Nguyen, Aivi T.

Microglia help limit the progression of Alzheimer’s disease (AD) by constraining amyloid-β (Aβ) pathology, effected through a balance of activating and inhibitory intracellular signals delivered by distinct cell surface receptors. Human leukocyte Ig-like receptor B4 (LILRB4) is an inhibitory receptor of the immunoglobulin (Ig) superfamily that is expressed on myeloid cells and recognizes apolipoprotein E (ApoE) among other ligands. Here, we find that LILRB4 is highly expressed in the microglia of patients with AD. Using mice that accumulate Aβ and carry a transgene encompassing a portion of the LILR region that includes LILRB4 , we corroborated abundant LILRB4 expression in microglia wrapping around Aβ plaques. Systemic treatment of these mice with an anti-human LILRB4 monoclonal antibody (mAb) reduced Aβ load, mitigated some Aβ-related behavioral abnormalities, enhanced microglia activity, and attenuated expression of interferon-induced genes. In vitro binding experiments established that human LILRB4 binds both human and mouse ApoE and that anti-human LILRB4 mAb blocks such interaction. In silico modeling, biochemical, and mutagenesis analyses identified a loop between the two extracellular Ig domains of LILRB4 required for interaction with mouse ApoE and further indicated that anti-LILRB4 mAb may block LILRB4-mApoE by directly binding this loop. Thus, targeting LILRB4 may be a potential therapeutic avenue for AD.

AbstractBackgroundThere is a gap in knowledge concerning which tangle maturity levels (Moloney, Alzheimers Dement, 2021) are associated with antemortem measures (tau‐PET, MRI, cognitive testing). We hypothesized antemortem measures would be more strongly correlated with markers recognizing middling to advanced tangle maturity. We additionally sought to assess the strength of the relationship with tangle and neuritic pathology recognized by tangle maturity markers.MethodPosterior hippocampi from n = 30 cases with antemortem measures ≤3 years prior to death were cut (5µm) and immunostained with tau antibodies to evaluate early (AT8), middling (PHF‐1), or advanced (2E9) tangle maturity. GENIE (pattern recognition software) separated tangle from neuritic burden (Figure). Hippocampus, parahippocampus, and fusiform were analyzed to correlate digital pathology with flortaucipir tau‐PET standard uptake value ratio (SUVr, cerebellar crus normalization), MRI volume/thickness, and cognitive measures which included Auditory Verbal Learning Test (AVLT) delayed recall, Mini‐Mental State Exam (MMSE), and Clinical Dementia Rating (CDR).ResultThe strongest relationships observed with flortaucipir were with PHF‐1 and 2E9 burden (Table). AT8 tangle and neuritic burden correlations with flortaucipir were relatively similar in the hippocampus but were stronger with neuritic pathology in parahippocampus and fusiform. PHF‐1 tangle and neuritic burden correlations were relatively similar with flortaucipir. 2E9 correlations were stronger with neuritic pathology in hippocampus. MRI measures were more strongly correlated with PHF‐1 and 2E9 burden. AT8 tangle and neuritic burden were correlated with fusiform thickness. The correlations between MRI measures and 2E9 tangle and neuritic burden were similar. Correlations between cognitive measures were similar between tangle maturity markers. AVLT delayed recall correlated with all tangle maturity markers, but neuritic burden was stronger for AT8 and 2E9 in hippocampus compared to tangle burden. MMSE correlated with neuritic burden for all tangle maturity markers. CDR remained statistically non‐significant.ConclusionInterpretation of the strength of the relationship between tau burden and antemortem measures differed across tangle maturity markers. Neuroimaging measures may be more sensitive to middle/advanced maturity markers with limited differences between tangle and neuritic pathology. These findings may be biologically relevant and imply either a limited recognition of tangles depending on antibody epitope, and/or may represent the clinicoradiologic impact of increasing neuritic burden.

