Decellularized canine placenta proteomic analysis
This work aims to map the main protein groups of the native (DOI: 10.17632/t8m5gf5jm9.1) and decellularized extracellular matrix of fetal side of canine placentas from 35 days of gestation (end of the middle third of gestation), hypothesizing that the cell removal protocol guarantees a biological matrix capable of maintaining protein groups and structural factors necessary for tissue rebuilding, as well as the processes of cell adhesion and interaction and the restoration of vasculogenesis.
Steps to reproduce
All experimental procedures were under rules of Ethics Committee number 6611181016 (FMVZ-USP). Three canine placentas with 35 days were collected, decellularized and validated as Matias et al (doi.org/10.1159/000492466). Then the fragments were submitted to mass spectrometry, were sample homogenization, precipitation, reduction, alkylation, digestion and purification followed HEDRICKS et al. (doi.org/10.1038/nrm3902). Finally 3 µg of total protein diluted in 15 µL of ABC was loaded into the reading tubes and then read on the spectrometer Orbitrap Fusion Lumos (Thermo Scientific). Data were exported to the MaxQuant software (version v184.108.40.206) to identify the peptide spectra and identify the protein list using the canine protein database downloaded from the Uniprot database. The list generated by MaxQuant contained the peptide count, raw intensities, LFQ intensity, iBAQ intensity and spectrum count. This list was then exported to the Inferno software (version 1.1.6970) based on R software (version 3.6.2). First, data quality was analyzed by means of correction graphs and principal component analysis. Then, statistical analysis was performed, generating mean values between groups, “fold change” using ANOVA and T test. Focusing on the objective of this project, proteins that were primarily related to cell component ontologies related to the extracellular matrix or cell junction were filtered. The Proteome fold change quantification data of 35-day canine placenta were used for Enrichment analysis and functional classification for Gene ontology terms. There was used the "enrichGO" function from R package clusterProfiler, which implements a hypergeometric model to test for gene set overrepresentation relative to a background gene set with a Bonferroni correction and an adjusted p value of 0.05 and OrgDb = org.Hs.eg.db . The GeneSymbol identifiers were converted to Entrez identifiers with bitR and then used "enrichKEGG" function from R package clusterProfiler to compute enrichment of proteins in KEGG pathways using an FDR cutoff of 0.05. The significant pathway results and respective proteins were used to create the Pathway diagrams with the Pathviews package . STRING v11.0  was used to explore the biological interactions of proteins identified as differentially abundant between control and 35-days Canine Placenta in two reference databases the Homo sapiens and Canis lupus familiaris, using all active interaction sources with evidence and medium interaction confidence of 0.4. To gain a systems-level understanding of the patterns and interactions of 35-days Canine placenta tissue, one of the steps required is the construction and analysis of the network involving the most informative proteins. For this, we used the NetworkAnalyst  with the protein-protein interaction (PPI) database based on Uterus Tissue Specific data collected from DifferentialNet for PPI interactions across human tissues using filter=15.