Abstract
Autism spectrum disorder (ASD) is an early-onset neurodevelopmental disorder characterized by impaired communication skills and social interaction, repetitive behaviour patterns and restricted and stereotyped interests. ASD is highly heritable, although the genetic factors involved are still largely unknown. We identified a rare deletion on chromosome 2, corresponding to a region including CAPG, which encodes for a post-synaptic protein, This evidence is important to understand the mechanisms underlying ASD considering the involvement of CAPG in the synaptic function, a pathway often dysregulated in ASD. The present project intends to 1) conduct a genetic, epigenetic and gene expression screening of 100 ASD subjects and 100 controls; 2)elucidate the role of CAPG for neuronal development and in ASD etiology thorough in vitro and in vivo studies; 3)investigate in Zebrafish and Drosophila ASD models impairment of neurons, synapses and locomotor behaviour by machine learning approaches.
Results achieved
Milestone 2 Sub-activity 2.2.1 Previous analyses revelead that an antibody raised against the human CapG protein is able to recognize a protein of similar weight in protein extract from Drosophila melanogaster suggesting the presence of a CapG homolog in Drosophila, which we have termed CapG-like. Using several bioinformatic tools, including BLASTP, we identified a strong sequence homology between human CapG and the sequence encoded by the fifth exon of the Drosophila Gelsolin gene. These analyses suggest that CapG-like may be produced as a short isoform of the bigger Gelsolin protein. Further investigations will be required to determine whether this isoform originates from alternative Transcription Start Sites, from an alternative splicing event or from post-translational proteolytic cleavage. The high degree of sequence similarity indicates that CapG is likely to be highly conserved during evolution. A search for the CapG-like short isoform across major protein databases revealed that it is not currently annotated. These observations suggest we could have identified the genomic locus responsible for the generation of CapG homologue of Drosophila. Sub-activity 2.2.2 Immunofluorescence analyses were performed on different regions of the Drosophila nervous system. In larval brains from wild-type individuals (Oregon R), we used antibodies against Elav and human CapG (hCapG). The results revealed complementary expression patterns: Elav labeled neuronal cell bodies in the optic lobes, whereas CapG-like (detected by the anti-hCapG polyclonal antibody Proteintech 10194-1-AP) was predominantly localized in axonal bundles and commissures within the ventral nerve cord (VNC). Importantly, the antibody used does not recognize Drosophila Gelsolin, from which the CapG-like isoform could derive, and Gelsolin itself displays a distinct expression pattern compared with CapG-like. Additional analyses were conducted to assess CapG-like expression at neuromuscular junctions (NMJs). Using double labeling with anti-HRP, a marker of the presynaptic motor neuron compartment, and anti-Disc Large, which labels the postsynaptic region of NMJs, we demonstrated the presence of CapG-like at the postsynaptic compartment. These findings suggest a degree of functional conservation of CapG-like. Indeed, in humans CapG expression is known to play a critical role in synaptic buttons formation and in the regulation of dendritic spine density. Sub-activity 2.2.3 Because the protein recognized by the hCapG antibody in both Western blot and immunofluorescence analysis could correspond to a short isoform of Drosophila Gelsolin, it was not possible within the timeframe of the project to generate a functional knock-out model specifically targeting CapG-like without simultaneously affecting the expression and function of the essential full-length Gelsolin protein. Despite this limitation, we experimentally demonstrated that CapG-like—consistent with its actin filament-capping function—is more highly expressed in adult flies older than 20 days. We therefore initiated the identification of this protein and its functional domains through mass spectrometry analysis. Thanks to the collaboration with the CEINGE research center in Naples, which will perform the analysis on the samples we provided (over 6 mg of total protein extract obtained from approximately 600 individuals 25 days old). We expect to better define the mechanisms underlying CapG-like generation and further confirm its evolutionary conservation. Activity 2.3 – Machine learning training to build prediction models (U2) Since it was not possible to generate a CapG-like knock-out model, we contributed to the training of the machine learning platform developed by U2 by providing climbing assay datasets obtained from Drosophila GBA-KO mutant flies. This Gaucher disease model was generated through gene editing of the GBA gene encoding acid glucocerebrosidase. Homozygous mutant animals (GBA-KO) display a well-characterized locomotor deficit that can be quantified using the climbing assay and progressively worsens with aging (20–25-day-old adults) (DOI: 10.1038/s42003-023-04813-2). We provided the U2 two different models, characterized by distinct genetic backgrounds characterized by a different body color (darker the wild type, pale yellowish the alternative background) and different degrees of motor impairment. This strategy increased the variability of the provided datasets and improved the training of the machine learning system. The large number of climbing assays supplied enabled the development of an algorithm capable of tracking, quantifying, and analyzing locomotor deficits at the level of individual animals.Project details
Unibo Team Leader: Roberto Bernardoni
Unibo involved Department/s:
Dipartimento di Farmacia e Biotecnologie
Coordinator:
Università degli Studi di CAGLIARI(Italy)
Total Unibo Contribution: Euro (EUR) 56.250,00
Project Duration in months: 24
Start Date:
30/11/2023
End Date:
28/02/2026
