Novel molecular tumour classification using MALDI-mass spectrometry imaging of tissue micro-array

• Published in: Analytical and bioanalytical chemistry 2010-05, Vol.397 (2), p.587-601
• Main Authors: Djidja, Marie-Claude, Claude, Emmanuelle, Snel, Marten F, Francese, Simona, Scriven, Peter, Carolan, Vikki, Clench, Malcolm R
• Format: Article
• Language: English
• Subjects: Adenocarcinoma - pathology; Amino Acid Sequence; Analytical Chemistry; Biochemistry; Biological and medical sciences; Characterization and Evaluation of Materials; Chemistry; Chemistry and Materials Science; Classification; