Date of Award

2019

Document Type

Thesis

Publisher

Edith Cowan University

Degree Name

Master of Medical Science

School

School of Medical and Health Sciences

First Supervisor

Associate Professor Simon Matthew Laws

Second Supervisor

Associate Professor Peter Roberts

Third Supervisor

Dr Tenielle Porter

Abstract

Background

In 2017 approximately 50 million people worldwide were living with dementia. With Alzheimer’s disease (AD), accounting for 50-70% of dementia cases making this debilitating disease, with no current effective prevention, treatment or cure, a critical healthcare concern. Genome wide association studies (GWAS) have identified a number of risk genes for late onset AD (LOAD); Apolipoprotein E (APOE), a gene involved in the cholesterol/lipid pathway is considered the gene with the greatest risk. The third most associated AD risk gene is Clusterin (CLU), is also involved in the cholesterol/lipid pathway. CLU has been implicated in both a protective and aggravating role in AD making it a compelling gene for further study. The identification of a fluid-based AD biomarker/s has become an increasingly important area of study to assist in the identification of individuals before substantial pathological or cognitive symptoms arise. Plasma clusterin is one such biomarker showing promise as levels differ between healthy individuals and those with mild cognitive impairment (MCI) and AD. Methylation of CLU, from brain samples, has been associated with Aβ pathology and lower risk of AD diagnosis. In addition, methylation of other AD associated risk genes has been shown to be altered in the periphery. These factors make CLU a prime candidate for peripheral methylation study.

Aims

The overarching aim of the study was to investigate methylation within the CLU promoter region in peripheral blood samples. Specifically, to determine whether there is a relationship between specific methylation sites and AD-related phenotypes. The first study aimed to crosssectionally assess CLU promoter region methylation and its association with clinical classification, genetic variation in CLU (rs9331888/rs11136000), and pathological biomarkers (Chapter 3.2). The second study aimed to analyse the influence of methylation on longitudinal cognitive performance (Chapter 3.3).

Aims

The overarching aim of the study was to investigate methylation within the CLU promoter region in peripheral blood samples. Specifically, to determine whether there is a relationship between specific methylation sites and AD-related phenotypes. The first study aimed to crosssectionally assess CLU promoter region methylation and its association with clinical classification, genetic variation in CLU (rs9331888/rs11136000), and pathological biomarkers (Chapter 3.2). The second study aimed to analyse the influence of methylation on longitudinal cognitive performance (Chapter 3.3).

Methods

Utilising DNA samples collected as part of the Australian Imaging, Biomarkers and Lifestyle Flagship Study of Ageing (AIBL). Methylation analysis was conducted on an initial 768 samples comprised of healthy controls (HC), MCI and AD participants. Methylation data was collected using the Agena MassArray platform across the CLU promoter region. Percentage methylation was determined by the EpiTYPER software. ANCOVA’s were used to study association between methylation levels and clinical classifications as well as genetic variants, whilst Spearman’s Partial correlations were used to investigate the relationship between methylation and brain imaging and fluid biomarker data (Chapter 3.2). Linear mixed models (LMMs) were employed to assess associations between methylation and cognitive performance, both at baseline and over 7.5 years (Chapter 3.3). Analyses were first conducted across the whole sample, followed by HC only, then HC with high brain Aβ burden (Aβ+). All P-values were corrected for the false discovery rate (FDR).

Results Analysis revealed an association between clinical classification and methylation. After FDR correction significant decreases in methylation from HC to MCI was seen at 3 loci, with CpG_14 and CpG_15 showing a significant decrease from HC to AD. CpG_25_26_27_28 also showed a decrease in methylation between HC and AD. When analysing methylation patterns with respect to rs11136000 a significant difference in methylation levels were observed between non-carriers and carriers of the protective T allele at CpG_14, CpG_16 and CpG_23_24. This was only observed when assessed across all participants, whilst no significant differences in methylation were observed with respect to rs9331888.

An increased percentage methylation at CpG_16 was associated with elevated plasma clusterin (⍴=0.1858, p=0.0034, q=0.0157) within the HC group, which remained as a trend in the HC Aβ+ after FDR correction (β=0.2704, p=0.0043, q=0.0817). With respect to cerebrospinal fluid (CSF) biomarkers no sites survived FDR correction in the whole cohort. CpG_37 was inversely correlated with phosphorylated tau (P-tau; ⍴=-0.4585, p=0.001, q=0.0155) within the HC cohort. Whilst with the Aβ+ HC group higher percentage methylation at CpG_1 (⍴=-0.6584, p=0.0016, q=0.0028), CpG_8_9 (⍴=-0.4704, p=0.0363, q=0.0297) and CpG_10 (⍴=-0.4433, p=0.0503, q=0.0297) was associated with elevated A42 CSF levels.

Increased methylation at CpG_1, CpG_2_3, CpG_8_9, CpG_14, CpG_16, CpG_19 and decreased methylation at CpG_20_21, CpG_23_24, CpG_25_26_27_28, CpG_34, CpG_36, CpG_38_39 and CpG_41 were associated with a higher brain Aβ burden across the whole cohort. Whilst methylation at CpG_15 positively correlated with cortical grey matter volume. Within HC’s, increased methylation at CpG_1, CpG_2_3, CpG_8_9, CpG_14, CpG_16, CpG_19, and decreased methylation at CpG_20_21, CpG_23_24, CpG_25_26_27_28, CpG_36, CpG_38_39 and CpG_41 were associated with a higher brain A burden. Whilst methylation at CpG_15, CpG_36 and CpG_41 positively correlated with cortical grey matter volume. Again, within the Aβ+ HC group, increased methylation at CpG_8_9, CpG_14, CpG_16, CpG_19, and decreased methylation at CpG_20_21, CpG_25_26_27_28, CpG_34, CpG_36, and CpG_41 were associated with a higher brain Aβ burden. Whilst a significant positive correlation between methylation at CpG_37 and cortical white matter volume was observed.

Whilst several methylation sites within the CLU promoter reached a nominal level of significance with respect to both baseline cognitive performance and change in cognitive performance over 7.5 years no sites retained significance after FDR correction. This finding was consistent when investigated across the entire cohort, within the HC and within the Aβ+ HC group.

Conclusions

Differential methylation was seen across the CLU promoter region in relation to clinical classification, SNP’s and both fluid and brain imaging biomarkers. These observations seen in the periphery provide further evidence of methylation, and specifically of methylation in the CLU promoter region, having the potential to be a biomarker in early AD diagnosis. Of particular interest is the decrease in methylation seen between clinical classifications and the associated impact on gene expression. Another area of noted interest is the association seen between CLU promoter region methylation and Aβ burden in the brain. Further exploration is warranted to validate and further examine these results by generating longitudinal methylation data to assess whether changes in methylation track with changes in AD clinical and pathological characteristics.

Download

Share

 
COinS