‘MGWAS’ Ties Brain Metabolites to Alzheimer’s, Parkinson’s, More

Millions of small molecules coursing through our bodies reflect biochemical processes that factor in health and disease. Studies have tied such metabolites in blood and urine to genetic variants. Now, a new line of investigation extends this to the brain. As reported in Nature Genetics on November 11, scientists led by Carlos Cruchaga, Washington University in St. Louis, correlated genetic variants with hundreds of metabolites in CSF and brain samples from more than 3,000 people. In CSF, the concentrations of 139 small molecules linked to variants at 96 loci, while in brain tissue, 31 metabolites mapped to 24 loci.

  • CSF/brain metabolite GWAS tie molecules to genetic variants.
  • Some of these variants associate with diseases and traits.
  • Examples: glycerophosphocholines with AD; O-sulfo-L-tyrosine with PD.

In total, 71 of these genetic loci have been implicated in human disease. For example, ApoE4, a risk factor for AD, FTD, and Type 2 diabetes, associated with phosphocholines in the CSF. The findings underscore that some aspects of metabolism are inherited. They also nominate biochemical pathways that deserve further investigation.

“This gene-to-molecule research is crucial for discovery in Alzheimer’s disease. It offers fresh perspectives on how genetic risk factors, such as APOE, influence metabolism and perhaps disease progression,” wrote Cristina Legido Quigley of King’s College London.

Metabolite GWAS have uncovered genetic connections to hundreds of molecules in thousands of blood or urine samples, helping illuminate the metabolic underpinnings of health and disease (Chen et al., 2023; Surendran et al., 2022; Schlosser et al., 2023). That said, the largest investigation into the genetics of the CSF metabolome included fewer than 300 participants (Panyard et al., 2021). For Cruchaga’s analysis, first author Ciyang Wang and colleagues upped that by an order of magnitude, surveying CSF and brain tissue metabolites from 2,602 and 1,016 people, respectively, integrating CSF and brain tissue samples from six AD research cohorts.

Mega Omics. Omics try to correlate clinical and biological data with thousands, or sometime millions, of molecular variables. Metabolomics must account for more than 200,000 small molecules. [Courtesy of Cristine Legido-Quigley.]

They measured 440 metabolites in CSF; 962 in brain samples. These included amino acids, carbohydrates, lipids, peptides, vitamins and cofactors, nucleotides, and xenobiotics coming from the diet. Next, they scanned donors’ genomes for variants that associated with the abundance of each metabolite.

For CSF metabolites, the scientists identified 205 independent metabolite/variant pairs, linking 139 molecules to variants within 96 loci. Some of these associations were pleiotropic. Some metabolites can be skewed by several independent variants within the same locus, while some loci influence levels of more than one metabolite. For brain tissue, the scientists uncovered 32 associations between individual variants and metabolites, tying 31 molecules to 24 loci.

M-GWAS in the Brain. The abundance of some metabolites in a person’s CSF and brain associated with genetic variants in their genome. [Courtesy of Wang et al., Nature Genetics, 2024.]

As expected, brain and CSF data substantially overlapped, such that 18 metabolite/gene locus pairs matched up. In contrast, the variants influencing metabolites in the brain and CSF were markedly different from those previously linked to blood or urine metabolites. To Cruchaga, this means it’s important to conduct metabolomics in relevant tissues. Notably, 14 loci in this study had never been associated with metabolites in any tissue or fluid before.

For each genetic variant, the scientists’ algorithm considered the location, expression, and function of genes close to it to try to identify the functional gene driving the association. This nominated 118 genes as possibly accounting for the 205 CSF signals, and 25 genes for the 32 brain associations. More than a third encoded proteins directly involved in the production, breakdown, or transport of the associated metabolites. One example? A variant in the CPS1 gene associated with CSF levels of homoarginine, glycine, and serine. CPS1 is the first enzyme in the urea cycle, which produces all of these downstream metabolites.

Many of these genes are chiefly expressed by astrocytes, in keeping with their well-known role as regulators of brain metabolism.

