Aims

The existence of individuals who remain cognitively intact despite the presence of histopathological signs of Alzheimer’s disease (AD), here referred to as “Non-demented with AD neuropathology” (NDAN), suggests that some unknown mechanisms are triggered to resist cognitive impairment. Synaptic dysfunction has been identified as one of the major AD causes and it is established that microglia, attracted to plaques, phagocyte damaged synapses. A possible mediator of this process is represented by TREM2, a recently identified AD risk factor. Based on TREM2 role in the scavenging function of microglia, we hypothesize that an efficient microglial phagocytosis underlies synaptic resilience in NDAN, thus protecting from memory deficits.

Methods

Using immunofluorescence microscopy, a comparative study of human post-mortem frontal cortices of aged-matched individuals, AD and NDAN subjects has been performed. The distribution of activated microglia (IBA1 and IBA1/CD68 positive cells) and the expression of microglia-related proteins (TREM2 and DAP12) were evaluated. Furthermore, to test the efficacy of microglia in removing debris and damaged synapses, preservation of synapses around plaques was assessed using MAP2 and tubulin βIII as dendritic and axonal markers, and PSD95 as a postsynaptic marker.

Results

NDAN individuals show higher microglial activation and TREM2 expression, as well as preserved axonal and dendritic structure around plaques vs. AD. Low levels of PSD95 around NDAN plaques may suggest a prompt removal of damaged synapses by efficient microglia.

Conclusions

Our results suggest a higher efficiency of TREM2-induced phagocytic microglia in removing damaged synapses, underlying synaptic resilience in NDAN individuals.

Supported by NIH/NIA R01AG069433, R01AG060718, R56063405 to GT

Aims

Glial and immune activation in neurodegenerative disease has received increasing attention as an important contributor to disease pathogenesis and progression in Alzheimer’s disease pathology. Here we aim to clarify the individual and collaborative effects of amyloid and tau pathology on responses of glial and immune cells in the brain.

Methods

We study cellular level changes of hippocampal cells by untargeted single cell RNA-seq using different AD mouse models that capture tau pathology, amyloid pathology, or combined pathology. We also examine how these disease-associated glial changes are affected by TREM2. We utilize in situ hybridization to corroborate major findings from single cell RNA-seq and to examine spatial distribution of gene expression changes in different cells.

Results

We observe distinct responses of microglia, astrocytes, oligodendrocytes and T cells to tau, Aβ or combined pathology. Interestingly, the disease-associated oligodendrocyte and astrocyte changes we observe show little dependence on TREM2 while the majority of microglial responses are dampened in Trem2 knockout mice.

Conclusions

Our study highlights distinct transcriptional states of microglia, oligodendrocytes, astrocytes, and T cells in the degenerating brain and how they are influenced by amyloidosis and a key Alzheimer’s Disease risk gene, Trem2. Our study may help identify diagnostic biomarkers and therapeutic targets related to disease progression.

Aims

The relative contributions of microglia and infiltrating monocyte-derived macrophages (MDMs) to containing Alzheimer’s disease (AD) are not fully understood. In the 5xFAD animal model of amyloidosis, disease-associated microglia (DAM) expressing the Triggering receptor expressed on myeloid cells 2 (TREM2), are found in close proximity to amyloid-beta (Aβ) plaques. Deletion of TREM2 results in the absence of DAM and in an increased Aβ-plaque load. However, the necessity of TREM2 and DAM for resolving AD pathology is still debatable.

Methods

Here, we activated systemic immunity by blocking the programmed cell death protein 1 / ligand (PD-1/PD-L1) pathway in TREM2-null and TREM2-intact 5xFAD mice, to decipher the roles of the different myeloid populations in mitigating AD pathology.

Results

We found that anti-PD-L1 treatment resulted in cognitive improvement in TREM2-null and TREM2-intact 5xFAD mice. In addition, in both TREM2-null/5xFAD and TREM2-intact/5xFAD, the treatment resulted in a reduction in water-soluble-Aβ, while reduction of insoluble-Aβ was observed only in TREM2-intact/5xFAD mice. Eliminating monocytes using anti-CCR2 antibody fully abrogated the observed effects of anti-PD-L1 treatment in TREM-null/5xFAD mice, and partially eliminated the effects in the TREM2-intact/5xFAD. Single-cell RNA-seq of myeloid cells isolated from TREM2-null/5xFAD brains revealed that MDMs express unique scavenger receptors, previously linked to soluble-Aβ removal, such as Macrophage scavenger receptor 1 (MSR1).

Conclusions

Overall, our findings highlight a novel TREM2-independent pathway by which cognitive improvement and removal of soluble-Aβ are achieved in an amyloidosis model. Thus, our results support the potential of MDM-harnessing immunotherapy in treating AD patients, irrespective of whether they carry a TREM2 mutation.

