Giacomina Rossi

Background: Mutations in progranulin gene ( GRN ) are a major cause of frontotemporal dementia (FTD). Most reported pathogenic mutations are nonsense, frameshift, or splicing mutations, resulting in a premature stop codon, degradation of mutated mRNA and consequent protein haploinsufficiency. Objectives: In this study, we analysed four subjects with FTD who had low plasma progranulin levels but no mutation detectable by sequencing of GRN , to disclose the underlying genetic cause of disease. Methods: Multiplex ligation-dependent probe amplification (MLPA) method was applied to search for rearrangements in GRN . Region-specific polymerase chain reaction (PCR) and Sanger sequencing were performed to define the breakpoint. Quantitative real-time PCR (qRT-PCR) on GRN mRNA and haplotype sharing analysis were also performed. Results: MLPA revealed in all the subjects the same heterozygous deletion, and a possible common ancestor was suggested by haplotype sharing. PCR and sequencing allowed us to define the size of the deletion (3028 bp), that removes part of GRN promoter, exon 1 including the transcription start site and most of the intron 1, and the breakpoints. qRT-PCR showed reduced level of mRNA, confirming the pathological nature of the deletion. Conclusions: In this study, we described a GRN heterozygous gross deletion which removes the consensus sequences for transcription factors and the transcription start site, leading to a reduced levels of plasma progranulin. Our study indicates that GRN rearrangements, although not common, should be investigated in patients with FTD who show low plasma progranulin levels but no GRN mutations detectable by DNA sequencing.

Traditional statistical approaches have advanced our understanding of the genetics of complex diseases, yet are limited to linear additive models. Here we applied machine learning (ML) to genome-wide data from 41,686 individuals in the largest European consortium on Alzheimer’s disease (AD) to investigate the effectiveness of various ML algorithms in replicating known findings, discovering novel loci, and predicting individuals at risk. We utilised Gradient Boosting Machines (GBMs), biological pathway-informed Neural Networks (NNs), and Model-based Multifactor Dimensionality Reduction (MB-MDR) models. ML approaches successfully captured all genome-wide significant genetic variants identified in the training set and 22% of associations from larger meta-analyses. They highlight 6 novel loci which replicate in an external dataset, including variants which map to ARHGAP25, LY6H, COG7, SOD1 and ZNF597. They further identify novel association in AP4E1, refining the genetic landscape of the known SPPL2A locus. Our results demonstrate that machine learning methods can achieve predictive performance comparable to classical approaches in genetic epidemiology and have the potential to uncover novel loci that remain undetected by traditional GWAS. These insights provide a complementary avenue for advancing the understanding of AD genetics.

A polygenic score (PGS) for Alzheimer’s disease (AD) was derived recently from data on genome-wide significant loci in European ancestry populations. We applied this PGS to populations in 17 European countries and observed a consistent association with the AD risk, age at onset and cerebrospinal fluid levels of AD biomarkers, independently of apolipoprotein E locus (APOE). This PGS was also associated with the AD risk in many other populations of diverse ancestries. A cross-ancestry polygenic risk score improved the association with the AD risk in most of the multiancestry populations tested when the APOE region was included. Finally, we found that the PGS/polygenic risk score captured AD-specific information because the association weakened as the diagnosis was broadened. In conclusion, a simple PGS captures the AD-specific genetic information that is common to populations of different ancestries, although studies of more diverse populations are still needed to better characterize the genetics of AD.

Among the more than 90 identified genetic risk loci for late-onset Alzheimer’s disease (AD) and related dementias, the apolipoprotein E gene ( APOE ) ɛ2/ɛ3/ɛ4 polymorphisms remains the longstanding benchmark for genetic disease risk with a consistently large effect across studies ^1–10 . Despite this massive signal, the exact mechanisms for how ɛ4 increases and for how ɛ2 decreases dementia risk is not well-understood. Importantly, recent trials of anti-amyloid therapies suggest less efficacy and higher risks of severe side effects in ε4 carriers ^11–13 , hampering the treatment of those with the highest unmet need. To improve our understanding of the genetic architecture of AD in the context of its main genetic driver, we performed genome-wide association studies (GWASs) stratified by ε4 and ε2 carrier status. Such insights may help to understand and overcome side effects, to impact clinical trial enrolment strategies, and to create the scientific basis for targeted mechanism-driven therapies in neurodegenerative diseases.

Familial cases of Alzheimer's disease (AD) with autosomal dominant transmission and early onset have a prevalence around 1%. Since only a small fraction of them has a monogenic inheritance due to APP, PSEN1 , and PSEN2 genes, genetic studies are ongoing to unravel the missing heritability. By sequencing panels including multiple dementia-related genes, we identified a novel likely pathogenic mutation in SORL1 in a pedigree including five members affected by AD. This loss of function mutation may lead to a reduction of SORL1 receptor, worsening amyloidogenic burden. As the contribution of SORL1 mutations to heritability of AD is presently not well established, we think that it is very important to signal new familial (likely) pathogenic SORL1 mutations in order to define the actual genetic involvement of SORL1 in AD pathogenesis.

