Abstract Body

Early deficits in the acetylcholine network have long been known for Alzheimer’s Disease (AD), and the cholinergic nuclei are damaged long before the emergence of disease symptoms. However, the role of non-coding RNA (ncRNA) regulators of the cholinergic network in the early emergence of damage to the deep brain nuclei and in the progression of the disease process remains unknown. Here, we report expression changes of specific transfer RNA fragments (tRFs) in the nucleus accumbens from AD patient brains. Transfer RNA fragments (tRFs) are a recently re-discovered small non coding RNA class shown to operate like microRNAs to block the expression of mRNA targets whose levels increase in blood cells from ischemic stroke patients (Winek et al., 2020). Specifically, we analyzed short RNA-sequencing datasets from 348 nucleus accumbens tissues of AD patients and matched controls from the Religious Orders Study. Our analysis identified significant changes in the levels of several tRFs, whose levels increased in correlation with the cognitive decline of the donor patients, but was not corelated with age. Notably, the identified tRFs carry complementary sequence motifs to several AD-related coding transcripts including the inflammation-regulating lipoxygenase ALOX5, the ATP-biding transporter ABCC2 and the beta-amyloid regulator SOAT1, which is known to catalyze the formation of fatty acid cholesterol esters. Our findings add these tRF candidates to the growing list of non-coding RNA regulators of gene expression in the brain of AD patients and open new venues for pursuing their roles as therapeutic targets.

Aims

The aim of this study was to generate predictive models for AD using plasma cell-free RNA species at different stages of the disease.

Methods

We generated cfRNA-Sequence data from AD cases at CDR=1 (N=44) and controls (N=45) and applied standard quality control. Gene expression was quantified with Salmon and corrected by library complexity and log transformed prior to analysis. Genes known to be involved in AD and other neurodegenerative diseases (N=25) were used to create a predictive model using step-wise discriminant analysis in the CDR=1. APOE genotype was included in the model afterwards. The predictive power was tested in early (N=27) and pre-symptomatic (N=21) stages of the disease.

Results

Out of the 25 genes, eight were included in the predictive model after step-wise discriminant analysis. After inclusion of APOE genotype, the area under the ROC curve was 0.96, 0.99 and 0.82 for CDR=1, CDR=0.5 and pre-symptomatic stages respectively (Figure1).

Conclusions

Cell-free RNA is a promising minimally invasive biomarker for AD with an accuracy comparable to the one obtained using CSF biomarkers. This approach can provide a new screening tool for AD that can be used at population level and to evaluate disease-modifying therapies that target amyloid beta and tau.

Aims

A fundamental gap in the field is the incomplete identification of the mechanisms by which tau promotes cellular dysfunction. We and others established that pathological tau shifts the translatome yielding a maladaptive response. However, the molecular mechanisms remain unknown. The overall objective of this study is to establish these mechanisms by defining 1) the impact of pathological tau on translation, 2) the components of RNA translation machinery that interact with tau, 3) alterations to ribosomal transport, and 4) how tau shifts ribosomal specificity promoting a maladaptive translatome.

Methods

We used puromycin to detect changes in newly-synthesized (pumocyinylated) proteins in vitro and in vivo. We then coupled puromycin assays with proteomics or microarrays in the brains of rTg4510 tau transgenic mice to identify proteins altered by tauopathy. We validated results in Alzheimer’s brains using WB, RT-PCR, and E-CLIP. We measured changes in neuronal transport with novel RiboTag-IRES-tau constructs. We also defined the region of tau-ribosome interface region. Finally, we measured changes in translation dynamics with RiboSeq.

Results

We determined that tau expression differentially shifts the transcriptome and the proteome and that the synthesis of ribosomal proteins is reversibly dependent on tau levels. One protein altered by tauopathy is ribosomal protein S6. Ribosomal abnormalities were also identified in Alzheimer’s samples.

Conclusions

Our data establish tau as a driver of RNA translation selectivity. Moreover, considering that regulation of protein synthesis is critical to learning and memory, aberrant tau-RNA/RBPs interactions in disease could explain the linkage between virtually every tauopathy and cognitive impairment and memory decline.

Aims

Diabetic-retinopathy(DR), the most frequent complication of diabetes-mellitus(DM), is characterized by microvascular changes in the retina. The diabetic marks are linked to ageing and cellular senescence, but the characterization is still unclear. Several studies suggested the involvement of miRNAs in DR, for the regulatory role in glucose-homeostasis and immune-cells functions. The study aims to evaluate serum levels of miRNA in DM/DR; DM/noDR and in HC, in order to find reproducible biomarkers for DR, in relation to ophthalmological parameters.

Methods

Recruited patients underwent to Microperimetry; Angiography-OCTA and serum sampling for miRNA-analysis. Best-corrected visual acuity, intraocular pressure, dilated ophthalmoscopy and miRNAs (let-7a-5p; let-7b-5p; 23a-3p; 27a-3p; 15a-5p; 320b; 495-3p) evaluations were done.

Descriptive statistical analysis has been applied, using Pearson correlation and Kruskal-Wallis test. R statistical environment v3.6 was used for data analysis.

Results

BCVA and HbA1c showed differences among groups. A reduction in retinal sensitivity was found in Diabetic's groups, with respect to control group. OCTA analysis, showed a significant reduction in DM/DR and DM/noDR, for perfusion density at DCP and CC level. Serum miRNA levels, showed a reduction for 15a-5p and 495-3p miRNAs in DM-patients compared to HC (p=0.027, p=0.049, respectively). Moreover, a positive significant correlation was found for miR-320b and BCVA (r=0.894). Radar-plot showed how DM/noDR have in higher median values of miRNAs overlapping to the DM/DR-group.

Conclusions

In conclusion, we find evidence of damage progression biomarkers in diabetic retinopathy in diabetic’s patients. Serum-miRNA are being considered to have a strong potential as a novel biomarker for early detection of diabetes complications.

Aims

microRNA profiles in the brain of Alzheimer’s disease (AD) patients are altered with disease progression, suggesting that microRNAs may contribute to AD pathology. Of particular interest is microRNA-132 (miR-132), which is consistently and robustly downregulated in AD patient brain and has been shown to be involved in pivotal processes in the central nervous system that are also affected in AD, such as amyloid plaque deposition, TAU phosphorylation, inflammation, synaptic plasticity and memory formation. Restoring miR-132 levels could therefore be a potential strategy to combat or modulate AD pathology at the molecular and functional level. In order to fully assess the therapeutic potential of this microRNA, the miR-132-regulated targets and underlying pathways that play a role in disease progression need to be systematically characterized. Taken the cellular complexity of brain tissue and its impact on AD pathology, single-cell genome-wide approaches are required to dissect the cell-specific regulatory repertoire of miR-132.

.

Methods

Intracerebroventricular infusions of miR-132 synthetic mimics or antisense oligonucleotides were employed in order to elevate or knock down miR-132 levels in the mouse brain. Hippocampal tissue was subsequently processed and subjected to complementary unbiased proteomics analysis (MaxLFQ) and single-cell RNA sequencing (10X Genomics) to identify miR-132 targets.

Results

Following differential expression, pathway enrichment approaches and integrative analysis of the resulting datasets, previously validated and new potential miR-132 targets were identified and are currently further validated.

Conclusions

Overall, our data suggest that miR-132 may be a potent regulator of previously reported and novel molecular pathways with multilayered implications in AD pathology.

Aims

In Alzheimer’s disease (AD), pathological protein accumulation is a major hallmark represented by neurofibrillary tangles and amyloid plaques. This protein accumulation might be due to a dysregulated protein synthesis. The regulation of eukaryotic protein synthesis is primarily effected by eukaryotic initiation factors (eIFs) at the translation initiation step. So far, the role of eIFs in neurodegeneration and their potential contribution to neurodegenerative processes are not fully understood. Here, we aimed to reveal the expression patterns of eIF subunits in AD.

