NEW YORK (GenomeWeb) – Using a bioinformatics algorithm to catalog structural variants in a non-coding region of the SNCA gene and phased sequencing, Duke University researchers reported they've discovered a haplotype that appears to increase a person's risk of developing Lewy body-type Alzheimer's.
In a paper published online in June in Alzheimer's & Dementia, researchers led by Ornit Chiba-Falek of Duke University Medical Center described how they found the so-called haplotype 3, located in a cytosine-thymine (CT)-rich region of SNCA intron 4, which "acts as an enhancer element" and is associated with heightened risk of Lewy body AD. This haplotype, the researchers concluded, is likely involved in causing Lewy bodies by upregulating SNCA expression.
Up to 20 percent of AD patients with dementia have Lewy bodies — distinctive masses inside nerve cells primarily comprising the alpha-synuclein protein encoded by the SNCA gene. These abnormalities are found in Parkinson's patients who develop cognitive difficulties. Autopsies have also found Lewy bodies in brain tissues of patients who had dementia, but not AD, because they had more psychiatric symptoms and not as much memory loss.
Very dense Lewy bodies are found in approximately 1 percent of AD pathology. Alzheimer's experts have "long wondered whether this is part of the functional pathology or a separate overlapping independent pathology," Allen Roses, Duke neurology professor and an author on the latest paper, told GenomeWeb. "It appears from our work with Lewy bodies in AD and Lewy bodies in Parkinson's, that there may be separate structural variants involved."
Although studies have shown that the SNCA gene is associated with neurodegenerative disorders that have Lewy bodies present, researchers have yet to pinpoint variants responsible for this phenotype. The authors of the Alzheimer's & Dementia paper believe their analysis demonstrates how important analysis of structural variants is to increasing understanding of complex neurological conditions.
Largely due to limitations of high-throughput platforms, researchers have focused on identifying SNPs associated with diseases. But studies have found structural variants, particularly common in introns, to be significantly associated with complex neurological disorders, such as SNCA Rep1 in Parkinson's and a G4C2 repeat in Lou Gehrig's disease and frontotemporal dementia. Moreover, citing that the ENCODE project highlighted the importance of non-coding regions in gene regulation, the authors of the Alzheimer's & Dementia paper contended that non-SNP structural variants may provide more precise insights than SNPs do into the biological processes underlying complex diseases and their symptoms.
One of the limitations currently is that NGS techniques can assess the position of structural variants (ie. insertion/deletions, block substitutions, inversions of DNA sequences, and copy number differences) but aren't very accurate when sequencing long repeat polymorphisms, according to Roses. So, his team has developed a pipeline using bioinformatics and Sanger-based phased sequencing for performing such analysis.
Using public databases, his team identifies areas of interest in the genome, and drills down to where structural variants are located. Roses' team has built a database for quickly locating structural variants. Next, they look for these structural variants in patients' samples using Sanger-based phased sequencing, which enables researchers to assess the polymorphic length of the markers on each copy of the gene. The variants are statistically analyzed for their association to specific disease phenotypes. Roses' group also conducts functional analysis to gain insights into a particular variant's role in causing a phenotype, such as Lewy bodies in Alzheimer's.
Roses' startup Zinfandel Pharmaceuticals used a similar approach a few years ago to discover that different lengths of the TOMM40 rs10524523 poly-T polymorphism, when combined with a person's APOE ε3 and APOE ε4 polymorphisms and age, can be used to predict his or her risk for mild-cognitive impairment due to Alzheimer's in five years. "There are four adjacent structural variants in intron 6 and the TOMM40-523 locus is the third one," Roses said.
For the work on Lewy bodies AD, researchers used their database of structural variants (housed at Polymorphic DNA Technologies) and a bioinformatics algorithm to score and catalogue regions of the SCNA gene with large number of alleles that might contain variants functionally important in the development of Lewy bodies AD. They genotyped samples obtained at autopsy from 120 patients with Lewy bodies AD and 361 AD patients without this abnormality, and performed statistical analysis to define associations between SNCA variants and Lewy Bodies pathology. Comparing their findings to the structural variant database, they homed in on three SNPs related to the Lewy body phenotype, but only the rs2298728 SNP was statistically significant.
