Search

Background: APOEε4 significantly increases the risk of developing Alzheimer's disease (AD). Cognitively healthy APOEε4‐carriers exist, suggesting potential protective mechanisms against APOEε4. We hypothesized that some APOEε4‐carriers may have genetic variations protecting them from developing APOEε4‐mediated AD pathology. We aim to identify these protective genetic variants. Methods: Whole genome sequencing (WGS) and cerebrospinal fluid (CSF) proteomics were performed from human cohorts to identify potential protective variants segregating exclusively among APOEε4 carriers. Bioinformatic analyses were performed to select candidate target genes. Immunohistochemistry on postmortem human brain tissues and mouse models expressing human APOE variants were performed along with in‐vivo functional studies in adult zebrafish AD model. Result: WGS analyses revealed 510 potential gene variants segregating exclusively among APOEε4 carriers, which included rare and loss‐of‐function (LOF) variants. Pathway analysis of these genes showed significant enrichment in extracellular matrix (ECM)‐related processes, suggesting protective effects of LOF in ECM proteins. This was further validated by CSF proteome profiling and subsequent analyses in APOEε4 carriers and non‐carriers. Fibronectin‐1 (FN1) and Collagen‐6A2 (COL6A2) were prioritized as candidate target genes for postmortem validation and in‐vivo functional studies. FN1 protein was increased in APOEε4 carriers resulting in thickened ECM at the basement membrane around the blood vessels, potentially impairing pathology‐induced responses such as clearance and immune system activity. This observation is validated in human brains, mouse models and zebrafish model; therefore, the pathological association of FN1 to AD is evolutionarily conserved. Supporting this hypothesis, in‐vivo functional study in zebrafish model with LOF mutations in fn1b revealed that fibronectin LOF enhanced gliovascular remodeling and microglial activation while reducing astrogliosis, suggesting that pathological accumulation of FN1 could impair toxic protein clearance, which is ameliorated with FN1 LOF. Conclusion: The vascular deposition of the ECM components FN1 and COL6A2 are increased in APOEε4 carriers. Rare variant in FN1 protect against APOEε4‐mediated pathogenesis in AD. We propose a new disease mechanisms and potential therapeutic intervention targets for vascular contribution to dementia to mitigate the risk of developing AD. [ABSTRACT FROM AUTHOR]

BACKGROUND: Few rare variants have been identified in genetic loci from genome‐wide association studies (GWAS) of Alzheimer's disease (AD), limiting understanding of mechanisms, risk assessment, and genetic counseling. METHODS: Using genome sequencing data from 197 families in the National Institute on Aging Alzheimer's Disease Family Based Study and 214 Caribbean Hispanic families, we searched for rare coding variants within known GWAS loci from the largest published study. RESULTS: Eighty‐six rare missense or loss‐of‐function (LoF) variants completely segregated in 17.5% of families, but in 91 (22.1%) families Apolipoprotein E (APOE)‐휀4 was the only variant segregating. However, in 60.3% of families, APOE 휀4, missense, and LoF variants were not found within the GWAS loci. DISCUSSION: Although APOE 휀4and several rare variants were found to segregate in both family datasets, many families had no variant accounting for their disease. This suggests that familial AD may be the result of unidentified rare variants. Highlights: Rare coding variants from GWAS loci segregate in familial Alzheimer's disease.Missense or loss of function variants were found segregating in nearly 7% of families.APOE‐휀4 was the only segregating variant in 29.7% in familial Alzheimer's disease.In Hispanic and non‐Hispanic families, different variants were found in segregating genes.No coding variants were found segregating in many Hispanic and non‐Hispanic families. [ABSTRACT FROM AUTHOR]

Background: Cardio and cerebrovascular risk factors (CVRFs) increase the risk of cerebrovascular disease and clinical Alzheimer's Disease (AD), and over 70% of the patients with AD coincident cerebrovascular pathology. We previously found that FMNL2 interacts with a burden score of hypertension, diabetes, heart disease, and body mass index (BMI) by altering the normal astroglial‐vascular mechanisms that underly amyloid clearance. Stroke, defined by history of a clinical stroke or brain imaging, is a moderately robust risk factor for AD and dementia. The goal here was to identify genes that interact with CVRFs, incorporating stroke as an additional factor, on AD in multi‐ethnic cohorts. Method: We conducted a genome‐wide gene‐CVRF score interaction analysis for AD, in 7,939 AD patients and 9,631 controls from eight multi‐ethnic cohorts of non‐Hispanic Whites, African Americans, and Hispanics including ADNI, NACC, NOMAS, WHICAP, EFIGA, and ROSMAP. A CVRF score was created from the first principal component of history of clinical stroke, hypertension, diabetes, and heart disease, and measured BMI. Gene‐based interaction test was performed with the adaptive gene‐environment interaction test. Results were summarized using a meta‐analysis. We investigated the association of pathological AD, amyloid‐β, or brain infarcts with gene expression and protein expression from the frontal cortex in ROSMAP using a generalized linear model. Age, sex, and the first three principal components were adjusted in the models. Result: The interaction of CVRF score with FMNL2 on AD (p = 1.02E‐05) was identified and additional genes were identified to interact with CVRF score, including SLC22A14 (p = 1.44E‐06), AMMECR1L (p = 2.74E‐06), PRG3 (p = 2.76E‐06), CFAP99 (p = 5.22E‐06), ADPGK‐AS1 (p = 8.58E‐06) and BRINP1 (p = 6.29E‐06). ADPGK‐AS1 and FMNL2 gene expressions were associated with pathological AD (p = 0.004 and p = 0.0002). FMNL2 and BRINP1 gene expressions were higher in the brains of patients with brain infarcts (p = 0.025 and p = 0.006). BRINP1 protein expression was associated with pathological AD (p = 0.0002) and was higher in the brains of patients with brain infarcts (p = 0.022). Conclusion: We identified novel candidate genes that interact with CVRFs on AD in multi‐ethnic cohorts. Understanding the interplay between genes, CVRFs, and AD has the potential to reveal novel molecular targets for prevention and treatment for AD. [ABSTRACT FROM AUTHOR]

Background: Genetic Studies of Alzheimer's Disease in Caribbean Hispanics (CH) has been ongoing for over 20 years. Our primary goal was to identify genes in large well‐phenotyped, multiplex families, and sporadic cases and controls. This longitudinal collection of data represents a unique resource for the scientific community. The project is led by collaborators in the United States (US) and the Dominican Republic (DR). For the past three years, our efforts have been centered on the collection of cerebrospinal fluid (CSF), plasma, serum and Paxgene tubes for RNA to study core biomarkers. Collection of blood samples for DNA to conduct genetic analyses is also ongoing. Method: The cohort consisted of participants aged 50 and older from the DR, including those living in the New York City area, willing and able to provide a blood and CSF samples. Interviewers in the DR were trained to collect data, blood samples and CSF. DNA samples were sent overnight to Columbia university (CU) for processing. CSF, Paxgene tubes, serum and plasma samples were processed in the DR and stored at ‐80F for batch shipping to CU. APOE genotyping was conducted in the US. Biomarkers, metabolomics, and proteomics data was generated at Columbia University. Methylation assays from Paxgene was conducted at Uniformed Services University (USUHS). Result: We have databased a total of 1899 individuals with plasma samples, 1707 with Paxgene, 1731 with serum and 203 with CSF. Blood samples from 1079 participants have genome wide arra, whole genome sequencing and APOE genotyping. 192 CSF samples have proteomics completed, 153 CSF samples have metabolomics data and 742 have plasma biomarkers (Ab40, Ab42, T‐tau, P‐tau181, NFL and GFAP)). Conclusion: Sample collection for genetic studies in developing countries present important logistical challenges. Despite these challenges, sample quality is comparable to those collected in New York. Biomarkers from CSF, serum and plasma can give us a closer look into underrepresented populations such as Caribbean Hispanics. Results suggest that similar protocols can be successful in Latin America. Future work will augment the biosample collection of the cohort by collecting samples from additional subjects. [ABSTRACT FROM AUTHOR]