Methods

eIFs were analyzed on mRNA and protein level in human post mortem brain tissue including regions of frontal and temporal cortices and the hippocampus. AD subjects and controls without neurodegenerative pathology were investigated. For the neuropathological assessment, Braak and Braak staging was performed for all samples.

Results

The expression of eIF4G on mRNA and protein level was decreased in higher Braak and Braak stages. Increased eIF4G levels were present in hippocampal tissue of AD subjects. Beside eIF4G, altered expression was observed for other eIF subunits such as eIF3 subunits.

Conclusions

In human post mortem brain tissue, several eIF subunits were found to be altered in AD, whereas alterations are already present on gene expression level. It might be that translation initiation is reduced in AD subjects with advanced AD pathology represented by the Braak and Braak staging. Overall, a potential involvement of an affected translation initiation and a contribution of individual eIF subunits to AD pathology can be assumed.

Aims

Circular RNAs (circRNAs) are a class of RNAs highly expressed in the nervous system and enriched in synaptoneurosomes. A recent study suggest that circCDR1-as is downregulated in the frontal cortex of Alzheimer’s Disease (AD) patients. A study investigating whether other circRNAs are differentially expressed in the context of AD remains outstanding. Here, we conduct an analysis of circRNA expression to explore the relevance of circRNA expression in AD.

Methods

We generated RNA-seq data from 83 individuals with AD and 16 control individuals. We performed circRNA differential expression (DE) analysis on the basis of clinical dementia rating (CDR). We replicated the DE analyses using publically available superior temporal cortex RNA sequencing data from the Mount Sinai Brain Bank (173 AD cases and 63 controls) and performed a meta-analysis.

Results

On meta-analysis, we observed six circRNAs to be significantly DE on the basis of CDR. This included the previously reported circCDR1-as (p-value: 1.66×10^-6), as well as five novel ones: circHOMER1 (p-value: 3.30×10-9), circDOCK1 (p-value: 4.17×10^-9), circPEX5L (p-value: 6.12×10^-6), circKCNN2 (p-value: 6.12 × 10^-6), and circMLIP (p-value: 8.45 × 10^-6). When included in a model with AD risk factors (number of APOE-4 alleles, gender, age at death, and ethnicity), the circRNAs collectively contributed more than 34% of the variation in CDR in our parietal dataset. In contrast, number of APOE-4 alleles, the most common genetic risk factor, only explained 9% of the variation in CDR. Multiple circRNAs are differentially expressed in AD brain tissues.

Conclusions

Together these circRNAs contribute substantially to the variation in CDR.

Aims

AD has a substantial genetic, molecular and cellular heterogeneity associated with its etiology. We sought to investigate the glial and neuronal pathways affected by AD at a cell specific resolution. To do so, we generated single-nuclei RNA-seq (snRNA-seq) from the parietal cortex of Mendelian mutation carriers, sporadic AD and neuropath-free donors from the Knight-ADRC and Dominantly Inherited Alzheimer Network banks.

Methods

We generated snRNAseq (10X chemistry v3) for 18 APP and PSEN1 mutation carriers, 36 sporadic AD and 9 controls. 336,892 nuclei remained after QC for clustering and analyses (Figure 1). Our analytical approach is based on the identification of cellular states, their characterization and identification of genes associated with genetic strata.

Results

We identified a myriad of transcriptional states for the most representative brain cell-types (Figure 1) with distinguishing expression profiles (mean of 600 genes overexpressed; FDR<0.05). We identified that neuronal and glial cells have specific transcriptional states enriched in nuclei from brains with APP and PSEN1 mutations. For example, a microglia cell state specific for these brains shows overexpression of genes related to apoptosis, NOD like receptor signaling, Parkin-Ubiquitin Proteasomal system and protein processing in ER, while the astrocyte cell state specific for these carriers shows overexpression of genes related to the lysosomal and Brain-Derived Neurotrophic Factor signaling pathways.

Conclusions

We developed a unique molecular atlas to study the pathways dysregulated in AD. Our analyses indicate that in the backdrop of neuropath free and even sporadic AD brains, ADAD samples have distinctive cell states and altered pathways in neurons and glia.