By Monica Heger
Next-generation sequencing has been successfully used to uncover the genetic basis of a number of Mendelian diseases, but parsing the genetics of more complex diseases like Alzheimer's remains challenging.
In an attempt to cut through this complexity, researchers from Washington University's School of Medicine are taking a two-part strategy.
First, as described in a study published this month in PLoS One, the team sequenced five candidate genes that have been linked to early-onset Alzheimer's and frontal temporal dementia to see if variants in these genes may also play a role in late-onset Alzheimer's, which is often classified as a separate illness. The team sequenced these genes in 439 individuals with late-onset Alzheimer's from families with a strong history of the disease.
Secondly, the team now plans to follow up with a broader sequencing strategy — either whole-exome or whole-genome sequencing — to uncover novel genes that may be involved in the disease. This study will involve patients for whom the targeted sequencing approach did not detect any pathogenic variants.
The goal of this effort, said senior author Alison Goate, professor of genetics in psychiatry at Wash U's School of Medicine, is to identify novel genes that are associated with Alzheimer's disease.
Sequencing for the study published in PLoS One was done within the school of medicine's Genome Technology Access Core on the Illumina platform. At the time, most of the sequencing was done on the Genome Analyzer, said Goate, but the core is now equipped with HiSeqs.
In the targeted sequencing study, the team sequenced APP, PSEN1, PSEN2, MAPT, and GRN, which have previously been linked to familial early-onset forms of Alzheimer's and familial frontal temporal dementia. They also checked the status of APOE4 — considered to be the strongest known genetic risk factor for sporadic and familial late-onset Alzheimer's — for all the samples.
Goate told Clinical Sequencing News that the team is following up on these samples by testing all of them for a repeat expansion that is also known to be associated with frontal temporal dementia, and then either exome or whole-genome sequencing on those patients that have neither the expansion nor a suspected pathogenic variant in any of the candidate genes.
As outlined in the PLoS One paper, the targeted approach identified 33 nonsense, missense, and splice-site sequencing variants in 60 of the 439 individuals, including five known pathogenic variants in 10 individuals, and three novel, potentially pathogenic alleles in four individuals. Pathogenic or likely pathogenic variants were found in all genes except PSEN2, although rare variants of unknown significance were found in PSEN2.
The fact that variants associated with early-onset Alzheimer's were found in late-onset patients indicates that, at least in some cases, both early- and late-onset disease have the same causes, the researchers said. "People who get the disease at younger ages probably have more risk factors and fewer protective ones, while those who develop the disease later in life may have more protective factors, but it appears the mechanism may be the same for both," Goate said in a statement.
Additionally, while the genes APP, PSEN1, and PSEN2 have been associated with early-onset Alzheimer's disease, MAPT and GRN are typically associated with familial frontal temporal dementia. Goate said that just as many patients had mutations to these genes as in the other ones, suggesting the patients may have been misdiagnosed.
The finding illustrates the importance of having a molecular diagnosis, said Goate, because in the future, as targeted therapies become available, a genetic diagnosis could change treatment options for these patients. Even though the phenotypes are similar, she said, "the mechanisms are likely to be different."
Another interesting finding from the study is that patients with the APOE4 allele were no less likely to have rare variants in the five sequenced genes than patients without the allele, suggesting that even in patients with that risk allele there are other contributing factors that impact disease risk, progression, and age of onset, said Goate.
Michal Janitz, a senior lecturer in the School of Biotechnology and Biomolecular Sciences at the University of New South Wales who uses transcriptome sequencing of post-mortem brain tissue samples, said that he has found similar results.
"This is quite interesting," he said, because the APOE4 allele is "considered to be the strongest genetic factor for developing early-onset Alzheimer's."
While the study points to additional genetic factors for familial Alzheimer's disease, Janitz noted that one problem with translating the findings to the clinic is that the variants only explain a very small portion of the disease and do not address sporadic forms of Alzheimer's, which make up the majority of Alzheimer's cases.
"If we think about whether we can use this for genetic testing," he said, it would be "limited to patients with a strong history of Alzheimer's disease, which is only a few percentage of all cases." However, he added, it's "still significant if you calculate the real numbers of individuals."
Additionally, while the team identified rare variants in genes that have previously been linked to early-onset Alzheimer's, the functional consequences of the variants are mostly unknown.
Goate said that the team is following up on these variants to try to determine their pathogenicity and how they impact disease.
Even though the researchers found more missense variants than would be expected from healthy individuals, "it's hard for us to say which ones are causing an increase in risk and which ones are benign," Goate said.
Allen Roses, a professor of neurobiology and neurology at Duke University and the director of the Deane Drug Discovery Institute at Duke who discovered the role of the APOE4 allele in Alzheimer's disease, said that the lack of correlation with disease is the study's biggest flaw.
One problem, he said, is that the study evaluated families and there are many benign polymorphisms that are private to families and not found in the general population. So finding a rare variant, even though it is in a gene associated with Alzheimer's, does not necessarily mean that that variant has anything to do with the disease, he said.
A useful control, he said, would have been to look at genes that had nothing to do with Alzheimer's to see how many rare variants clustered by family in those genes.
Janitz agreed that the study was not able to conclude whether the variants were disease-causing or not, which he said is a problem with doing targeted sequencing studies in complex diseases, particularly in diseases where there is "a complex interplay between genes and the environment."
For instance, he said, while it is clear that the aggregation of amyloid protein plays a role in the disease, it's not clear whether mutations to the APP gene cause that aggregation.
Additionally, he said that in RNA-seq studies he has conducted that have evaluated not only familial cases of Alzheimer's disease but also sporadic ones, the five genes the authors sequenced were not at the top of the list of the genes that his team has found to be differentially expressed.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.