort, comprising subjects with no disease diagnoses and in which brains were judged to be free of pathology. Second, aging sensitive transcripts used to gauge molecular brain age changed from early ages and in a continuous way, well before any disease processes could have set in. Third, the association of older molecular brain age with a variety of diseases suggests a causal link between brain aging and Alzheimer’s dementia rather than accelerated aging being a consequence of the pathological brain. For example, if the aging were an effect, then we would have to conclude that both Alzheimer’s dementia pathology centered in the cortex and PD pathology centered in the dopaminergic neurons could both somehow speed up the aging of the PFC. Fourth, molecular age strongly associated with the rate of cognitive decline in subjects without diseases.

These findings suggest that slowing brain aging might delay or favorably slow progression of Alzheimer’s dementia, PD, cognitive decline, and potentially other neurological conditions of old age. Slowing molecular aging may be most effective as a preventative strategy before irreversible neuronal loss has occurred. In this regard, the detailed analysis of molecular brain aging may lead to specific genes and pathways that regulate the rate of aging and offer therapeutic targets for intervention to impact a broad spectrum of neurological diseases and deficits.

Definitions

It is to be understood that the methods described in this disclosure are not limited to particular methods and experimental conditions disclosed herein; as such methods and conditions may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting.

Furthermore, the experiments described herein, unless otherwise indicated, use conventional molecular and cellular biological and immunological techniques within the skill of the art. Such techniques are well known to the skilled worker, and are explained fully in the literature. See, e.g., Ausubel, el al, ed., Current Protocols in Molecular Biology, John Wiley & Sons, Inc., NY, N.Y. (1987-2008), including all supplements, Molecular Cloning: A

Laboratory Manual (Fourth Edition) by MR Green and J. Sambrook and Harlow et al , Antibodies: A Laboratory Manual, Chapter 14, Cold Spring Harbor Laboratory, Cold Spring Harbor (2013, 2nd edition), incorporated herein by reference in its entirety.

Unless otherwise defined herein, scientific and technical terms used herein have the meanings that are commonly understood by those of ordinary skill in the art. In the event of any latent ambiguity, definitions provided herein take precedent over any dictionary' or extrinsic definition. Unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular. The use of "or" means "and/or" unless stated otherwise.

The use of the term "including", as well as other forms, such as "includes" and "included", is not limiting. Generally, nomenclatures used in connection with cell and tissue culture, molecular biology, immunology, microbiology, genetics and protein and nucleic acid chemistry and hybridization described herein are those well-known and commonly used in the art. The methods and techniques provided herein are generally performed according to conventional methods well known in the art and as described in various general and more specific references that are cited and discussed throughout the present specification unless otherwise indicated. Enzymatic reactions and purification techniques are performed according to manufacturer's specifications, as commonly accomplished in the art or as described herein. The nomenclatures used in connection with, and the laboratory procedures and techniques of, analytical chemistry, synthetic organic chemistry, and medicinal and pharmaceutical chemistry described herein are those well-known and commonly used in the art. Standard techniques are used for chemical syntheses, chemical analyses, pharmaceutical preparation, formulation, and delivery, and treatment of patients.

That the disclosure may be more readily understood, select terms are defined below. As used herein, the term "nucleic acid" refers to a polymer of two or more nucleotides or nucleotide analogues (such as ribonucleic acid having methylene bridge between the 2 -0 and 4’-C atoms of the ribose ring) capable of hybridizing to a complementary nucleic acid. As used herein, this term includes, without limitation, DNA, RNA, LNA, and PNA.

The term "subject", as used herein, refers to a mammal, e.g. a human, a domestic animal or a livestock including a cat, a dog, a rabbit, cattle and a horse. Mammals also include rodents. Rodents can include mice and rats. Mammals also include non-human primates including monkeys.

The term“molecular age” as used herein, refers to an age assigned to the brain capacity of a subject. Brain function varies with chronological age, but varies unevenly across the human population. There are several genetic and epigenetic factors that influence the variance in brain function actually experienced by subjects as they age. The molecular age is a measure of brain function at a particular chronological age that takes in genetic and epigenetic variance that occurs across a population of subjects.

The term“Aage” as used herein, refers to the difference in predicted brain function expressed as a difference between an average chronological age across a subject population and the molecular age of a particular subject.

The term“age-related gene” as used herein, refers to a number of genes that have been shown to be differentially expressed as subjects age on average. For example, these genes include those listed in Table 1 and represented by the heat map in FIG. 4.

The term“transcriptome” as used herein, refers to a measure of the expression of messenger RNA (mRNA) molecules in a sample for a selected number of genes. In certain embodiments, the transcriptome being studied encompasses the expression of mRNA from one or more age-related genes. The term“biological sample” as used herein, refers to a sample from a subject. In some embodiments, the sample is from a specific tissue or bodily fluid from the subject. In some specific embodiments, the tissue is blood or epithelial cells. In other embodiments, the tissue is skin, hair or neuronal tissue. In other embodiments, the bodily fluid is serum, saliva, urine, semen, tears, lymph, mucus, cerebrospinal fluid, synovial fluid, bile, plasma, pus, breast milk or amniotic fluid.

The term“machine-learning” refers to data analysis that automates analytical model building. In some embodiments, the term“machine learning,” as used herein, may refer to a complex mathematical model for data classification that is generated using machine-learning techniques. In other embodiments,“machine learning” may include any process that tunes a number of parameters to be simultaneously optimal on training dataset using one or more machines.

