Author(s)

  • Mélanie Moreau (Presenting Author) | CBMN UMR5248 | 146 Rue Léo Saignat, CARF, 33076, Bordeaux CEDEX, France
  • Jade Jaubert (Presenting Author) | Bordeaux Proteome | 146 Rue Léo Saignat, CARF, 33076, Bordeaux CEDEX, France
  • Mehdi Boubaddi | BRIC U1312/ CHU Bordeaux | 146 Rue Léo Saignat, BBS, 33076, Bordeaux CEDEX, France
  • Stephane Claverol | Bordeaux Proteome | 146 Rue Léo Saignat, CARF, 33076, Bordeaux CEDEX, France
  • Samuel Amintas | BRIC U1312/ CHU Bordeaux | 146 Rue Léo Saignat, BBS, 33076, Bordeaux CEDEX, France
  • Christophe Laurent | BRIC U1312/ CHU Bordeaux | CHU Bordeaux, 33000, Bordeaux, France
  • Caroline Tokarski | CBMN UMR5248 & Bordeaux Proteome | 146 Rue Léo Saignat, CARF, 33076, Bordeaux CEDEX, France
  • Sandrine Dabernat | BRIC U1312/ CHU Bordeaux | 146 Rue Léo Saignat, BBS, 33076, Bordeaux CEDEX, France
  • Nicolas Desbenoit | CBMN UMR5248 | 146 Rue Léo Saignat, CARF, 33076, Bordeaux CEDEX, France

Abstract

Human samples are precious, extracting maximum information from it is crucial. We propose multiomics and multimodal workflow combining Bottom-up LC-MS/MS, MALDI Mass Spectrometry Imaging (MALDI-MSI) and immunohistochemistry (IHC) to obtain morphological, lipidomic and proteomic data with reduced material requirements.

To develop our workflow, we used human tumoral (T) and non-tumoral (NT) pancreatic samples from resected pancreatic adenocarcinoma.
For LC-MS/MS approach, 48 samples of T and NT regions were microdissected from FFPE tissue. They were prepared and proteolysed using the PreOmics iST-FFPE kit, optimized for our samples. Peptides were analyzed using Fusion Lumos mass spectrometer. The adapted workflow enabled the identification of hundreds of proteins, demonstrating the feasibility of deep proteomic profiling from limited material.
For spatial approach, we used frozen sections of T or NT sample. IHC, lipid and peptide detections were performed on the same slide, via an AP-MALDI-Orbitrap. Data are treated using public databases for lipids and home-made database (with proteins identified by LC-MS/MS and in silico digestion) for peptides. Targeted analyses can be run focusing on well-defined areas like specific cells or structures.

Combining LC-MS/MS, MSI and IHC, we obtain a huge amount of data from a low quantity of material. This novel approach enhances the understanding of pancreatic cancer and facilitates the discovery of prognostic signatures and biomarkers.