Missense variant in TREML2 protects against Alzheimer's disease

Abstract

TREM and TREM-like receptors are a structurally similar protein family encoded by genes clustered on chromosome 6p21.11. Recent studies have identified a rare coding variant (p.R47H) in TREM2 that confers a high risk for Alzheimer's disease (AD). In addition, common single nucleotide polymorphisms in this genomic region are associated with cerebrospinal fluid biomarkers for AD and a common intergenic variant found near the TREML2 gene has been identified to be protective for AD. However, little is known about the functional variant underlying the latter association or its relationship with the p.R47H. Here, we report comprehensive analyses using whole-exome sequencing data, cerebrospinal fluid biomarker analyses, meta-analyses (16,254 cases and 20,052 controls) and cell-based functional studies to support the role of the TREML2 coding missense variant p.S144G (rs3747742) as a potential driver of the meta-analysis AD-associated genome-wide association studies signal. Additionally, we demonstrate that the protective role of TREML2 in AD is independent of the role of TREM2 gene as a risk factor for AD.

Keywords: Alzheimer's disease; Association; Conditional analysis; Endophenotype; Gene; Genome-wide association studies; TREM2.

Fig. 1

Odds ratios for rs9381040 (IGAP hit), rs3747742 (TREML2, p.S144G), and rs75932628 (TREM2, R47H) among AD patients, as compared with control subjects, at each study center and overall. Shown are the combined estimates of the AD risk of possessing rs9381040 (IGAP hit), combined odds ratios analyses were homogeneous (p = 0.69, by Woolf test for heterogeneity). Panel (A), the rs3747742 (TREML2, p.S144G) (p = 0.81, by Woolf test for heterogeneity), panel (B), the rs75932628 (TREM2, p.R47H) (p = 0.97, by Woolf test for heterogeneity), panel (C), rs75932628 (TREM2, p.R47H) after conditioning for rs3747742 (TREML2, p.S144G) panel (D). The triangles represent ADGC study, the inverted triangles represent ARUK study, squares represent GERAD study, circles represent EADI study and the diamonds represent the summary odds ratio. The horizontal lines indicate the 95% confidence intervals of the estimates. Abbreviations: ADGC, Alzheimer's disease genetic consortium; ARUK, Alzheimer's Research UK; EADI, European Alzheimer's disease initiative; IGAP, international genomics of Alzheimer's project; GERAD, genetic and environmental risk for Alzheimer's disease.

Fig. 2

Association of TREM2 and TREML2 variants with CSF ptau levels. Panel (A) CSF ptau181 levels by rs9381040 genotype (IGAP meta-analysis most significant SNP). AG + GG versus AA p = 0.04. (Panel B) CSF ptau181 levels by rs3747742 genotype (TREML2, missense variant p.S144G). AG + GG versus AA p = 0.02. Panel (C) CSF ptau181 levels by rs75932628 genotype (TREM2, missense variant p.R47H). AG versus AA p = 0.0016. Panel (D) CSF ptau181 levels by rs3747742 genotype (TREML2, missense variant p.S144G). AG + GG versus AA excluding the variant p.R47H carriers p = 0.03. The mean and the standard error of the mean (SEM) for the normalized residuals CSF ptau181 levels are shown in blue. Abbreviations: CSF, cerebrospinal fluid; IGAP, international genomics of Alzheimer's project; SNP, single nucleotide polymorphisms.

Fig. 3

Microglial expression of TREM2 and TREML2 show opposing effects in the presence of IL-1b. TREM2 panel (A) and TREML2 panel (B) gene expression were analyzed in primary mouse microglia and astrocytes activated by 0.2 ng/mL IL-1β for 24 hours. Induction of Saa3 expression panel (C) serves as a positive control for IL-1β stimulated activation. Abbreviation: IL-1β, interleukin-1 beta.