The term“algorithm,” as used herein, is a broad term and is used in its ordinary sense, including, but not limited to, the computational processes. In some embodiments the term “algorithm” means any self-consistent set of ordered steps specifying definable operations upon data and leading to a particular result.

The term“single-nucleotide polymorphism” or“SNP”, used interchangeably herein, refers to variation in nucleotide sequence that differs at a single base pair. In some embodiments, the SNP can alter the structure and function of the corresponding gene product (i.e. protein). In other embodiments, the SNP affects transcription of the gene with which it is associated.

The term‘^polygenic risk score” or“PRS” used interchangeably herein, refers to a number based on variation in multiple genetic loci and their associated weights. In some embodiments, the linear or nonlinear function of the estimated statistical parameters includes per gene variant allele effect size mean and/or estimates of variability. In some embodiments, computing comprises linear weighting of each gene variant by its estimated posterior effect size divided by its estimated posterior variance. In some embodiments, the process or determining the PRS further comprises the step of obtaining maximal correlation of genetic risk scores with phenotypes in de novo subject samples by obtaining posterior effect size estimates for each SNP modulated by genie annotations and/or strength of association with pleiotropic phenotypes.

Brain aging refers to the increase in cognitive decline and incidence of many common neurodegenerative diseases as the brain ages. These neurodegenerative diseases include mild cognitive impairment, dementia, Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, tauopathy, basal ganglia disorders, schizophrenia, Lou Gehrig’s disease, and brain disease that is age gated, i.e., that have specific ages of onset. Brain aging can also be shown by decreases in reaction time and mobility in a subject. Brain aging is also associated with reduction in brain plasticity, reduction in gray matter, changes in neuronal morphology, reduction of attention span and reduction of memory function. The instant disclosure provides methods of detecting advanced or slowed brain aging in a subject at a given chronological age. The instant disclosure also provides methods of detecting a predisposition for or against the development of neurodegenerative diseases associated with brain aging. In particular, the disclosure provides methods of detecting a predisposition for onset of

Alzheimer’s disease. The detection of these predispositions can lead to subjects being prescribed regimens that could delay the effects of aging. These regimens include physical exercise, being intellectually engaged, maintaining social networks and maintaining a healthy diet.

Cognitive decline may refer to a modest disruption of memory often referred to as age-associated cognitive impairment” or“mild cognitive impairment” (MCI) that manifests as problems with memory or other mental functions such as planning, following instructions, or making decisions that have worsened over time while overall mental function and daily activities are not impaired. In some embodiments a severe cognitive decline or impairment manifests as more severe problems with memory, learning, concentration, or making decisions that affect daily life. Non-limiting examples of measures of cognitive decline include memory, reaction time, learning, thinking, language, judgment, decision-making, and motor coordination. Diseases or disorders associated with cognitive decline include dementia, Alzheimer's disease, delirium, and amnesia.

According to certain embodiments, one treatment of age related neurodegenerative diseases or other neural defects associated with aging is activation of sirtuin. In certain embodiments, the sirtuin is SIRT1, 2, 3, 4, 5, 6, or 7. According to a specific embodiment, the sirtuin is SIRT1.

Furthermore, in accordance with the present disclosure there may be employed conventional molecular biology, microbiology, and recombinant DNA techniques within the skill of the art. Such techniques are explained fully in the literature. See, e.g., Sambrook,

Fritsch & Maniatis, Molecular Cloning: A Laboratory Manual, Second Edition (1989) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (herein "Sambrook et al., 1989"); DNA Cloning: A Practical Approach, Volumes I and II (D. N. Glover ed. 1985);

Oligonucleotide Synthesis (M. J. Gait ed. 1984); Nucleic Acid Hybridization [B. D. Hames & S. J. Higgins eds. (1985)]; Transcription And Translation [B. D. Hames & S. J. Higgins, eds. (1984)]; Animal Cell Culture [R. I. Freshney, ed. (1986)]; Immobilized Cells And Enzymes [IRL Press, (1986)]; B. Perbal, A Practical Guide To Molecular Cloning (1984); F. M.

Ausubel et al. (eds.), Current Protocols in Molecular Biology, John Wiley & Sons, Inc.

(1994).

The following examples are provided to further elucidate the advantages and features of the present application, but are not intended to limit the scope of the application. The examples are for illustrative purposes only.

EXAMPLES

Example 1: Approach to determine aee-sensitive transcripts.

The goal of this study was to determine the molecular age of the brain to predict the development of a neurological disease in an individual. The flow chart of the strategy is shown in FIG. 1. A large cohort of 239 human subjects free of neurological disease between the ages 25-97 years, the Common Mind (CM) cohort, was used to determine transcripts that increase or decrease with age. The Common Mind cohort was used as the training cohort for determination of the molecular age. We found ~7% of all transcripts differ by age in a monotonic way across the entire cohort. Other cohorts such as PsychEncode (PE) (n=216), GTEx (n=87) (The GTEx Consortium 2015), and BrainCloud (n=127), were used to validate the predictive power of the method developed in the CM cohort. Samples from the prefrontal cortex (PFC) was used for all the cohorts as PFC is highly affected by aging and a variety of neurodegenerative diseases but shows little to no significant neuronal death with age unlike many other regions of the brain. Therefore, expression studies in PFC should be minimally confounded by changing cell-type numbers. Thus, 834 transcripts that decrease with age and 537 transcripts that increase with age were identified (Table 1). Importantly, the

transcriptionally-defined brain aging begins in early adulthood and progresses linearly thereafter. A computational model (elastic net regression controlling for potential sources of noise such as sex and RNA quality) was then used to calculate a molecular age for each brain.

