Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Genetic interaction study of Alzheimer's disease quantitative biomarkers : A polygenic risk score analysis and evaluation

Kumar, Atul LU orcid ; Janelidze, Shorena LU ; Shoai, Maryam ; Palmqvist, Sebastian LU orcid ; Stomrud, Erik LU orcid ; Hardy, John ; Hansson, Oskar LU orcid and Mattsson-Carlgren, Niklas LU orcid (2021) In Alzheimer's & dementia : the journal of the Alzheimer's Association 17.
Abstract

BACKGROUND: Relationship between genetic factors and pathological features of Alzheimer's disease (AD) may be studied with biomarker, using both polygenic risk scores (PRSs), as well as individual genetic variants in genome-wide association studies (GWAS). METHOD: In the Swedish BioFINDER study, we used a priori PRS for AD based on findings in recent GWAS, and AD related biomarkers in cerebrospinal fluid (CSF) in cognitively unimpaired individuals (n = 751), Mild Cognitive Impairment (n = 212), and AD (n = 150) patients. AD related biomarkers were rank-based inverse normal transformed to be used as dependent variables in linear regression models adjusted for age, gender, education, APOE ε4, ε2 count and significant principal components.... (More)

BACKGROUND: Relationship between genetic factors and pathological features of Alzheimer's disease (AD) may be studied with biomarker, using both polygenic risk scores (PRSs), as well as individual genetic variants in genome-wide association studies (GWAS). METHOD: In the Swedish BioFINDER study, we used a priori PRS for AD based on findings in recent GWAS, and AD related biomarkers in cerebrospinal fluid (CSF) in cognitively unimpaired individuals (n = 751), Mild Cognitive Impairment (n = 212), and AD (n = 150) patients. AD related biomarkers were rank-based inverse normal transformed to be used as dependent variables in linear regression models adjusted for age, gender, education, APOE ε4, ε2 count and significant principal components. We also tested individual genetic variants in GWAS for each biomarker. Analyses were performed in the total sample, and after stratification on MMSE results. RESULT: The PRS was associated with higher CSF P-tau 181 (p ≤1.2e-05) and T-tau (p ≤ 8.14e-05) and lower CSF Aβ42/40 (p ≤ 0.006) and Aβ42 (p ≤ 0.04) (Figure 1, 2). Gene Enrichment of PRS 5 genes [containing 1850 genetic variants mapped to 1607 genes] for Tau biomarker showed 13 Gene Ontology (GO) Biological Process (BP) terms at p-value < e-03 ("Dendrite Morphogenesis": top hit; p-value ≤ 9.20e-06) and 16 KEGG pathway term enriched for genes of PRS 5 ("Phosphatidylinositol signaling system": top hit; p-value ≤ 5.5e-06) (Figure 3, 4). Gene enrichment of PRS 7 [containing 62 genetic variants mapped to 58 genes] for Aβ biomarker returned 12 GO terms ("Integrin-Mediated Signaling Pathway": top hit; p-value ≤1.20e-03) and 1 term enriched for KEGG pathway (Hematopoietic cell lineage). In our predefined list of genes interacting with MAPT (22 genes) and APP (69 genes) we found 3 genes from MAPT and APP set that were involved in PRS 5 and PRS 7 respectively. We also found 9 genes from APP set that was involved in PRS 5. CONCLUSION: Elevated levels of AD related biomarkers are associated with polygenic risk scores in AD. These findings further strengthens the link between genetic and biomarker disease predictors and indicate a potential role for these markers in disease prediction and patient stratification in AD.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Alzheimer's & dementia : the journal of the Alzheimer's Association
volume
17
article number
e053556
publisher
Wiley
external identifiers
  • scopus:85124058578
  • pmid:35109036
ISSN
1552-5279
DOI
10.1002/alz.053556
language
English
LU publication?
yes
id
df9712f0-2605-4542-8a0a-09b1991eee97
date added to LUP
2022-04-05 14:50:14
date last changed
2023-12-20 03:02:32
@article{df9712f0-2605-4542-8a0a-09b1991eee97,
  abstract     = {{<p>BACKGROUND: Relationship between genetic factors and pathological features of Alzheimer's disease (AD) may be studied with biomarker, using both polygenic risk scores (PRSs), as well as individual genetic variants in genome-wide association studies (GWAS). METHOD: In the Swedish BioFINDER study, we used a priori PRS for AD based on findings in recent GWAS, and AD related biomarkers in cerebrospinal fluid (CSF) in cognitively unimpaired individuals (n = 751), Mild Cognitive Impairment (n = 212), and AD (n = 150) patients. AD related biomarkers were rank-based inverse normal transformed to be used as dependent variables in linear regression models adjusted for age, gender, education, APOE ε4, ε2 count and significant principal components. We also tested individual genetic variants in GWAS for each biomarker. Analyses were performed in the total sample, and after stratification on MMSE results. RESULT: The PRS was associated with higher CSF P-tau 181 (p ≤1.2e-05) and T-tau (p ≤ 8.14e-05) and lower CSF Aβ42/40 (p ≤ 0.006) and Aβ42 (p ≤ 0.04) (Figure 1, 2). Gene Enrichment of PRS 5 genes [containing 1850 genetic variants mapped to 1607 genes] for Tau biomarker showed 13 Gene Ontology (GO) Biological Process (BP) terms at p-value &lt; e-03 ("Dendrite Morphogenesis": top hit; p-value ≤ 9.20e-06) and 16 KEGG pathway term enriched for genes of PRS 5 ("Phosphatidylinositol signaling system": top hit; p-value ≤ 5.5e-06) (Figure 3, 4). Gene enrichment of PRS 7 [containing 62 genetic variants mapped to 58 genes] for Aβ biomarker returned 12 GO terms ("Integrin-Mediated Signaling Pathway": top hit; p-value ≤1.20e-03) and 1 term enriched for KEGG pathway (Hematopoietic cell lineage). In our predefined list of genes interacting with MAPT (22 genes) and APP (69 genes) we found 3 genes from MAPT and APP set that were involved in PRS 5 and PRS 7 respectively. We also found 9 genes from APP set that was involved in PRS 5. CONCLUSION: Elevated levels of AD related biomarkers are associated with polygenic risk scores in AD. These findings further strengthens the link between genetic and biomarker disease predictors and indicate a potential role for these markers in disease prediction and patient stratification in AD.</p>}},
  author       = {{Kumar, Atul and Janelidze, Shorena and Shoai, Maryam and Palmqvist, Sebastian and Stomrud, Erik and Hardy, John and Hansson, Oskar and Mattsson-Carlgren, Niklas}},
  issn         = {{1552-5279}},
  language     = {{eng}},
  publisher    = {{Wiley}},
  series       = {{Alzheimer's & dementia : the journal of the Alzheimer's Association}},
  title        = {{Genetic interaction study of Alzheimer's disease quantitative biomarkers : A polygenic risk score analysis and evaluation}},
  url          = {{http://dx.doi.org/10.1002/alz.053556}},
  doi          = {{10.1002/alz.053556}},
  volume       = {{17}},
  year         = {{2021}},
}