No increase of CD8+ TEMRA cells in the blood of healthy adults at high genetic risk of Alzheimer's disease

The adaptive immune system plays a crucial role in the pathophysiology and progression of Alzheimer's disease (AD).1 In this context, we read with great interest the recent publication by Gericke et al.,2 who reported a positive association of CD8+ CD45RA-reactivated T effector memory (TEMRA) cells with the presence of brain β-amyloid and AD-specific changes in plasma biomarkers in ∼180 cognitively healthy individuals. While an increase of TEMRA cells had been reported previously in clinically affected AD patients compared to controls (e.g., ref.3), the results by Gericke et al.2 extended these findings to the pre-clinical disease phase. In the current study, we investigated whether the alteration of TEMRA cells may manifest even earlier in the course of the disease, that is, in younger and in older, cognitively healthy adults at high genetic risk of AD based on polygenic risk score (PRS) analyses. Detailed methods can be found in the Supplementary methods. Briefly, quality-controlled genome-wide single nucleotide polymorphism (SNP) data (n = 7,536,015)4 and CD8+ TEMRA cell proportions of total CD8+ T cells in blood (Table S1)5 were available for 370 cognitively healthy individuals comprising a group of young adults (n = 120, mean age: 29 years, range 23–35) and a group of older people (n = 250, mean age: 69 years, range 60–82). All participants were recruited as part of the Berlin Aging Study II (BASE-II), a multi-institutional and multidisciplinary project aimed to characterize the individual changes of healthy participants (at the time of recruitment) in the aging process.6 All participants included here had a Mini-Mental State Examination score ≥27. The PRS was calculated using R and PRSice-27 and was based on apolipoprotein E (APOE) e4 (rs429358) and 70 of the 83 genome-wide significantly associated (α = 5.0E-8), independent SNPs from the largest AD genome-wide association studies (GWAS) to date based on 487,511 individuals.8 For each age group, we performed linear regression analyses of TEMRA cells (log-transformed; Figure S1) on the AD PRS as a continuous score. In addition, we performed linear regression analyses on the established AD risk variants e4 and e2 (rs7412) in APOE. As a positive control, we performed linear regression analyses of TEMRA cell counts on anti-cytomegalovirus (CMV) antibody titer status and older versus young age.9, 10 All analyses were adjusted for sex and four principal components. Significance was defined as α = 0.05. We performed a range of sensitivity analyses by using an optimized AD PRS that explained the largest amount of variance of the TEMRA cell counts, adjusting for CMV as additional covariate, contrasting the top versus lower PRS quintile to allow for a non-linear model, and using arcsinh transformation of the TEMRA cell data as applied in the original study.2 As expected, we observed a strong and statistically significant increase of CD8+ TEMRA cells in CMV-seropositive participants (young adults: p = 1.17E-15, older: p = 2.46E-24) as well as with increasing age (p = 3.78E-13, Table 1, Figure S2). However, the PRS on genome-wide significant AD risk SNPs did not show statistically significant association with CD8+ TEMRA cells, neither in young adults (incremental phenotypic variance explained [ΔR2] = 0.00003, p = 0.954) nor in the older (ΔR2 = 0.0007, p = 0.670, Table 1, Figure S2). Furthermore, and in contrast to the original study,2 APOE rs429358 and rs7412 also failed to show significant associations with the number of TEMRA cells (Table 1, Figure S2). These results did not change in sensitivity analyses: Optimizing the PRS resulted in a score composed of 7234 SNPs (threshold for variant selection: p = 4.15E-03) in young adults and 76,312 SNPs (p-value threshold = 9.71E-02) in older adults, but the analysis still yielded non-significant results. None of the other sensitivity analyses, including contrasting PGS extremes, changed the results substantially (Table 1). In summary, we did not detect a statistically significant increase of TEMRA cells in cognitively healthy younger or older adults with a high genetic risk of AD. While we cannot exclude a lack of power to detect small effects, we investigated a substantially larger sample compared to that of Gericke et al.,2 observed the well-established effects of CMV status and age on TEMRA cell counts, and performed a range of carefully executed additional analyses, all of which showed null effects. Thus, while prior work2 suggests that the increase of TEMRA cells in the blood occurs shortly before the clinical onset of AD, that is, specifically when pathology is already present in the brain, our study shows that this increase does not manifest in cognitively healthy participants at increased genetic AD risk earlier in life. Further functional studies are needed to fully characterize the mechanisms and timing underlying the dynamic and intricate role of CD8+ TEMRA cells in AD. We are grateful to all BASE-II participants. The BASE-II research project (co-PIs: Lars Bertram, Ilja Demuth, Denis Gerstorf, Ulman Lindenberger, Graham Pawelec, Elisabeth Steinhagen-Thiessen, and Gert G. Wagner) has been supported by the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF) under grant numbers #16SV5536K, #16SV5537, #16SV5538, #16SV5837, #01UW0808, 01GL1716A, and 01GL1716B, and by the Max Planck Institute for Human Development, Berlin, Germany. Additional contributions (e.g., equipment, logistics, personnel) are made from each of the other participating sites. The responsibility for the contents of this publication lies with its authors. C.M.L. was supported by the Cure Alzheimer's Fund and the Heisenberg program of the German Research Foundation (DFG; LI 2654/4-1). The authors declare no conflicts of interest. Author disclosures are available in the Supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.