Molecular age and was well correlated to the chronological age in all cohorts. Aage was calculated as the difference between each individual data point and the regression line in all cohorts. FIG. 2A-F show validation of our predicted molecular age with the chronological age in all cohorts.

The CM cohort showed a high correlation of molecular age and chronological age at time of death (FIG. 2A). The other three cohorts, PsychEncode (PE) (n=216), GTEx (n=87)(The GTEx Consortium2015), and BrainCloud (n=127), were used to validate the predictive power of the method developed in the CM cohort. The molecular age and chronological age were well correlated in all cohorts (FIG. 2). It is notable that the older subjects in the CM and PE cohorts appeared to have a reduction in the Pearson correlation coefficient (R value) relating molecular and chronological age compared to younger subjects (FIG. 2).

Another cohort tested was ROS-MAP, a large cohort (n=430) of subjects (67-108 years) that is a combination of subjects from the Religious Order Study and the Rush

Memory and Aging Project (Table 3). This cohort is much older than the others, which may explain the somewhat lower R value for chronological versus molecular age (FIG. 2). Note this correlation is still highly significant (p^O ⁰⁸). Older subjects may generally show a reduction in correlation due to a time-dependent divergence in differences in aging rates among individuals in the cohorts.

A proxy for the rates of brain aging was obtained as the difference between molecular age and chronological age far each brain, termed Aage (see FIG. 1). Aage was calculated for each brain as the difference between each individual data point and the regression line in all cohorts, which we return to below.

Example 2. Comparison of transcription-based and methvlation-based assignments of aces.

Methylation data was available for subjects in the BC and ROS-MAP cohorts and were used as a second way to assign molecular age. As shown in FIGs 3A and 3B, there was a strong correlation between molecular and chronological age in BC (R=0.98), and ROS- MAP (R=0.67, p<2.2 x 10^~16), as had been previously shown. Next Aage in these two cohorts as determined by both methods the transcriptional and methylation methods was compared to ascertain whether it would show similar deviations from normal aging. This showed a highly significant correlation (R=0.69, p=1.2 x 10^-11) in BC and in ROS-MAP (R=0.43, p<2.2 x 10^~16)between the two methods (FIGs 3C and 3D). It is unlikely that the two methods reveal the same molecular events since methylation data applies to many different tissues showing highly variable transcriptional changes with aging. It was confirmed that the DNA methylation markers were not proximal to age-regulated transcripts (data not shown), as has been observed in other tissues.

Example 3: Characterization of aging-sensitive transcripts

FIG. 4 shows the heat map of 537 increasing and 834 decreasing transcripts described in Table 1 from the Common Mind and PsychEncode cohorts. Each column represents one subject and shows the 537 transcripts that increase with age (top in red) or the 834 transcripts that decrease with age (bottom in blue). It was observed that most transcripts show continuous incremental differences with the age of the subjects, suggesting that brain aging is a continual process from early in adulthood. This finding along with the published data on PFC argued that the transcriptional changes are not simply due to neuronal loss and a correspondingly higher glial composition. It was also evident that some brains appear exceptional, either slowed or advanced for aging compared to the average. These exceptional cases showed the highest absolute values of Aage in our analyses described above.

The aging-sensitive transcripts were functionally grouped by Ingenuity software. Neurological disease genes were a top disease category for the transcripts that were lower in older people. Such genes were related to PD (p=2 x 10³), Tauopathy (p=7 x 10³),

Huntington’s (p=2 x 10 ¹¹), ALS (p=4 x 10³), basal ganglia disorders (p=l x 10 ¹⁰), and other neurodegenerative disorders as well as to psychiatric disorders including anxiety (p=l x 10⁴), depression (p=2 x 10⁴), bipolar disorder (p=l x 10³) and schizophrenia (p=2 x 10⁷). More specific pathway categories for down regulated genes include“glutamate signaling”,

“dopamine feedback in cAMP signaling” (not shown, p=l x 10⁵), and“rho GTPases” (FIG. 4A). The rho-GTPases are particularly interesting, because they are known regulators of synaptic spine formation and actin cytoskeletal dynamics. Thus, these changes are consistent with earlier findings showing deficits in synaptic function and neuronal signaling.

Three of the categories of down regulated genes are“mitochondrial dysfunction,” “oxidative phosphorylation” (not shown, p=2 x 10⁵), and“sirtuin signaling”. One exciting possibility is that defective sirtuin function contributes to mitochondrial and oxidative phosphorylation defects, which then impair neuronal function. However, whether mitochondrial dysfunction causes defects in neuronal function, neuronal functional defects cause mitochondrial defects, or the two are causally unlinked, was not discemable.

Among upregulated categories, were genes involved in cell morphology, immune cell trafficking, cancer, and cell to cell signaling were top categories. Also among upregulated categories are pathways involved in inflammation and DNA damage/cancer, consistent with earlier findings and illuminated the deteriorating environment of the aging brain.

