Jensen, Samantha
Alzheimer’s Risk Genes and Pathways of Expression
Faculty Mentor: John Kauwe, Biology
Although Alzheimer’s disease (AD) has been the subject of research for nearly 100
years, these decades of research have not led to a unified understanding of the
mechanisms behind the disease pathology1 ,2. Recent improvements in whole genome
sequencing have made it possible for researchers to find a large number of mutations
implicated in AD. Due to a high frequency of false positives and confounding factors,
many mutations being investigated for therapies may not even have a causative effect
on Alzheimer’s expression. Without concrete knowledge of the interactions of the
proteins and genes implicated in the disease, drug development has been
unsuccessful.
However, these previous genetic discoveries can help to illuminate the pathways
through which AD works. Using expression data, whole genome sequencing, and
association and overrepresentation analyses, we were able to show pathways of
expression that may be promising for further study.
Methodology
A complete list of AD associated single nucleotide polymorphisms (SNPs) were taken
from a Dr. Perry Ridge’s recent review paper3. The locations of these SNPs were used
to find areas of interest and any SNP in these areas found in an individual in the
Alzheimer’s Disease Genetics Consortium (ADGC) dataset was compared against that
individual’s case-control status for AD in order to generate a list of loci that may have a
significant effect on disease risk.
These genetic mutations determined to be related to AD were then compared in the
Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset against each expression
locus in order to find places where expression was significantly changed in individuals
with the mutated allele. The genes containing affected loci were then used in an
overrepresentation analysis using the online toolkit Reactome4 in order to find biological
pathways that were most affected by AD SNPs.
Results
We found nine genes through our eQTL analysis where loci were most often
differentially expressed (see Table 1). These genes indicate locations where possession
of an AD SNP significantly changed levels of expression.
The pathways most commonly affected by known AD variants were cell-cell
communication (Figure 1A), lipid metabolism (Figure 1B), vesicle-mediated transport
(Figure 1C), and transmembrane transport of small molecules (Figure 1D).
Results
Alzheimer’s disease progresses as lipid plaques and tangles build up in the brain, so
our finding that pathways involving lipid metabolism and transport seem to be promising
mechanisms for AD pathology. We also saw some evidence that other AD implicated
mechanisms, including the immune system (see Figure 1) were affected by AD SNPs.
Although these findings seem to be in line with what we know about AD, this project
involved many complex calculations and needs replication in other datasets to insure
accuracy and generalizability.
Conclusion
Using the gene-gene and gene-protein interactions found (see Table 1), specific
experimentation on the interactions between these genes and proteins can illuminate
possible drug targets. Understanding the full pathways – from DNA to protein
interactions – involved in AD allows us to focus therapies on the proteins that may have
therapeutic value. Future research plans include determination of causal variants in
differential expression and experimental verification of the effect of certain variation on
protein expression.
1 Alzheimer, A. About a peculiar disease of the cerebral cortex. Alzheimer Dis Assoc Disord 1, 3–8 (1907).
2 Khachaturian, Z. S. Diagnosis of Alzheimer’s disease: two-decades of progress. J. Alzheimers Dis. 9, 409–415
(2006).
3 Ridge, P. G. et al. Assessment of the genetic variance of late-onset Alzheimer’s disease. Neurobiology of Aging
41, 200.e13–200.e20 (2016)
4 “Reactome Pathway Browser.” N.p., n.d. Web. 28 Dec. 2016. http://www.reactome.org/PathwayBrowser/