Given the role of metabolism in many diseases and traits, the scientists next cross-referenced their metabolite-linked genetic variants with hits from genome-wide association studies for 23 traits or disorders. These included neurodegenerative diseases such as AD, PD, and FTD, as well as related traits, including Type 2 diabetes, obesity, brain volume, and cognitive performance.

The scientists matched 33 CSF metabolite variants to 12 loci associated with seven traits. For variants linked to brain metabolites, three overlapped with two loci that were implicated in six traits.

Some examples? A standout came from none other than the ApoE gene, in which a polymorphism corresponding to the ApoE4 variant was associated with four different glycerophosphocholines in CSF. The same variant was also linked to increased risk for AD and FTD, as well as with cognitive performance and Type 2 diabetes.

Metabolites Meet Disease. CSF metabolites (left) associate with variants in genetic loci (right). These loci also associate with human traits and diseases (listed on top). Co-localization between trait and metabolite variants is indicated by colored boxes, where orange and blue signal positive associations, respectively, between metabolite levels and traits. Bordered boxes indicate genome-wide significant associations with traits. Nonsynonymous variants and eQTLs are coded by red boxes and pink boxes, respectively. [Courtesy of Wang et al., Nature Genetics, 2024.]

“The association between APOE and glycerophospholipids in CSF is particularly intriguing, especially in the context of the unsaturated fatty acids attached to phosphatidylcholines,” wrote Quigley. Her lab has shown that most fatty acids dip in the brain in people with AD (Whiley et al., 2014). “Understanding the interplay between genetics and lipids in Alzheimer’s, and finding ways to achieve ‘lipid repair,’ is an exciting area of research,” she said.

While his group’s findings do not directly implicate glycerophospholipids in ApoE4-driven AD risk, still, Cruchaga thinks this metabolic pathway could be involved in the disease, or could at least be a disease biomarker reflecting other metabolic malfunctions.

Here’s another example. Wang and colleagues linked low levels of CSF O-sulfo-L-tyrosine with PD risk through variants near the gene encoding arylsulfatase A (ARSA). This enzyme works as an α-synuclein chaperone, implying that its weakening could enable α-synuclein aggregation (Lee et al., 2019).

The researchers also tied CSF metabolites to traits that are risk factors for AD. Genetic variants at multiple loci associated both with metabolites and with a high BMI-adjusted waist-to-hip ratio, a measure of obesity. The correlation suggests that genetic variants that influence metabolism might also genetically predispose to obesity. For example, a variant in the ABCG2 gene, which encodes a transporter of xanthine, associated with less xanthine in CSF. The purine is known to plummet in CSF in people with obesity.

For those wanting to dig into other metabolites, genetic variants, and diseases investigated in the study, the data are available online at https://ontime.wustl.edu/.—Jessica Shugart 

Comments

  1. This gene-to-molecule research is crucial for discovery in Alzheimer’s disease, offering fresh perspectives on how genetic risk factors, such as APOE, influence metabolism and perhaps disease progression.

    The association between APOE and glycerophospholipids in CSF is particularly intriguing, especially in the context of the unsaturated fatty acids attached to phosphatidylcholines reported in the paper. We have shown in human participants that most fatty acid levels in the brain decrease in Alzheimer’s disease (Whiley et al., 2014), and that particular PCs in the blood are also reduced at asymptomatic stage (Snowden et al., 2017). Others have shown that in animal models the 14:0 myristic acid is key to memory formation (Akefe et al., 2024). Understanding the interplay between genetics and lipids in Alzheimer’s and finding ways to achieve "lipid repair" is an exciting area of research.

    Studying omics for metabolism is totally dependent in advances in technology. Each lab develops its own methods and is trying for the most precise measurements of thousands of molecules. The wide range of concentrations of molecules makes this a challenge but also an opportunity for discovery.