To test whether the changes in the cellular composition accounted for the transcriptional changes, the transcript data was mapped onto the various cell types of the brain. A recent study identified consensus brain cell type-specific transcriptional markets based on the overlap of five murine and human single cell RNA-seq studies. Transcripts were grouped as“neuronal” or“glial” (encompassing astrocytes, oligodendrocytes, and microglia) and queried the fraction that decreased (FIG. 4B) or increased (FIG. 4C) with aging. If there were significant loss of neurons in the PFC with age or gain of glial in the CM dataset, it might be expected that all neuronal-specific transcripts would decrease with age and that all glial-specific transcripts would increase with age (with some small margin of error for statistical chance).

As evident in the Venn diagrams (FIG. 4B and 4C), this was not the case as only a small fraction of the neuronal or glial specific transcripts changed with age, providing further evidence for the surmise that changes in cellular composition occur minimally in the aging PFC and do not account for the aging sensitive transcripts. Further, the methylation clock, which is likely independent of transcription and thus not similarly susceptible to cellular composition changes, was totally consistent with the transcriptional gauge. We thus conclude that changes in cellular composition are unlikely to explain the aging-regulated transcripts. Example 4: Transcriptional Aape associates with Alzheimer’s dementia. PD and other phenotypes

FIG. 5 shows the association of Aage with Alzheimer’s dementia (AD), PD and other diseases. ROS-MAP subjects had been followed longitudinally by neurological examination for a variety of diseases and phenotypes including AD, dementia, motor and cognitive function, and their brains were characterized for disease-related pathology after death. APOE genotype data was also available for ROS-MAP subjects. The relationship of Aage in ROS- MAP as applied to 38 clinical and pathological phenotypes were investigated. The characteristics of 430 subjects in the ROS-MAP cohort is listed in Table 3. The average age of this cohort at the time of death was 87 years and about 2/3 of subjects had a pathological or clinical diagnosis of neurological disease. AROEe4, a known biomaiker for AD, was covaried, when assessing the relationship of Aage to variables and likewise co- varied for Aage when assessing the relationship of AROEe4 to variables in order to isolate their effects. P- values were corrected for multiple testing.

From the data, positive Aage was associated with risk of clinical diagnosis of AD as subjects show significantly older aging compared to non-disease controls (FIG. 5A).

Corroborating this finding, subjects with older Aage values also performed worse on the clinical Alzheimer’s dementia mini mental exam (FIG. 5B). The ROS-MAP brains were also quantified for levels of a pathological marker of AD, tangles, which was significantly and positively associated with a positive value of Aage (FIG. 5C).

A positive Aage was also associated with Parkinsonian sign score assessed across all 430 ROS-MAP subjects, which is a composite score of clinical signs of PD comprising rigidity (muscle stiffness), tremor (involuntary oscillation of limbs), gait (shortened, shuffling walking), and bradykinesia (FIG. 5D) and with rigidity and gait separately (FIG. 5E, Table4). Likewise, positive Aage was also significantly associated with the presence of PD pathology

(p=0.006, FIG. 5F). An association with PD diagnosis was not observed, but this may be because of the low number of PD-diagnosed subjects in the cohort (n=31, Table 3).

The strongest association of positive Aage was with global cognition slope (p=5 x 10^{" 5}, FIG. 5G), which is the composite rate of cognitive decline over time for five different cognitive domains (episodic memory, visual-spatial ability, perceptual speed, semantic memory, and working memory). This data included all 430 ROS-MAP subjects, although strong associations were found individually in Alzheimer’s dementia and disease and pathology free subjects (Table 2). Each of the above cognitive domains were also individually significant for association with positive Aage (FIG. 5H, Table 4).

Table 2 shows P-values obtained from multiple regression of Aage with variables of interest segregated into control and disease cases. Red p-values indicate significantly increased risk with older D age. Those in blue indicate the inverse relationship. Cognitive measures were performed longitudinally by clinicians.

Because our goal was to probe any association between aging and risk alleles in Alzheimer’s dementia, we tested for an association between AROEe4 and Aage by stratifying the entire ROS-MAP population into groups with 0, 1 or 2 AROEe4 alleles. There was a weak association found between Aage and one copy of AROEe4 (FIG. 51). As expected, the AROEe4 allele showed a highly positive association with AD- and cognition-related measures (Table 4).

These findings suggested that positive Aage had a large impact on risk for a variety of common late-life diseases and impairments, and to a similar extent as AROEe4. While most brain-related diagnoses and phenotypes associated significantly with Aage, peripheral phenotypes such as cardiovascular phenotypes and pathology, thyroid disease, and cancer did not (Table 2). This dichotomy may reflect the feet that our methods used to calculate Aage are specific to brain, or might indicate that different tissues age at different rates in the same person, or it may reflect greater power for associations for conditions assessed with greater veracity in two studies of aging and dementia

Example 5: Comparing transcriptional and methvlation ages in relation to Alzheimer’s dementia and related variables

Subset of ROS-MAP subjects that had both methylation and transcriptional data available was used for this study (n=336) to directly compare the impact of transcriptional delta ages to methylation delta ages (Table 4).

P-values obtained from multiple regression of either methyl D age or transcriptional D age with variables of interest. Red p-values indicate significantly increased risk with older D age. Those in blue indicate the inverse relationship. Cognitive measures were performed longitudinally by clinicians.