Semantic and right temporal variant of frontotemporal dementia (svFTD and rtvFTD) are rare clinical phenotypes in which, in most cases, the underlying pathology is TDP-43 proteinopathy. They are usually sporadic disorders, but recent evidences suggest a higher frequency of genetic mutations for the right temporal versus the semantic variant. However, the genetic basis of these forms is not clear. In this study we performed a genetic screening of a single-center cohort of svFTD and rtvFTD patients, aiming at identifying the associated genetic variants. A panel of 73 dementia candidate genes has been analyzed by NGS target sequencing including both causal and risk/modifier genes in 23 patients (15 svFTD and 8 rtvFTD) and 73 healthy age-matched controls. We first performed a single variant analysis considering rare variants and then a gene-based aggregation analysis to evaluate the cumulative effects of multiple rare variants in a single gene. We found 12 variants in nearly 40% of patients (9/23), described as pathogenic or classified as VUS/likely pathogenic. The overall rate was higher in svFTD than in rtvFTD. Three mutations were located in MAPT gene and single mutations in the following genes: SQSTM1, VCP, PSEN1, TBK1, OPTN, CHCHD10, PRKN, DCTN1. Our study revealed the presence of variants in genes involved in pathways relevant for the pathology, especially autophagy and inflammation. We suggest that molecular analysis should be performed in all svFTD and rtvFTD patients, to better understand the genotype–phenotype correlation and the pathogenetic mechanisms that could drive the clinical phenotypes in FTD.

Frontotemporal Dementia (FTD) represents a highly heritable neurodegenerative disorder. Most of the heritability is caused by autosomal dominant mutations in the Microtubule-Associated Protein Tau (MAPT), Progranulin (GRN), and the pathologic exanucleotide expansion of C9ORF72 genes. At the pathological level, either the tau or the TAR DNA-binding protein (TDP-43) account for almost all cases of FTD. Pathogenic mechanisms are just arising, and the emerging role of non-coding RNAs (ncRNAs), such as microRNAs (miRNA) and long non-coding RNAs (lncRNAs), have become increasingly evident. Using specific arrays, an exploratory analysis testing the expression levels of 84 miRNAs and 84 lncRNAs has been performed in a population consisting of 24 genetic FTD patients (eight GRN, eight C9ORF72, and eight MAPT mutation carriers), eight sporadic FTD patients, and eight healthy controls. The results showed a generalized ncRNA downregulation in patients carrying GRN and C9ORF72 when compared with the controls, with statistically significant results for the following miRNAs: miR-155-5p (Fold Change FC: 0.45, p = 0.037 FDR = 0.52), miR-15a-5p (FC: 0.13, p = 0.027, FDR = 1), miR-222-3p (FC: 0.13, p = 0.027, FDR = 0.778), miR-140-3p (FC: 0.096, p = 0.034, FRD = 0.593), miR-106b-5p (FC: 0.13, p = 0.02, FDR = 0.584) and an upregulation solely for miR-124-3p (FC: 2.1, p = 0.01, FDR = 0.893). Conversely, MAPT mutation carriers showed a generalized robust upregulation in several ncRNAs, specifically for miR-222-3p (FC: 22.3, p = 7 × 10⁻⁶, FDR = 0.117), miR-15a-5p (FC: 30.2, p = 0.008, FDR = 0.145), miR-27a-3p (FC: 27.8, p = 6 × 10⁻⁶, FDR = 0.0005), miR-223-3p (FC: 18.9, p = 0.005, FDR = 0.117), and miR-16-5p (FC: 10.9, p = 5.26 × 10⁻⁵, FDR = 0.001). These results suggest a clear, distinctive pattern of dysregulation among ncRNAs and specific enrichment gene pathways between mutations associated with the TDP-43 and tau pathologies. Nevertheless, these preliminary results need to be confirmed in a larger independent cohort.

Frontotemporal lobar degeneration (FTLD) is a complex disease, characterized by progressive degeneration of frontal and temporal lobes. Mutations in progranulin (GRN) gene have been found in up to 50% of patients with familial FTLD. Abnormal deposits of post-translationally-modified TAR DNA-binding protein of 43 kDa (TDP-43) represent one of the main hallmarks of the brain pathology. To investigate in peripheral cells the presence of the different TDP-43 forms, especially the toxic 25 kDa fragments, we analyzed lymphoblastoid cell lines (LCLs) and the derived extracellular vesicles (EVs) from patients carrying a GRN mutation, together with wild-type (WT) healthy controls. After characterizing EV sizes and concentrations by nanoparticle tracking analysis, we investigated the levels of different forms of the TDP-43 protein in LCLs and respective EVs by Western blot. Our results showed a trend of concentration decreasing in EVs derived from GRN-mutated LCLs, although not reaching statistical significance. A general increase in p-TDP-43 levels in GRN-mutated LCLs and EVs was observed. In particular, the toxic 25 kDa fragments of p-TDP-43 were only present in GRN-mutated LCLs and were absent in the WT controls. Furthermore, these fragments appeared to be more concentrated in EVs than in LCLs, suggesting a relevant role of EVs in spreading pathological molecules between cells.