3D genome organization during human decidualization

ANT, LUCIANA; LE DILY, FRANCOIS; PANIZZO, SILVANA; VALLEJO, GRISELDA; PISCIOTTANO, FRANCISCO; BEATO, MIGUEL; SARAGÜETA, PATRICIA

Congreso; XI Congreso Argentino de Bioinformática y Biología Computacional; 2021

Asociación Argentina de Bioinformática y Biología Computacional

3D genome organization during human decidualizationLuciana Ant 1 , Francois Le Dily 2 , Silvana Panizzo 1 , Griselda Vallejo 1 , Francisco Pisciottano 1 ,Miguel Beato 2 , Patricia Saragüeta 1 .1) Instituto de Biología y Medicina Experimental (IBYME-CONICET), Buenos Aires, Argentina.2) Centro de Regulación Genómica (CRG), Barcelona, Spain.Background:With the advent of High throughput Chromosome Conformation Capture (Hi-C) analysis it has becomepossible to query the organization of the entire genome simultaneously at the sequence level trough thequantification of its interactions (Lieberman-Aiden et al., 2009). The resulting contact frequency maps areable to display self-associating domains formed by short-range interactions among contiguous segments ofthe genome and can be visualized as ?triangles? present at the diagonal of Hi-C heatmaps. Classically, thesecontact domains are called compartments and Topologically Associating Domains (TADs).Compartments are defined by Principal Component Analysis (PCA), normally using Hi-C data binned at0.5-1.0 Mb resolution and, as a consequence, compartments are normally considered to be larger than 1 Mbin size. Compartments can contain sequences in an active (A) or silenced (B) transcriptional state and theyinteract with other compartments in the same state to give the plaid pattern observed in Hi-C heat-maps.On the other hand, neighbouring regions in separate compartmental domains interact less frequently andrepresent a compartmental switch or border. In this way, the compartmentalization of the genome createsboth local compartmental domains and distant compartmental interactions.In consequence, the chromatin organization can play an important role in transcriptional regulation, forexample creating a microenvironment where genes regulated by a transcription factor aggregate together tocoordinate their expression, or insulating an enhancer from its target gene to downregulate it.Results:In this work we explore the dynamics of the chromatin landscape and the transcriptional reprogramingduring decidualization of tHESC cells, a trans-differentiation process inherent of the human stromalendometrial cells. Decidualization was induced by exposure to cAMP, E2 and either progesterone or thesynthetic progestin R5020 and Hi-C and RNAseq assays were performed at 60 minutes, 3 and 6 days later.Implementing only a few basic packages we were able to create custom Python scripts to extract theinformation from the raw data of RNAseq and Hi-C that provides some insight into the mechanisms ofgenome 3D organization that takes place in this process.The PC1 values from the PCA of the Pearson correlation matrices were used to evaluate the globalcompartment reconfiguration, analysing correlation between different time points. The positive correlation(p-value<0.01) between B to A compartment changes and differential gene expression during humandecidualization was also verified. Furthermore, heatmaps were directly generated from the aligned Hi-Creads using costumed Python scripts, without the need of any software or a lot of computing power.Conclusions:The analysis of the differential interaction heat-maps lead to the discover a specific region in chromosome7, which comprises the Hox genes cluster, which showed dramatic changes in looping conformation, andalso presented a correlation with changes in decidual specific genes expression after 6 days of treatment,indicating the increased interaction frequency between this cluster and the neighbouring gene promoters.Submission track: Genomics, Transcriptomics, and Metagenomics