As with the 430 subjects shown above, it was found that older transcriptional Aages continued to be significantly associated with Alzheimer’s dementia (p=1.3 x 10 ^L), faster global cognitive decline (p=T .8 x 10^-5), episodic memory decline (p=l .2 x 10^-4), visual spatial ability decline ip 1.3 x 10^-4), perceptual speed decline (p=5 x 10) ⁵), working memory decline (p=0.02), greater Parkinsonian sign score (p=0.01), more severe dementia (p=7.6 x 10^{- 0 .05}), tangles (p=1.0 x 10^-5), and Lewy body pathology (p=0.005)(Table 5). With the exception of dementia grade (p=0.006), these variables were not significantly associated with methylation Aages (Table 5). However, depression score (p=0.03), amyloid (p=0.05), Parkinsonian gait (p=0.03), bradykinesia (p=0.05), and PD pathology (p=0.001) were associated with methylation delta ages. These findings suggest that both methods are useful predictors, and that the transcriptional gauge may have a broader reach in associating with the risk of brain dysfunctional endpoints in aging.

Example 5: Synergistic effect nf Aage and AROEe4 in AD

FIG. 6 shows transcriptional Aage and AROEe4 are synergistic risk factors for AD.

The odds of having a clinical diagnosis of AD were calculated with respect to each. Aage was binned into younger (-1 standard deviation from the mean or— 5 molecular years), neutral (-5 years to +5 years), or older (+1 standard deviation from the mean or +5 molecular years), and combined subjects that were homozygous for AROEe4 with those that were heterozygous because there were only four homozygous subjects. It was found that Aage and AROEe4 are synergistic risk factors for Alzheimer’s dementia (FIG. 6A and 6B). For example, subjects who are +5 years and bear one or two APOE e4 alleles have more than 5x the average odds of having Alzheimer’s dementia. However, subjects who are - 5 molecular years and with one or two AROEe4 alleles have no elevated chance of Alzheimer’s dementia compared to subjects of average molecular age with no APOEM alleles. These findings suggest that younger Aage can protect against AROEe4 alleles. In summary, our findings suggest that AROEe4 and older Aage contribute syneigistically to risk of Alzheimer’s dementia and a variety of age-related neurological diseases and dysfunction, and AROEe4 is unlikely to increase risk by simply increasing the rate of brain aging.

Example 6. Brain Aging Polygenic Risk Score Prediction of Neurological Phenotypes. A polygenic risk score has been created based on a set of approximately 1000 single nucleotide polymorphisms (SNPs) that can predict the deviation of biological brain age from chronological brain age, (Aage), which is a proxy for the rate of brain aging. Biological age was determined from transcriptome profiling. Aage is predictive of a variety of brain-related health and wellness measures, including rate of cognitive aging and Alzheimer’s disease.

This simple test for Aage using DNA from blood or a cheek swab would be useful for identifying at risk individuals for faster brain aging and disease.

As described above, transcriptomic analysis was used to assign a biological age to prefrontal cortex (PFC) samples from a series of cohorts from brain banks, and calculated the deviation of biological age from chronological age (Aage) as a proxy for the rate of aging. These methods were developed by entraining the algorithm on one cohort of disease-free brains and applying them to four additional cohorts. This method showed a highly significant correlation between biological and chronological age for all cohorts. It was then showed that positive Aage (foster brain aging) associates with risk for AD, PD signs, and longitudinal cognitive decline with a high significance in one cohort for which data has had been collected for human postmortem transcriptome data, genetic data, and clinical variables. Meta-analysis of genome-wide association was used to identify SNPs that associate with Aage. Machine learning was also used to create an algorithm that weights these SNPs to create a polygenic risk score (PRS) that can be used as a proxy for Aage. This is shown in Appendix A - Table 6, incorporated by reference, herein. This PRS was highly predictive of Aage and was also predictive of the same clinical variables that Aage itself predicted. PRS will be associated with neurological measures and neurological disease using linear regression in many several larger cohorts including the Baltimore Longitudinal Study and the UK biobank.

Other genetics tests for subjects at risk cognitive aging and disease include APOE4, an AD PRS, a cognitive aging PRS, and a mcthylation based biological aging test. It was shown in our initial cohort that our Aage PRS was a better predictor than any of these other tests. It was also additive with APOE4 so would improve on that genetic test when used in combination. Example 7: Prediction of co^gnitive resilience in a large living human study.

We tested whether our initial polygenic risk score comprised of ~8000 snps would predict cognitive resilience in a large living human study, the UK Biobank, and show' here that it is predictive. The UK biobank is a 500,000 person study of people in the UK who have genetic information as well as cognitive testing over time at 1-3 years intervals. We show that for a variety of tests our score is predictive of how well people retain their peak cognitive performance as they age. Tests range from reaction time, to general cognition, to memory. This score should be useful clinically or commercially for determining risk or resilience to cognitive aging.

In Table 7, tests of cognition are presented in column 1. Successive columns show results of these tests on successive visits overtime. The first three columns provide data regarding correlation of the tests of cognition with age. The subsequent visits show that these measures do indeed decline overtime as expected. Columns marked“n” are number of subjects,“t” are t-values or a measure of how much of a relationship age or PRS relates to these values, and“p” are the p-values or significance of the relationship. Statistically significant p-values <0.05, are shown in bold.

REFERENCES

Akbarian S, Liu C, Knowles JA, Vaccaiino FM, Famham PJ, Crawford GE, Jaffe AE, Pinto D, Dracheva S, Geschwind DH, et al. 2015. The PsychENCODE project. Nat Neurosci 18: 1707-1712. http://www.nature.com/doifinder/10.1038/nn.4156.