Background: Genetic Studies of Alzheimer's Disease in Caribbean Hispanics (CH) has been ongoing for over 20 years. Our primary goal was to identify genes in large well‐phenotyped, multiplex families, and sporadic cases and controls. This longitudinal collection of data represents a unique resource for the scientific community. The project is led by collaborators in the United States (US) and the Dominican Republic (DR). For the past three years, our efforts have been centered on the collection of cerebrospinal fluid (CSF), plasma, serum and Paxgene tubes for RNA to study core biomarkers. Collection of blood samples for DNA to conduct genetic analyses is also ongoing. Method: The cohort consisted of participants aged 50 and older from the DR, including those living in the New York City area, willing and able to provide a blood and CSF samples. Interviewers in the DR were trained to collect data, blood samples and CSF. DNA samples were sent overnight to Columbia university (CU) for processing. CSF, Paxgene tubes, serum and plasma samples were processed in the DR and stored at ‐80F for batch shipping to CU. APOE genotyping was conducted in the US. Biomarkers, metabolomics, and proteomics data was generated at Columbia University. Methylation assays from Paxgene was conducted at Uniformed Services University (USUHS). Result: We have databased a total of 1899 individuals with plasma samples, 1707 with Paxgene, 1731 with serum and 203 with CSF. Blood samples from 1079 participants have genome wide arra, whole genome sequencing and APOE genotyping. 192 CSF samples have proteomics completed, 153 CSF samples have metabolomics data and 742 have plasma biomarkers (Ab40, Ab42, T‐tau, P‐tau181, NFL and GFAP)). Conclusion: Sample collection for genetic studies in developing countries present important logistical challenges. Despite these challenges, sample quality is comparable to those collected in New York. Biomarkers from CSF, serum and plasma can give us a closer look into underrepresented populations such as Caribbean Hispanics. Results suggest that similar protocols can be successful in Latin America. Future work will augment the biosample collection of the cohort by collecting samples from additional subjects. [ABSTRACT FROM AUTHOR]

Background: Cardio and cerebrovascular diseases co‐exist in up to a third of the Alzheimer's disease (AD) patients, and the presence of cardiovascular/cerebrovascular risk factors (CVRFs) are associated with the risk of AD in middle age and later. The relationship between CVRFs such as hypertension, body mass index, diabetes and coronary heart disease, and AD is well known, but there has been limited mechanistic evidence directly linking these vascular risk factors in AD to the presence of ischemic microvascular pathology. Each of these vascular factors has the capacity to impair the blood‐brain barrier and glio‐vascular units. Arterial pulses and flow are required for glymphatic clearance of molecules including amyloid β. Few studies have directly linked genetic variation related to the risk factors and the molecular mechanisms underlying the interplay of CVRF and genetics in AD. APOE plays a role in the metabolism of cholesterol and other lipids in the brain. Other genetic variants that have been found to interact with cardio and cerebrovascular risk factors include the CLU, PICALM, CR1, BIN1, ACE, AGT, and ALDH2. Methods: We analyzed several large multi‐ethnic AD cohorts to understand the role of cardiovascular risk in the disease. We created a CVRF score comprised of hypertension, diabetes, heart disease and body‐mass‐index and tested the interaction of the score with genetic variants in conferring risk of AD. Results: We found that genetic variants in FMNL2, BRINP1, RFC2, and IGFN1 interact with enhanced risk score for CVRFs, to modify the risk of developing AD. FMNL2 and BRINP1 brain expression is increased in patients that have pathological hallmark of AD and have chronic infarcts. Causal mediation analysis found that both amyloid and tau pathology increased FMNL2 expression which in turn resulted in higher risk of AD, suggesting that AD brains enhance the activity of this gene. Conclusions: We propose that that FMNL2 is part of a larger set of genetic and molecular network that underlies the interaction between CVRF and AD. Clinical studies and basic science with animal models could be harmonized to discover new disease mechanisms, which could open new clinical directions and drug development efforts. [ABSTRACT FROM AUTHOR]