AbstractBackgroundPlasma biomarkers for Alzheimer’s disease (AD) such as phosphorylated tau protein 181 (p‐tau 181), neurofilament light chain (NfL), amyloid‐β 42/40 ratio (Aβ42/40), and glial fibrillary acidic protein (GFAP) have shown high performance in the discrimination of those with clinical mild cognitive impairment (MCI) or dementia from cognitively unimpaired individuals, but few population‐based autopsy studies have evaluated the neuropathological correlates of plasma biomarkers.MethodWe utilized a population‐based prospective study of 350 participants with autopsy and antemortem plasma biomarker data. Plasma biomarkers Aβ42/40 ratio, p‐tau181, GFAP, and NfL were measured using Quanterix Simoa assays. We predicted AD neuropathological change (ADNC), Braak staging, and neuritic plaque score using these plasma biomarkers and investigated the added utility of cognitive and demographic information. We utilized a variable selection procedure in cross‐validated (CV) logistic regression models by minimizing the Bayesian Information Criterion to identify the best set of predictors for a given outcome among plasma biomarkers, demographic variables, and neuropsychological tests included in the Mayo Clinic Preclinical Alzheimer Cognitive Composite (Mayo‐PACC).ResultADNC was best predicted with plasma GFAP, NfL, p‐tau181 biomarkers along with APOE ε4 carrier status and Mayo‐PACC cognitive score (CV AUC = 0.821). Braak staging was best predicted using plasma GFAP, p‐tau181, and cognitive scores (CV AUC = 0.788). Neuritic plaque score was best predicted using plasma Aβ42/40 ratio, p‐tau181, GFAP, and NfL biomarkers (CV AUC = 0.778). Inclusion of plasma biomarkers demonstrated significantly better prediction of Braak staging (p<0.01) and neuritic plaque score (p<0.01), but not ADNC (p = 0.13), when compared to models that included only cognitive and demographic information. Univariate models consistently underperformed in comparison to multivariate models.ConclusionPlasma biomarkers can provide valuable predictive information about the presence of AD pathology, which can facilitate earlier recognition and diagnosis of AD. Incorporation of demographic and cognitive variables along with plasma variables in multivariable models will improve prediction accuracy of AD pathology, specifically Braak staging and ADNC.

AbstractBackgroundIncreasing evidence suggests that transactive response DNA‐binding protein 43 (TDP‐43) pathology in Alzheimer’s Disease (AD), or AD‐TDP, can be diffuse or limbic‐predominant. Understanding whether diffuse AD‐TDP has genetic, clinical, and pathological features that differ from limbic AD‐TDP could have clinical and research implications. We aimed to better characterize the clinicopathologic features of diffuse AD‐TDP and differentiate it from limbic‐predominant AD‐TDP pathology.MethodThree hundred sixty‐three participants were prospectively recruited and followed in the Mayo Clinic Study of Aging, Alzheimer’s Disease Research Center, and Neurodegenerative Research Group and died between May 1999 and August 2021. All underwent genetic, clinical, neuropsychologic, and neuropathologic evaluations. At postmortem evaluation, they showed Alzheimer’s disease neuropathologic changes and TDP‐43‐immunoreactive inclusions. The distribution of TDP‐43 pathology was assessed using the Josephs 6‐stage scheme. Participants with TDP‐43 stages 1‐3 were classified as Limbic, while those with stages 4‐6 were classified as Diffuse. Neuropathologically defined frontotemporal lobar degeneration cases were excluded. Multivariable logistic regression was used to identify clinicopathological features that predicted diffuse TDP‐43 pathology at postmortem. We hypothesized that the Diffuse group would show impairment in cognitive domains beyond episodic memory and more hippocampal sclerosis.ResultsThe cohort was 61% female and old at onset (median:76 years [IQR:70‐82]) and death (median:88 years [IQR:82‐92]). Fifty‐four percent (n = 197) were classified as Limbic and 46% (n = 166) as Diffuse. A “clinical” multivariate logistic regression fit to 85 participants with all clinical variables of interest showed that later age at onset (OR:1.09, 95%CI:1.01‐1.19;P = 0.03), longer disease duration (OR:1.17, 95%CI:1.02‐1.36;P = 0.03) and higher Neuropsychiatric Inventory scores (OR:1.14, 95%CI:1.01‐1.28;P = 0.03) all independently increased the odds of being Diffuse, while higher Trails Making Test‐B (OR:0.99, 95%CI:0.98‐0.99;P = 0.02) and Block Design Test scores (OR:0.91, 95%CI:0.85‐0.99;P = 0.023) independently decreased the odds of being Diffuse. A “pathology” multivariate logistic regression fit to 233 cases with all neuropathological data revealed that widespread amyloid‐β pathology corresponding to Thal phases 3‐5 all significantly decreased the odds of diffuse pathology by 80‐90%, while hippocampal sclerosis increased it sixfold (OR:6.18, 95% CI:2.93‐13.03;P<0.001).ConclusionsDiffuse AD‐TDP shows clinicopathological features that differ from Limbic AD‐TDP, which could have public health implications relating to treatment and research on AD and dementia.