Avramopoulos D, Szymanski M, Wang R, Bassett S. 2011. Gene expression reveals overlap between normal aging and Alzheimer’s disease genes. Neurobiol Aging 32:

2319.e27-34. http://www.sciencedirect.com/sciencc/article/pii/SO 197458010001843.

Beach TG, Sue LI, Walker DG, Roher AE, Lue L, Vedders L, Connor DJ, Sabbagh MN, Rogers J. 2008. The sun health research institute brain donation program: Description and experience, 1987-2007. Cell Tissue Bank 9: 229-245.

Benjamini Y, Hochberg Y. 1995. Controlling the false discovery' rate: a practical and powerful approach to multiple testing. J R Stat Soc 57: 289-300.

Bennett DA, Buchman AS, Boyle PA, Barnes LL, Wilson RS, Schneider JA. 2018. Religious Orders Study and Rush Memory and Aging Project. J Alzheimer’s Dis. Bennett DA, Schneider JA, Arvanitakis Z, Wilson RS. 2012a. Overview and findings from the religious orders study. Curr Alzheimer Res 9: 628-45.

http://www.ncbi.nlm.nih.gOv/pubmed/22471860%5Cnhttp://www.pubmedcentral.nih.

gov/articlerender.fcgi?artid=PMC3409291.

Bennett DA, Wilson RS, Arvanitakis Z, Boyle PA, De Toledo-Morrell L, Schneider JA. 2012b. Selected findings from the religious orders study and rush memory and aging project. Adv Alzheimer’s Dis 3: 397-403.

Bertram L, Lill CM, Tanzi RE. 2010. The genetics of Alzheimer disease: back to the future. Neuron 68: 270-81. http://www.ncbi.nlm.nih.gov/pubmed/20955934.

Bossers K, Wirz KTS, Meerhoff GF, Essing AHW, Van Dongen JW, Houba P, Kruse CG, Verhaagen J, Swaab DF. 2010. Concerted changes in transcripts in the prefrontal cortex precede neuropathology in Alzheimer’s disease. Brain 133: 3699-3723.

Bras J, Guemeiro R, Darwent L, Parkkinen L, Ansorge O, Escott-Price V, Hernandez DG, Nalls M a, Clark LN, Honig LS, et al. 2014. Genetic analysis implicates APOE, SNCA and suggests lysosomal dysfunction in the etiology of dementia with Lewy bodies. Hum Mol Genet 23: 1-8. http://www.ncbi.nlm.nih.gov/pubmed/24973356.

Cao K, Chen-Plotkin AS, Plotkin JB, Wang L-S. 2010. Age-correlated gene expression in normal and neurodegenerative human brain tissues. PLoS One 5: 9.

http://dx.plos.org/10.1371/joumal.pone.0013098. Colantuoni C, Lipska BK, Ye T, Hyde TM, Tao R, Leek JT, Colantuoni E a., Elkahloun AG, Herman MM, Weinberger DR, et al. 2011. Temporal dynamics and genetic control of transcription in the human prefrontal cortex. Nature 478: 519-523.

http://dx.doi.org/10.1038/naturel0524.

De Jager PL, Ma Y, McCabe C, Xu J, Vardarajan BN, Felsky D, Klein HU, White CC, Peters MA, Lodgson B, et al. 2018. Data descriptor: A multi-omic atlas of the human frontal cortex for aging and Alzheimer’s disease research. Sci Data 5: 1-13.

Dean DC, Jerskey BA, Chen K, Protas H, Thiyyagura P, Roontiva A,

O’Muircheartaigh J, Dirks H, Waskiewicz N, Lehman K, et al. 2014. Brain differences in infants at differential genetic risk for late-onset Alzheimer disease: a cross-sectional imaging study. JAMA Neurol 71: 1 1-22.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4056558&tool=pmcentrez&rende rtype=abstract.

Erraji-Benchekroun L, Underwood MD, Arango V, Galfalvy H, Pavlidis P,

Smymiotopoulos P, Mann JJ, Sibille E. 2005. Molecular aging in human prefrontal cortex is selective and continuous throughout adult life. Biol Psychiatry 57: 549-58.

http://www.sciencedirect.eom/science/article/pii/S0006322304011187.

Frontier M, Roussos P, Sieberts SK, Johnson JS, Kavanagh DH, Perumal TM, Ruderfer DM, Oh EC, Topol A, Shah HR, et al. 2016. Gene expression elucidates functional impact of polygenic risk for schizophrenia. Nat Neurosci 19: 1442-1453.

http://www.nature.com/doifinder/10.1038/nn.4399.

Glorioso C, Oh S, Douillard GG, Sibille E. 2011. Brain molecular aging, promotion of neurological disease and modulation by sirtuin 5 longevity gene polymorphism. Neurobiol Dis 41: 279-90.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3014380&tool=pmcentre z&rendertype=abstract.

Glorioso C, Sibille E. 2011. Between destiny and disease: genetics and molecular pathways of human central nervous system aging. Prog Neurobiol 93: 165-81.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3034779&tool=pmcentre z&rendertype=abs tract.

Glorioso C, Sibille E. 2010. Between destiny and disease: genetics and molecular pathways of human central nervous system aging. Prog Neurobiol.

http://www.ncbi.nlm.nih.gov/pubmed/21130140. Hastie T, Tibshirani R, Narasimhan B, Chu G. 2016. impute:impute: Imputation for microarray data. R package version 1.46.0.

