Spencer Foutz and Dr. John Kauwe, Department of Biology
Introduction
For the last several years, the e4 allele in the APOE gene region has been a strong determinant of the development of Amyloid Beta plaques in the brain’s cerebral spinal fluid. As these plaques aggregate, the deposits become increasingly deadly to neuronal cells, and can eventually lead to Alzheimers’ Disease. While the e4 allele is a strong indicator of change in Amyloid Beta levels, there are believed to be other genetic variants within the APOE gene region that also greatly influence this protein. Our purpose was to further evaluate previous studies on this gene region in order to determine other variants which could possibly play a role in the genetic pathology of the onset of Alzheimers’ Disease.
Methodology
The methods used in this project had been tested in a previous experiment. That experiment had been conducted in the fall of 2012 and had shown that a particular single nucleotide polymorphism (SNP) was a significant variant of interest in our research. In an effort to evaluate the validity of the SNP (rs769449) in regards to its influence on AB42 protein levels, and in an attempt to discover other SNPs of interest, we designed our study with the following aims: Aim 1: Perform analyses of genetic variants (SNPs) in the APOE region with and without adjusting for the effects of the APOE e4 locus.
In order to analyze genetic variants in the APOE gene region, we first had to prepare a new dataset given to us by Washington University of St. Louis consisting of 10,000 patients. I then used a genome web browser to form a list of genetic variants found within the APOE gene region. This narrowed our variants of interest from 6 million to 247. Using the prepared dataset and our file of pertinent variants, I then used PLINK software to do statistical analyses which would determine the significance of our variants in this particular dataset.
The first analysis performed was conducted in all patients, without any adjustment for those who carried the harmful e4 variant or the protective e2 variant. Next, we repeated this analysis, this time controlling the effects of the e4 and e2 alleles. Our study was then repeated in patients who were homozygous for the e3 allele. The purpose of this third analysis was to verify that our first two analyses were valid, as it would be impossible for people homozygous for the e3 allele to be influenced by the e2/e4 alleles.
Aim 2: Perform bioinformatics analyses of the identified variants to further validate results and determine possible mechanisms of function for each genetic variant.
To further evaluate the results of our study, we used SAS statistical software to perform additional analyses to validate the results of our experiment.
Results
As of right now we are unable to show that Rs769449 is significantly associated with changes in AB42 protein levels in the CSF. We were also unable to find any other SNPs as candidates to independently alter AB42 protein levels. The first two analyses showed high significance for rs769449, confirming the results of our analyses previously performed in our separate dataset. Even when the e2 and e4 alleles were controlled for, rs769449 was a significant SNP in all patients in the dataset. However, the results of our analysis in group of patients homozygous for the e3 allele did not come back significant. This was the opposite result we had received from previous data. The pvalue was so insignificant that we questioned the validity of our study for quite sometime. After repeating all analyses twice and checking the dataset for errors, a SAS study was conducted in order to determine why this SNP was not significant in this subset of people. It was determined that out of all of the e3e3 carriers, none were carriers of the rs769449 variant. We then reviewed the data in our previous dataset and found that an unproportionally small portion of e3e3 carriers were also carriers for rs769449, making our claim of significance unvalid.
Discussion
Our study thus far has presented challenges as well as new opportunities that will be explored throughout this next year. First, the fact that there are so few carriers of variant rs769449 within e3 homozygous carriers makes it difficult to validate our claim of significance, and we will have to come up with other methods to do so. It also presents the question as to why so few people homozygous for the e3 allele are also carriers of the variant rs769449. Could the e3 allele be protective for this variant as well? It is also of interest to find out why so few people with the e3 allele have this variant when it has been shown to be completely independent from the e4 allele. Liinkage disequilibrium studies need to be done in order to discover the relationship between the e3 allele and rs769449. Pathway analyses should also be performed to determine other variants influencing our SNP of interest.
Conclusion
I wish to thank all those who helped me in the project, especially Mr Harker, who so generously donated money to this project. I owe Dr. Kauwe special thanks for mentoring me throughout this project. His teaching, patience, and support are what made it possible for me to work on this project. I would also like to thank Dr. Perry Ridge and Matthew Bailey for their continuous support and involvement in this project. Finally, I would like to thank Josue Gonzalez and Ivan Aranoand from the Kauwe lab for continuing this project in my absence this next year. My involvement in this project and others in the Kauwe lab has been the highlight of my university experience and I am grateful for the time, money and trust that everybody has put into my education here at Brigham Young University.