mRNA expression profiles were examined for the following genes: DNA methylation writers, DNMT1 (NM_001130823), DNMT3A (NM_022552.5) and DNMT3B (NM_006892.4); readers, MeCP2 (NM_004992.4), ZBTB4 (NM_001128833.2), ZBTB33 (NM_001184742.2), ZBTB38 (NM_001376113.1) and UHRF1 (NM_001048201.3); and erasers, TET1 (NM_030625.3), TET2 (NM_001127208.3), TET3 (NM_001287491.2), TDG (NM_003211.6), MBD4 (NM_001276270.2), AICDA (NM_020661.4), GADD45A (NM_001924.4), GADD45B (NM_015675.4) and GADD45G (NM_006705.4). RNA methylation effector genes examined were as follows: writers, NOP2/NSUN1 (NM_001258308.2), NSUN2 (NM_017755.6), NSUN3 (NM_022072.5), NSUN4 (NM_199044.4), NSUN5 (NM_148956.4), NSUN6 (NM_182543.5), NSUN7 (NM_024677.6) and TRDMT1/DNMT2 (NM_004412.7); readers, ALYREF (NM_005782.4) and YBX1 (NM_004559.5); and erasers, ALKBH1 (NM_006020.3), TET1 (NM_030625.3), TET2 (NM_001127208.3) and TET3 (NM_001287491.2). Expression data were obtained from normalised fragments per kilobase of transcript per million (FPKM) values derived from RNA sequencing and presented as a z-score value of expression.
Individuals were categorised by clinical diagnosis AD versus control and grouped by age of death, APOE4 allele carrier status and Braak and CERAD staging in HIP and STG (Supplementary Table 3) and IPC or WM (Supplementary Table 4). For Braak staging, three groups were generated to represent levels of pathology. Group 1 corresponded to individuals representing low levels of pathology with a Braak stage between 0 and II. Group 2 had a Braak stage between III and IV and had moderate pathology. In the third group, samples had Braak stages between V and VI, indicating the highest level of pathology. To analyse by CERAD ranking, samples were divided into two groups with a CERAD score of 0–1 indicating low amyloid load and a CERAD score of 2–3 representing high amyloid load.
Supplementary Fig. 1 provides a flow chart of our analysis pipeline. In a first phase, differences between age, sex, APOE4 allele status and Braak and CERAD staging in the groups AD and control were assessed. We then tested for significant correlations between diagnosis status, Braak, CERAD and APOE4 across each of the four brain regions. In a second phase, we compared RNA abundance profiles across the 32 effector proteins in AD and control groups for each brain region. Subsequently, differences in gene expression were examined in individuals classified by Braak and CERAD rankings. In the third arm of the study, we assessed differential expression in individuals who self-reported TBI. For this analysis, the cohort was divided into five groups: a TBI-control group of aged individuals without TBI and without AD (referred to as ‘TBI-control’); a group of individuals which includes all TBI individuals, i.e. with and without dementia (All TBI); a third group of individuals with TBI but no AD (TBI + no AD); a fourth group which includes individuals with TBI and AD (TBI + AD); and a final group with no TBI but have AD (no TBI + AD). No differences were observed between gender, age and brain tissue post-mortem interval across the TBI groupings (Supplementary Tables 5 and 6). Similarly, in individuals who reported head injuries, we found no difference between age of first TBI incident, number of TBI incidents with loss of consciousness or duration or across the TBI groupings (Supplementary Table 7).
Free full text: Click here