Haug H, Kuhl S, Mecke E, Sass NL, Wasner K. 1984. The significance of morphometric procedures in the investigation of age changes in cytoarchitectonic structures of human brain. J Himforsch 25: 353-374.

http://www .ncbi.nlm.nih.gov/entrez/query ,fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=6481152.

Horvath S. 2013. DNA methylation age of human tissues and cell types. Genome Biol

14: R115.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4015143 &tool=pmcentre z&rendertype=abstract.

Jellinger K a. 2013. Pathology and pathogenesis of vascular cognitive impairment-a critical update. Front Aging Neurosci 5: 1-19.

Lambert JC, Ibrahim-Veibaas CA, Harold D, Naj AC, Sims R, Bellenguez C, DeStafano AL, Bis JC, Beecham GW, Grenier-Boley B, et al. 2013. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 45: 1452- 8. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3896259&tool=pmcentre z&rendertype=abstract.

Lefort R 2015. Reversing synapse loss in Alzheimer’s disease: Rho-guanosine triphosphatases and insights from other brain disorders. Neurotherapeutics 12: 19-28.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4322073&tool=pmcentre z&rendertype=abstract.

Lipska BK, Deep-Soboslay A, Weickert CS, Hyde TM, Martin CE, Herman MM, Kleinman JE. 2006. Critical Factors in Gene Expression in Postmortem Human Brain: Focus on Studies in Schizophrenia. Biol Psychiatry 60: 650-658.

Lu T, Pan Y, Kao SY, Li C, Kohane I, Chan J, Yankner BA. 2004. Gene regulation and DNA damage in the ageing human brain. Nature 429: 883-891.

Mayeux R, Stem Y. 2012. Epidemiology of Alzheimer disease. Cold Spring Harb Perspect Med 2.

McKenzie AT, Wang M, Haubetg ME, Fullard JF, Kozlenkov A, Keenan A, Hurd YL, Dracheva S, Casaccia P, Roussos P, et al. 2018. Brain Cell Type Specific Gene

Expression and Co-expression Network Architectures. Sci Rep. Miller JA, Oldham MC, Geschwind DH. 2008. A systems level analysis of transcriptional changes in Alzheimer’s disease and normal aging. J Neurosci 28: 1410-1420. http://www.jneurosci.Org/content/28/6/1410.fiill.pdf.

Morrison JH, Baxter MG. 2012. The ageing cortical synapse: hallmarks and implications for cognitive decline. Nat Rev Neurosci 13: 240-250.

http://www.ncbi.nlm.nih.gov/pubmed/22395804.

Morrison JH, Hof PR. 1997. Life and death of neurons in the aging brain. Science

278: 412-419.

Nussbaum RL, Ellis CE. 2003. Alzheimer’s Disease and Parkinson’s Disease. N Engl J Med 348: 1356-1364.

http://www.nejm.Org/doi/firll/10.1056/NEJM2003ra020003%5Cnhttp://www.nejm.org /doi/pdff 10.1056/NEJM2003ra020003.

Nussbaum RL, Mclnnes RR, Willard HF. 2015. Thompson & Thompson Genetics in Medicine. 8th ed. Elsevier.

Plassman BL, Langa KM, Fisher GG, Heeringa SG, Weir DR, Ofstedal MB, Burke JR, Hurd MD, Potter GG, Rodgers WL, et al. 2007. Prevalence of dementia in the United States: The aging, demographics, and memory study. Neuroepidemiology 29: 125-132.

Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, Mailer J, Sklar P, de Bakker PIW, Daly MJ, et al. 2007. PLJNK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81 : 559-575.

Saetre P, Jazin E, Emilsson L. 2011. Age-related changes in gene expression are accelerated in Alzheimer’s disease. Synapse 65: 971-4.

http://www.ncbi.nlm.nih.gOv/pubmed/21425351.

Stegle O, Parts L, Piipari M, Winn J, Durbin R. 2012. Using probabilistic estimation of expression residuals (PEER) to obtain increased power and interpretability of gene expression analyses. Nat Protoc 7: 500-7.

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3398141&tool=pmcentre z&rendertype=abstract.

The GTEx Consortium. 2015. The Genotype-Tissue Expression (GTEx) pilot analysis: Multitissue gene regulation in humans. Science (80- ) 348: 648-660.

http://www.ncbi.nlm.nih.gov/pubmed/25954001.

Tsuang D, Leverenz JB, Lopez OL, Hamilton RL, Bennett D a, Schneider J a, Buchman AS, Larson EB, Crane PK, Kaye J a, et al. 2013. APOE e4 increases risk for dementia in pure synucleinopathies. JAMA Neurol 70: 223-8. http://www.pubmedcentral.nih. gov/articlerender.fcgi?artid=3580799&tool=pmcentre z&rendertype=abstract.

Warren J. Strittmatter ADR. 1996. Apolipoprotein E and Alzheimer's disease. Annu Rev Neurosci 19: 53-77.

Williams-Gray CH, Goris A, Saiki M, Foltynie T, Compston DAS, Sawcer SJ, Barker R a. 2009. Apolipoprotein E genotype as a risk factor for susceptibility to and dementia in Parkinson’s disease. J Neurol 256: 493-8. http://www.ncbi.nlm.nih.gov/pubmed/19308307.