Background: Cardiovascular risk factors (CVRFs) increase the risk of cerebrovascular disease and AD, and over 30% of the patients with AD coincident cerebrovascular pathology. We previously found that FMNL2 interacts with CVRFs (p = 6.6e‐07) by altering the normal astroglial‐vascular mechanisms that underly amyloid clearance. The goal here was to identify additional genes that contribute to the interaction between CVRFs and AD. Method: A collection of eight multi‐ethnic cohorts was explored to study genome‐wide gene‐CVRF score interaction analysis for AD, in 7,441 AD patients and 10,453 controls. The cardiovascular risk factors score (CVRF score) was created from a first principal component of the four vascular risk factors; self‐reported history of hypertension, diabetes, and heart disease, and measured body mass index. Gene‐based interaction test was performed using the adaptive gene‐environment interaction (aGE) test. Results were summarized using a meta‐analysis. We investigated the association of gene expression with pathological AD and AD phenotype (amyloid‐β, tau, or brain infarcts) in 1,092 samples from the frontal cortex in ROSMAP. The pathway enrichment analysis was performed on the differentially expressed genes. Age and sex were adjusted in the models. Result: The previous interaction of CVRF score with FMNL2 (p = 3.64e‐07) was strengthened and additional genes were identified including BRINP1 (p = 2.45e‐06), CFAP99 (p = 3.3e‐06), and PRG3 (p = 3.71e‐06). FMNL2 encodes a formin‐related protein important in regulating actin and microtubules. FMNL2 and BRINP1 expressions were higher in the brains of patients with brain infarcts (p = 0.025 and p = 0.006, respectively). The expression level of BRINP1 interacted with brain infarcts on pathological AD (p = 0.02) and with amyloid‐β or tau on brain infarcts (p = 0.02 and p = 0.01, respectively). The pathways were relevant to cellular interactions between astrocytes and endothelia as well as the immune system reaction in the glio‐vascular niche. Conclusion: The four novel genes are likely to be involved in the complex interaction between Alzheimer's disease pathologies (amyloid and phosphorylated tau deposition) and cerebrovascular pathology, for example, at the glia‐vascular interface during AD progression. Understanding how these genes interfere with the mechanisms underlying the clearance of amyloid and tau increasing their deposition in brain will be essential. [ABSTRACT FROM AUTHOR]

Background: Energy, amino acid, and lipid metabolism are dysregulated in Alzheimer's Disease. We investigated circulating plasma metabolites to capture systemic biochemical changes associated with Alzheimer's disease (AD) using the clinical diagnosis and followed by the addition of plasma‐based biomarkers. Method: Exogenous and endogenous metabolites in plasma were measured in 300 individuals with a clinical diagnosis of AD and 430 healthy individuals without dementia of Caribbean Hispanic ancestry using untargeted liquid‐chromatography, performed on HILIC (+ ionization) and C18 (‐ ionization) columns, coupled to a Thermo Orbitrap HF‐X mass spectrometer. Genome‐wide SNP data and plasma biomarkers including pTau181, Aβ40, Aβ42, neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) were obtained on all participants. Metabolite association with the clinical diagnosis of AD and for the AD biomarker, pTau‐181, levels were assessed after adjustment for age and sex. Result: More than 6000 metabolic features were measured with high accuracy in the sample. 270 metabolic features were associated with pTau‐181 levels and 66 metabolic features were associated with clinical AD status, after correction for multiple testing. Valyl‐Serine (p = 1.1e‐7) and creatinine (p = 9.6e‐7) were strongly associated pTau‐181 levels. When using a diagnostic cutoff of pTau181 levels for AD as the outcome (biological AD), we also observed significant associations with a lysophospholipid (p = 9.5e‐5). Malic acid (p = 4.2e‐4) and tryptophan (p = 3.9e‐4) were significantly associated with clinical AD status. Metabolites associated with pTau‐181 levels were enriched in glycan biosynthesis, energy, amino acid, urea cycle/amino group and carbohydrate metabolism pathways. Patients with both clinical and biological AD showed altered tyrosine and tryptophan metabolism. Conclusion: Metabolites, specifically amino acids and lipids, were associated with AD status and pTau181 levels. Metabolite profiling can identify perturbed pathways in clinical and pre‐clinical AD and integration with genome‐wide SNP data will identify mechanisms underlying genetic association with disease. [ABSTRACT FROM AUTHOR]