AbstractThe neuroimmune response in human aging and Alzheimer’s disease (AD) is complex and heterogeneous. Numerous genome wide associated studies have shown several AD genetic risk loci related to innate immunity and microglial function, of which APOE E4 and TREM2 R47H have been shown to influence microglial activation states. Microglial heterogeneity, characterized by morphology and function, has long been documented in human and murine literature. Using single‐nuclei RNA sequencing, one distinct reactive microglial subpopulation termed amyloid‐responsive microglia (ARM) was previously characterized in postmortem human AD brain tissue harboring various genetic risk variants, including APOE E4 and TREM2 R47H, and further validated in an expanded autopsy cohort using an immunohistochemical (IHC) marker for ARM. Herein described is an abbreviated study exploring the clinical utility of postmortem ARM IHC and its potential relationship to antemortem biomarkers, such as fluid and neuroimaging biomarkers, and cognition. Also discussed is the integration of digital pathology and artificial intelligence (AI)‐based algorithms to translate postmortem microglial findings into potential research application and model development.

Abstract Objectives Active RA has been associated with an increased risk of both cardiovascular and peripheral vascular disease. We aimed to compare cerebrovascular changes in patients with and without RA, both with and without a neuropathologic diagnosis of neurodegenerative disease. Methods Patients with RA (n = 32) who died and underwent autopsy between 1994 and 2021 were matched to non-RA controls (n = 32) on age, sex and level of neurodegenerative proteinopathy. Routine neuropathologic examination was performed at the time of autopsy. Cerebrovascular disease severity was evaluated using modified Kalaria and Strozyk scales. Clinical dementia diagnoses were manually collected from patients’ medical records. Results Prior to death, 15 (47%) RA patients and 14 (44%) controls were diagnosed with dementia; 9 patients in each group (60% and 64%, respectively) had Alzheimer’s disease. The prevalence of cerebral amyloid angiopathy, microinfarcts, infarcts or strokes was found to be similar between groups. Patients with RA were more likely to have more severe vascular changes in the basal ganglia by Kalaria scale (P = 0.04), but not in other brain areas. There were no significant differences in the presence of large infarcts, lacunar infarcts or leukoencephalopathy by Strozyk scale. Among patients with RA and no clinical diagnosis of dementia, the majority had mild–moderate cerebrovascular abnormalities, and a subset of patients had Alzheimer’s disease neuropathologic changes. Conclusion In this small series of autopsies, patients with and without RA had largely similar cerebrovascular pathology when controlling for neurodegenerative proteinopathies, although patients with RA exhibited more pronounced cerebrovascular disease in the basal ganglia.

Abstract Given the prevalence of dementia and the development of pathology-specific disease-modifying therapies, high-value biomarker strategies to inform medical decision-making are critical. In vivo tau-PET is an ideal target as a biomarker for Alzheimer’s disease diagnosis and treatment outcome measure. However, tau-PET is not currently widely accessible to patients compared to other neuroimaging methods. In this study, we present a convolutional neural network (CNN) model that imputes tau-PET images from more widely available cross-modality imaging inputs. Participants (n = 1192) with brain T1-weighted MRI (T1w), fluorodeoxyglucose (FDG)-PET, amyloid-PET and tau-PET were included. We found that a CNN model can impute tau-PET images with high accuracy, the highest being for the FDG-based model followed by amyloid-PET and T1w. In testing implications of artificial intelligence-imputed tau-PET, only the FDG-based model showed a significant improvement of performance in classifying tau positivity and diagnostic groups compared to the original input data, suggesting that application of the model could enhance the utility of the metabolic images. The interpretability experiment revealed that the FDG- and T1w-based models utilized the non-local input from physically remote regions of interest to estimate the tau-PET, but this was not the case for the Pittsburgh compound B-based model. This implies that the model can learn the distinct biological relationship between FDG-PET, T1w and tau-PET from the relationship between amyloid-PET and tau-PET. Our study suggests that extending neuroimaging’s use with artificial intelligence to predict protein specific pathologies has great potential to inform emerging care models.