Yankner BA, Lu T, Loerch P. 2008. The aging brain. Annu Rev Pathol 3: 41-66. http://www.ncbi.nlm.nih.gOv/pubmed/l 8039130.



Claims

1. A method for determining the Aage of a subject comprising:

a) measuring a transcriptome from a brain sample of the subject;

b) determining gene expressions of one or more age-related genes from the isolated transcriptome;

c) comparing the gene expressions from the subject to corresponding gene expressions in a pre-determined standard;

d) using the comparison to model a molecular age of the subject; and

e) calculating the difference between the molecular age of the subject and a standard molecular age at the chronological age of the subject to obtain Aage of the subject.

2. A method of determining the polygenic risk score (PRS) of a subject comprising: a) performing meta-analysis of genome-wide association to identify one or more single nucleotide polymorphisms (SNPs) in a biological sample that is associated with the Aage calculated in claim 1 ; and

b) using machine learning to create an algorithm that weights these SNPs to create the

PRS.

3. The method of claim 1 or 2, wherein the subject is a mammal.

4. The method of claim 3, wherein the mammal is a human.

5. The method of claim 2, wherein the biological sample is blood, serum, or epithelial cells.

6. The method of claim 1, wherein the one or more age-related genes comprise one or more of the genes shown in Table 1.

7. The method of claim 1, wherein the pre-determined standard is generated from a cohort selected from a Common Mind cohort.

8. The method of claim 1, wherein the modeling comprises using elastic net regression controlling for potential sources of noise.

9. The method of claim 2, wherein the one or more SNPs comprise one or more of those shown in Appendix A - Table 6.

10. A method of determining the propensity of a human subject to be afflicted with neurological disease and/or increased cognitive decline comprising,

a) measuring atranscriptome from a brain sample of the subject;

b) determining gene expressions of one or more age-related genes from the isolated transcriptome;

c) comparing the gene expressions from the subject to corresponding gene expressions in a prc-determined standard;

d) using the comparison to model a molecular age of the subject; and

e) calculating the difference between the molecular age of the subject and a standard molecular age at the chronological age of the subject to obtain Aage of the subject, wherein if the molecular Aage is +5 or more, the probability of neurologicaldisease and/or increased cognitive decline is greater than average .

11. The method of claim 10, where the neurodegenerative diseases comprise Alzheimer’s disease, Paikinson’s disease, frontotemporal dementia, depression, Huntington’s disease, or amyotrophic lateral sclerosis.

12. The method of claim 10, where the neurological disease is Alzheimer’s disease.

13. The method of claim 10 further comprising determining the presence of an APOE allelic variant in a sample from the subject wherein:

i) if the molecular Aage is +5 and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurologicale diseases and/or increased cognitive decline; and

ii) if the molecular Aage is about 0 and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurologicale disease and/or increased cognitive decline.

14. The method of claim 12, w'here the neurological diseases comprise Alzheimer’s disease, Paikinson’s disease, frontotemporal dementia, depression, Huntington’s disease, or amyotrophic lateral sclerosis..

15. The method of claim 12, where the neurological disease is Alzheimer’s disease.

16. The method of claim 10, wherein the subject is a mammal.

17. The method of claim 10 or 13, wherein the mammal is a human.

18. The method of claim 2 or 5, wherein the biological sample is blood, serum, or epithelial cells.

19. A method of treating a symptom associated with brain aging in a subject in need thereof comprising administering to the subject a therapeutically effective amount of a sirtuin activator.

20. The method of claim 19, wherein the sirtuin activator activates SIRT1.

21. The method of claim 16, wherein the subject is a human.

22. A method of determining the polygenic risk score (PRS) of a subject comprising: a) identifying one or more SNPs in a biological sample from the subject that associate with Aage; and

b) calculating a polygenic risk score (PRS) for the subject based on the presence of absence of the one or more SNPs.

23. The method of claim 22, wherein the subject is a mammal.

24. The method of claim 22, wherein the mammal is a human.

25. The method of claim 22, wherein the biological sample is blood, serum, or epithelial cells.

26. The method of claim 22, w'hercin the one or more SNPs comprise one or more of those shown in Appendix A - Table 6.

27. A method of determining the propensity of a subject to be afflicted with

neurodegenerative disease and/or increased cognitive decline comprising,

a) identifying one or more SNPs in a biological sample from the subject that associate with Aage; and

b) calculating a PRS for the subject based on the presence of absence of the one or more SNPs wherein if the molecular PRS is correlated with positive Aage, the probability of neurological disease is greater than average.

28. The method of claim 27, where the neurological diseases comprise Alzheimer’s disease, Parkinson’s disease, frontotemporal dementia, depression, Huntington’s disease, or amyotrophic lateral sclerosis.

29. The method of claim 27, where the neurological disease is Alzheimer’s disease.

30. The method of claim 27 further comprising determining the presence of an APOE allelic variant in a sample from the subject wherein if the PRS is correlated with positive Aage and at least one allele of AROEe4 is present, the individual has greater than average propensity of being afflicted with neurological disease and/or increased cognitive decline.

31. The method of claim 30, where the neurological diseases comprise Alzheimer’s disease, Parkinson’s disease, frontotemporal dementia, depression, Huntington’s disease, or amyotrophic lateral sclerosis.

32. The method of claim 30, where the neurological disease is Alzheimer’s disease.