Aim Formalin-fixed and paraffin-embedded (FFPE) tissues handling has long been a major issue in tissue proteomics. This requires dedicated procedures to access the identity of the largest panel of proteins. Some proteomics experiments need working with restricted tissues sizes, especially when proteomes of cancers of early stages are explored. We designed different procedures in the context of sample amount downscaling. Methods Several extraction and digestion approaches, adapted from previous tissue proteomics procedures, were tested on breast cancer FFPE tissue samples obtained by laser micro-dissection (LMD), bearing approximately 2500 cells. The preparations were followed by nanoLC MS/MS analyses using Waters nanoACQUITY UPLC and Thermo QExactive instrumentations. Data processing was performed using MaxQuant and Perseus. Results First, experiment sets suggested that raw LMD sample has to be kept during sample preparation. Proteolytic digestion occurs directly on tissues and proteins are not fully extracted. The number of steps has to be limited to the minimum to retrieve an adequate number of proteins. Trypsin solution for on-tissue proteins digestion also has to be set to high concentration to compensate the fact that digestion is rather performed on raw samples than on protein extracts. Finally, an adapted procedure of the Citric Acid Antigen Retrieval approach, combined with adequate 2D-UPLC separation, allowed us to retrieve more than 1400 proteins on these samples of limited size. A proof of concept was proposed for the validation of the method, consisting in comparing 5 triple negative breast cancer (TNBC) of no special type (NST), and 5 of invasive lobular carcinoma (ILC). Even though the high similitude of the two cancer types, 10 proteins were found statistically significant more or less abundant between the tissues. These developments were applied to the analysis of proteomes of early stages of cervix cancer, restricted to few thousand of cells. This gave us the first insight of early events occurring for the development of this affliction, supported by the identification of several involved biomarkers. Conclusions Our study highlights essential issues for FFPE tissue preparation, regarding sample loss. This will open new opportunities for biomarker discovery from tissues with restricted size. In the next future, this approach would serve as a supportive method for the identification of proteins detected by MALDI imaging.
A laser microdissection-based workflow for FFPE tissue microproteomics : from biomarker discovery to a supportive method for identification from MALDI Imaging analyses
Casadonte R;
2016-01-01
Abstract
Aim Formalin-fixed and paraffin-embedded (FFPE) tissues handling has long been a major issue in tissue proteomics. This requires dedicated procedures to access the identity of the largest panel of proteins. Some proteomics experiments need working with restricted tissues sizes, especially when proteomes of cancers of early stages are explored. We designed different procedures in the context of sample amount downscaling. Methods Several extraction and digestion approaches, adapted from previous tissue proteomics procedures, were tested on breast cancer FFPE tissue samples obtained by laser micro-dissection (LMD), bearing approximately 2500 cells. The preparations were followed by nanoLC MS/MS analyses using Waters nanoACQUITY UPLC and Thermo QExactive instrumentations. Data processing was performed using MaxQuant and Perseus. Results First, experiment sets suggested that raw LMD sample has to be kept during sample preparation. Proteolytic digestion occurs directly on tissues and proteins are not fully extracted. The number of steps has to be limited to the minimum to retrieve an adequate number of proteins. Trypsin solution for on-tissue proteins digestion also has to be set to high concentration to compensate the fact that digestion is rather performed on raw samples than on protein extracts. Finally, an adapted procedure of the Citric Acid Antigen Retrieval approach, combined with adequate 2D-UPLC separation, allowed us to retrieve more than 1400 proteins on these samples of limited size. A proof of concept was proposed for the validation of the method, consisting in comparing 5 triple negative breast cancer (TNBC) of no special type (NST), and 5 of invasive lobular carcinoma (ILC). Even though the high similitude of the two cancer types, 10 proteins were found statistically significant more or less abundant between the tissues. These developments were applied to the analysis of proteomes of early stages of cervix cancer, restricted to few thousand of cells. This gave us the first insight of early events occurring for the development of this affliction, supported by the identification of several involved biomarkers. Conclusions Our study highlights essential issues for FFPE tissue preparation, regarding sample loss. This will open new opportunities for biomarker discovery from tissues with restricted size. In the next future, this approach would serve as a supportive method for the identification of proteins detected by MALDI imaging.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