     

    View all comments by Cristina Legido Quigley
    • User Profile Image Tobias Madl
      Medical University of Graz / Division Medicinal Chemistry
    • Posted:

    This publication presents a comprehensive genome-wide association study on brain and CSF metabolite levels, and reveals significant metabolic changes associated with various diseases, including Alzheimer’s. The authors have also developed a webserver that allows further exploration of their data on metabolite genome-wide association studies and related molecular traits, providing a valuable resource for the research community.

    Wang et al. combined several datasets of brain metabolites collected from samples of the parietal lobe cortex, dorsolateral prefrontal cortex, and temporal cortex. However, since metabolite profiles can vary significantly between different brain regions, this approach might introduce biases related to the distribution of sample sizes across these areas. Furthermore, by merging datasets that exhibit regional variations in metabolite levels, potentially important changes could be obscured. For example, our study highlighted notable differences in the levels of several amino acids, lactate, taurine, and creatine across the cerebellum, frontal, and occipital cortices (Zhang et al., 2023). It would be interesting for future studies to consider analyzing metabolite profiles from individual brain regions to pinpoint more specific changes. It would also be beneficial to determine which brain tissues most closely reflect CSF metabolite levels.

    One finding from Wang et al. that I found most exciting is the association of the APOE locus with multiple CSF glycerophosphocholines (GPCs) and its co-localization with several conditions, including Type 2 diabetes, Alzheimer’s disease, frontotemporal dementia (FTD), and cognitive performance. This aligns with our previous findings, which indicated an increase in the catabolic metabolite glycerophosphocholine—a breakdown product of glycerophospholipids—in FTD and Alzheimer's disease compared to controls. This has been further corroborated by studies that have shown that glycerophosphocholine levels are elevated in both the CSF and blood of Alzheimer’s patients (Jia et al., 2021; Walter et al., 2004). Glycerophosphocholine, involved in both the synthesis and degradation of cell membranes, reflects changes in the structural integrity of neural cells (Klein 2000).

    Wang et al. not only advances our understanding of the genetic factors influencing metabolite levels in the brain and CSF but also reinforces the role of the APOE locus in neurodegenerative diseases, such as Alzheimer's. It underscores the complexity of metabolite variations across different brain regions and their implications for disease mechanisms. Future research should focus on dissecting these regional differences more deeply and correlating them with CSF metabolite profiles to enhance our understanding of the underlying pathology.

    References:

    Jia L, Yang J, Zhu M, Pang Y, Wang Q, Wei Q, Li Y, Li T, Li F, Wang Q, Li Y, Wei Y. A metabolite panel that differentiates Alzheimer's disease from other dementia types. Alzheimers Dement. 2021 Nov 17; PubMed.

    Klein J. Membrane breakdown in acute and chronic neurodegeneration: focus on choline-containing phospholipids. J Neural Transm (Vienna). 2000;107(8-9):1027-63. PubMed.

    Walter A, Korth U, Hilgert M, Hartmann J, Weichel O, Fassbender K, Schmitt A, Klein J. Glycerophosphocholine is elevated in cerebrospinal fluid of Alzheimer patients. Neurobiol Aging. 2004 Nov-Dec;25(10):1299-303. PubMed.

    Zhang F, Rakhimbekova A, Lashley T, Madl T. Brain regions show different metabolic and protein arginine methylation phenotypes in frontotemporal dementias and Alzheimer's disease. Prog Neurobiol. 2023 Feb;221:102400. Epub 2022 Dec 26 PubMed.

    View all comments by Tobias Madl

Make a Comment

To make a comment you must login or register.

