INVESTIGADORES
ARMAS Pablo
congresos y reuniones científicas
Título:
Human disease-related G4-Vars ? genetic variations can affect G-quadruplex formation and function
Autor/es:
GISMONDI, M.; MARGARIT, E.; LORENZATTI, A.; PIGA, E.J.; BINOLFI, A.; CALCATERRA, N.B.; ARMAS, P.
Lugar:
Conferencia Virtual
Reunión:
Simposio; 5to Simposio Argentino de Jóvenes Investigadores en Bioinformática.; 2020
Institución organizadora:
Argentine Regional Student Group
Resumen:
Human variation databases contain a vast amount of genetic variations including single nucleotide polymorphisms (SNPs) that can show molecular effects on coding and non-coding regions. Although many variations were associated with different human diseases by DNA sequencing studies from patient populations (Genome Wide Association Studies or GWAS), the molecular mechanisms for most of them are unknown. G-quadruplexes (G4) are one of few secondary structures on nucleic acids that can be transiently folded when single-stranded G-rich regions are exposed, and were involved in different nucleic acids processes, including transcriptional gene expression regulation. The main goal of this work was to identify human diseases-related G4-Vars: naturally-ocurring variations associated with human diseases whose versions possess contrasting and drastic abilities to affect G4 structure formation. First, 417071 sequences flanking variations present in proximal promoter regions (PPRs, - 1000 bp upstream from transcription start sites) and their corresponding polymorphism information were downloaded from COSMIC, ClinVar, dbSNP and HGMD-PUBLIC human variation databases. We substituted the polymorphism position with reference or alternative versions. A G4 lineal pattern analysis showed that about 3% of the total sequences possess a G4 motif in one or more of their associated substituted sequences. Comparing reference and alternative versions for each sequence allowed us to get 2312 G4-Var sequences whose versions allowed or avoided G4 motif occurrence. The implementation of a novel G4 folding predictor (QUADRON) with a machine learning model based on in vitro G4 genomic data, supported 657 G4-Vars to have the best chances to differentially occur due to polymorphisms in genomic sequences. Following in silico predictions, disease-related G4-Vars on the PPRs of GRIN2B, F7, CSF2 and SIRT1 genes were experimentally validated by in vitro analyses and transcriptional regulation assays using reporter genes in cultured cells. Results suggest G4-Vars as a novel mechanism affecting differential transcription and point that they should be considered to understand the molecular effect of variations on the predisposition or establishment of human diseases and as potential targets for chemotherapeutic treatments.