To further investigate this area with the rs2298728 SNP, Chiba-Falek's team performed phased sequencing, performed by Polymorphic, on a subset of 95 samples from AD patients with and without Lewy bodies, and identified nine variations. When researchers further analyzed these variations for their haplotype phase, they were able to narrow their search to four haplotype combinations that showed a strong association for Lewy bodies. But only one of these, haplotype 3, conferred risk for developing extensive Lewy bodies AD.
Phased sequencing, which has been critical to Roses' work with TOMM40 and Lewy bodies AD, has enabled researchers to figure out whether a variant lies on a particular chromosome and which variants are inherited together. Knowing the arrangement of variants along chromosomal copies can help advance understanding of how a gene influences a phenotype in a complex disorder like AD.
For example, with phased sequencing data, Roses' team could determine how haplotype 3 inheritance impacted the level of Lewy bodies in a patient. "All the autopsies with extensive Lewy bodies showed haplotype 3 homozygosity," Roses said. Patients' samples that had an intermediate level of Lewy bodies, were heterozygous for haplotype 3, and none of autopsies without this abnormality were haplotype 3.
Although NGS is able to perform phased analysis, Roses isn't convinced of the accuracy of available platforms to determine structural variants. "When and if NGS becomes accurate for repeat lengths, this could be much easier," he said, adding that for now, using Sanger sequencing is sufficient for focused research purposes.
Additional analysis by Chiba-Falek and colleagues showed that haplotype 3 was linked to higher SNCA mRNA expression, with the highest levels in homozygotes. Epigenetic analysis further corroborated this finding. "We did expression analysis and found out that there seems to be an enhancer region in that CT-rich area," Scott Sundseth, senior director of pharmacogenetics consulting at Cabernet Pharmaceuticals (another wine-themed startup headed by Roses that conducts PGx consulting for drugmakers), recently said at the Festival of Genomics in Boston. "There was increased expression in haplotype 3 that is potentially responsible for the functional change that's happening."
Importantly, haplotype 3 is also tagged by the minor allele of the rs2298728 SNP, which researchers said could be a proxy for identifying people with this haplotype associated with risk of Lewy bodies AD. According to Sundseth, the haplotype 3 finding related to Lewy bodies AD has clinical implications. "It could be developed into a diagnostic," he said. "We discovered this a year ago, [but] we haven't done anything with it yet."
This research may be important for identifying patients without AD who are still at risk for Lewy body dementia, and Roses' team is looking at autopsies of non-AD patients to study this further. Researchers have also seen patients with Lewy bodies who don't have dementia, similar to people whose brain tissue after autopsy show beta-amyloid — commonly associated with AD — who didn't have dementia. Roses' group will study haplotype 3 in other cohorts to validate their finding and further characterize why Lewy bodies are associated with dementia or Alzheimer's in some patients but not others.
Chiba-Falek and colleagues wrote in the paper that they will use biological systems and genome editing approaches to confirm haplotype 3's regulatory role on SNCA expression and Lewy bodies formation. Additionally, Sundseth said his colleagues are applying the structural variants research strategy to Amyotrophic Lateral Sclerosis in order to better understand how alternative mRNA splicing could affect disease pathogenesis.
They are also collaborating with other groups to investigate structural variants in non-central nervous system disorders, such as cardiovascular and metabolic diseases. He and his colleagues are interested in seeing if "structural variants, which we have predicted would have an important impact on regulating genes and gene structures, indeed do and [if] the pathogenic mechanisms are very common," Sundseth said.
The latest study was partly funded by the National Institute of Neurological Disorders and Stroke. Duke's Kathleen Price Bryan Brain Bank provided the brain tissue samples for analysis in the study.