Aims

Trimethylamine-N-oxide (TMAO), a metabolite originated from microbial metabolism, is associated with cardiovascular disease, a relationship potentially explained by increased TMAO-induced expression of pro-inflammatory cytokines in the periphery; its relationship with the central nervous system (CNS) immune response is however unknown. Studies in aging mice demonstrate that reducing circulating TMAO reduces cognitive decline and neuroinflammation and, in humans, CSF TMAO associates with greater Alzheimer’s disease (AD) pathology as shown by CSF. Here, our primary objective was to explore associations of TMAO with CSF biomarkers of glial activation and inflammation measured across the AD spectrum.

Methods

CSF was collected from participants in the Wisconsin Registry for Alzheimer’s Prevention and the Wisconsin Alzheimer’s Disease Research Center (cognitively unimpaired [n=510], mild cognitive impairment [n=41], dementia [n=45]). Biomarkers were measured with NeuroToolKit robust prototype assays (Roche Diagnostics). TMAO relative abundance in CSF and blood plasma were measured via Metabolon’s UHPLC/MS metabolomics platform. CSF levels of YKL-40, GFAP, IL-6, S100, and sTREM2 were log-transformed and regressed on TMAO levels, adjusting for age, sex, APOE ε4 status, and parental history of AD.

Results

CSF TMAO levels were significantly positively associated with CSF sTREM2 (Figure). Positive correlations with GFAP and YKL-40 were also observed, although sTREM2 was the only CSF biomarker significantly associated with TMAO relative abundance in CSF or plasma.

Conclusions

In this study, TMAO levels were associated with sTREM2, and additional studies in animal models are needed to determine whether TMAO is mechanistically associated with AD pathology and microglial activation.

Aims

Our recent discovery cerebrospinal fluid (CSF) proteomics study, revealed potential AD-specific biomarkers related to inflammatory pathology, such as migration inhibitory factor (MIF) and triggering receptor expressed on myeloid cells 1 (TREM-1). Here, we aimed to validate these findings including also samples from patients with dementia with Lewy Body (DLB) using an accessible technology, with the purpose for clinical implementation.

Methods

MIF and TREM-1 were measured in CSF from patients with AD (n=38), DLB (MIF: n=50, TREM-1: n=40) and non-demented controls (cognitively unimpaired, n=37) using assays on the antibody-based SimplePlex technology, that were validated for CSF analysis. Differences in protein levels between diagnostic groups and their correlation to AD CSF biomarkers and mini-mental state examination (MMSE) scores were tested.

Results

CSF MIF levels were increased in AD (1.2-fold, p=0.002) and DLB (1.14-fold, p=0.036) compared to controls. MIF levels correlated with both total- and phosphorylated-Tau (t-Tau, r=0.652, p<0.001; p-Tau, r=0.682 p<0.001) but not with Amyloid-beta 42 (Aβ42, p>0.05) or MMSE (p>0.05). CSF TREM-1 levels were increased in AD compared to DLB (1.4-fold, p=0.016) and controls (1.5-fold, p=0.003) and correlated with CSF biomarkers (t-Tau: r=0.317, p=0.001; p-Tau: r=0.297, p=0.001; Aβ42: r=-0.221, p=0.018) but not with MMSE (p>0.05).

Conclusions

Inflammatory-related proteins MIF and TREM-1 are increased in AD, thereby, validating our previous proteomics study in an independent cohort using a technology that can be used to reach clinical implementation. These novel biomarkers reflect an underlying neuroinflammatory process in AD, and further studies are needed to establish their diagnostic potential for the discrimination of AD from other dementias.

Abstract Body

The innate immune receptor TREM2 is a strong genetic risk factor for development of Alzheimer’s disease and other neurological disorders. TREM2 loss-of-function mutations exacerbate Alzheimer’s pathology, and activation of TREM2 in a disease state provides therapeutic benefits. Conversely, amassing evidence indicates that TREM1 functions in the opposite direction by acting as a maladaptive toxic amplifier of immune responses. In a series of studies, we have been investigating a possible role for TREM1 in Alzheimer’s Disease development. Similar to TREM2, our new data from post-mortem autopsy tissues indicate that TREM1 increases in Alzheimer’s Disease brain and correlates with overall level of brain Amyloid. In clinical cohort studies, we find that CSF soluble levels of TREM1 (sTREM1) positively associate with CSF Amyloid and increase at preclinical AD stages in absence of elevated Tau. In mouse model experiments, TREM1 amplifies proinflammatory responses in aging and AD, and blocking its signaling promotes immune metabolic homeostasis. These new findings indicate that TREM1 inhibition may be an effective treatment strategy for Alzheimer’s Disease and possibly other neurodegenerative disorders.