New Analysis Reveals Gene Activity That May Underlie Parkinson’s
Studies on twins and families have suggested that susceptibility to Parkinson’s disease has a substantial genetic component. Now, an analysis of gene activity in different brain tissues has identified gene candidates that may be directly involved in causing the neurodegenerative disorder.
The study, “A transcriptome-wide association study identifies susceptibility genes for Parkinson’s disease,” was published in the journal NPJ Parkinson’s Disease.
One method to identify genetic risk factors is a genome-wide association study or GWAS, which compares genomic data from people with and without Parkinson’s to find significant differences that may be linked to the disease. The genome is the entire set of genetic instructions present in a cell.
However, many of the GWAS signals are often found in genomic regions outside genes, or the segments of DNA that carry instructions for proteins; thus, few have been implicated in specific biological mechanisms.
“Although recent GWAS has seen great strides in identifying risk [points] associated with [Parkinson’s disease], the functional significance of these associations remains elusive,” the researchers wrote.
Given that, a team of researchers at the Hunan University of Chinese Medicine, in China, now turned to another approach, called a transcriptome-wide association study or TWAS. This approach investigates the relationship between disease risk and gene expression — gene activity — which is directly tied to protein production and biological function.
Using TWAS gene activity data from 13 different brain tissue samples, the researchers sought to identify gene candidates that may directly be involved in susceptibility to Parkinson’s.
The team included data from 15,056 Parkinson’s cases and 449,056 people without the condition (controls), as well as 18,618 UK Biobank proxy cases, which are individuals who do not have Parkinson’s but have a first-degree relative — a parent, sibling, or child — who does.
A total of 18 genes were found whose expression was significantly associated with Parkinson’s. The most significant gene, called LRRC37A2, was associated in all 13 brain tissues, including the hypothalamus and basal ganglia, an area that contains nerve cells affected by the disorder.
There also were six genes found in single tissues: RNF40 and CPLX1 were discovered in the cerebellar hemisphere, VKORC1 in the cortex, MAP3K14 and GAK in the cerebellum, and CENPV in the frontal cortex. After validation, 14 of the 18 remained significant, “suggesting that their signals were genuine and not due to chance,” the researchers wrote.
Next, the TWAS data was cross-referenced with Parkinson’s GWAS data to find disease-related genetic changes that may impact gene activity. The researchers identified eight independent genes, including LRRC37A2, MMRN1, CD38, RNF40, GPNMB, ZSWIM7, GAK, and CPLX1. Of these, four — CD38, LRRC37A2, RNF40, and ZSWIM7 — were newly identified and are located in areas of the genome associated with Parkinson’s.
One of these genes, CD38, encodes for a protein strongly expressed in brain cells. Its activity increases with age, the primary risk factor for most neurodegenerative diseases, including Parkinson’s. Experiments show that mice lacking this gene are protected from neurodegenerative injury.
Among the eight genes identified, five — LRRC37A2, MMRN1, GPNMB, ZSWIM7, CPLX1 — were significantly associated with Parkinson’s in at least one brain tissue in a validation replication analysis. Again, LRRC37A2 was disease-associated in all 13 tissue samples.
A heritability analysis found genes in 10 of 13 brain tissues significantly contribute to Parkinson’s heritability. The TWAS genes identified in all 13 tissues explained 22.17% of the estimated heritability.
The TWSA genes were then cross-referenced to the Genetic Association Database (GAD), which associates genetic changes with diseases. The analysis detected six conditions, including Parkinson’s, as well as tobacco use disorder, cholesterol and LDL cholesterol, bone mineral density, and cleft lip and cleft palate. LDL cholesterol is often termed “bad” cholesterol and can significantly increase a person’s risk of premature cardiac disease and stroke.
“Tobacco use disorder, cholesterol, and bone mineral density, also have been reported to be associated with [Parkinson’s],” the researchers wrote.
TWSA gene-associated biological pathways included the mTOR signaling pathway, with alterations being a common hallmark of human neurological disorders. Pathways related to the metabolism of selenium-related compounds also were associated, as was the PPAR signaling pathway, which is activated by naturally occurring fatty acids and plays an essential role in the regulation of numerous biological processes.
Finally, the team identified 166 characteristics (phenotypes) linked to genetic changes in the TWAS genes, which were neurological, psychiatric, and cognitive. A further correlation between Parkinson’s and 122 identified traits with current GWAS data found a positive correlation with older age at first sexual intercourse and Parkinson’s.
Five traits that negatively correlated with Parkinson’s included heel bone mineral density, current tobacco smoking, and three measures of leg fat. “Importantly, tobacco use disorder and bone mineral density have been identified through GAD disease enrichment analyses,” the researchers noted.
The team pointed out that studies show “smoking is associated with a lower incidence of [Parkinson’s disease],” and nicotine stimulates nerve cells that are damaged in Parkinson’s and protects against nerve cell damage in animal models.
“In summary, by using the TWAS method that we generated recently, we identified 18 genes associated with [Parkinson’s disease],” the researchers wrote. “We identified eight conditionally independent genes, and we demonstrated that several of the GWAS significant signals on [Parkinson’s disease] could be driven by genetically regulated gene expression.”
These findings are expected to further research into the effects of these specific genes in possibly causing Parkinson’s.
“In conclusion, we prioritized genes that are likely to affect [Parkinson’s disease] by using a TWAS approach and identified phenotypes associated with [Parkinson’s disease],” the team wrote.