References

Mutations Citations

  1. APOE C130R (ApoE4)

Paper Citations

  1. Chen Y, Lu T, Pettersson-Kymmer U, Stewart ID, Butler-Laporte G, Nakanishi T, Cerani A, Liang KY, Yoshiji S, Willett JD, Su CY, Raina P, Greenwood CM, Farjoun Y, Forgetta V, Langenberg C, Zhou S, Ohlsson C, Richards JB. Genomic atlas of the plasma metabolome prioritizes metabolites implicated in human diseases. Nat Genet. 2023 Jan;55(1):44-53. Epub 2023 Jan 12 PubMed.
  2. Surendran P, Stewart ID, Au Yeung VP, Pietzner M, Raffler J, Wörheide MA, Li C, Smith RF, Wittemans LB, Bomba L, Menni C, Zierer J, Rossi N, Sheridan PA, Watkins NA, Mangino M, Hysi PG, Di Angelantonio E, Falchi M, Spector TD, Soranzo N, Michelotti GA, Arlt W, Lotta LA, Denaxas S, Hemingway H, Gamazon ER, Howson JM, Wood AM, Danesh J, Wareham NJ, Kastenmüller G, Fauman EB, Suhre K, Butterworth AS, Langenberg C. Rare and common genetic determinants of metabolic individuality and their effects on human health. Nat Med. 2022 Nov;28(11):2321-2332. Epub 2022 Nov 10 PubMed.
  3. Schlosser P, Scherer N, Grundner-Culemann F, Monteiro-Martins S, Haug S, Steinbrenner I, Uluvar B, Wuttke M, Cheng Y, Ekici AB, Gyimesi G, Karoly ED, Kotsis F, Mielke J, Gomez MF, Yu B, Grams ME, Coresh J, Boerwinkle E, Köttgen M, Kronenberg F, Meiselbach H, Mohney RP, Akilesh S, GCKD Investigators, Schmidts M, Hediger MA, Schultheiss UT, Eckardt KU, Oefner PJ, Sekula P, Li Y, Köttgen A. Genetic studies of paired metabolomes reveal enzymatic and transport processes at the interface of plasma and urine. Nat Genet. 2023 Jun;55(6):995-1008. Epub 2023 Jun 5 PubMed.
  4. Panyard DJ, Kim KM, Darst BF, Deming YK, Zhong X, Wu Y, Kang H, Carlsson CM, Johnson SC, Asthana S, Engelman CD, Lu Q. Cerebrospinal fluid metabolomics identifies 19 brain-related phenotype associations. Commun Biol. 2021 Jan 12;4(1):63. PubMed.
  5. Whiley L, Sen A, Heaton J, Proitsi P, García-Gómez D, Leung R, Smith N, Thambisetty M, Kloszewska I, Mecocci P, Soininen H, Tsolaki M, Vellas B, Lovestone S, Legido-Quigley C, . Evidence of altered phosphatidylcholine metabolism in Alzheimer's disease. Neurobiol Aging. 2014 Feb;35(2):271-8. PubMed.
  6. Lee JS, Kanai K, Suzuki M, Kim WS, Yoo HS, Fu Y, Kim DK, Jung BC, Choi M, Oh KW, Li Y, Nakatani M, Nakazato T, Sekimoto S, Funayama M, Yoshino H, Kubo SI, Nishioka K, Sakai R, Ueyama M, Mochizuki H, Lee HJ, Sardi SP, Halliday GM, Nagai Y, Lee PH, Hattori N, Lee SJ. Arylsulfatase A, a genetic modifier of Parkinson's disease, is an α-synuclein chaperone. Brain. 2019 Sep 1;142(9):2845-2859. PubMed.

External Citations

  1. https://ontime.wustl.edu/

Further Reading

No Available Further Reading

Primary Papers

  1. Wang C, Yang C, Western D, Ali M, Wang Y, Phuah CL, Budde J, Wang L, Gorijala P, Timsina J, Ruiz A, Pastor P, Fernandez MV, Dominantly Inherited Alzheimer Network (DIAN), Alzheimer’s Disease Neuroimaging Initiative (ADNI), Panyard DJ, Engelman CD, Deming Y, Boada M, Cano A, Garcia-Gonzalez P, Graff-Radford NR, Mori H, Lee JH, Perrin RJ, Ibanez L, Sung YJ, Cruchaga C. Genetic architecture of cerebrospinal fluid and brain metabolite levels and the genetic colocalization of metabolites with human traits. Nat Genet. 2024 Dec;56(12):2685-2695. Epub 2024 Nov 11 PubMed.

Annotate

To make an annotation you must Login or Register.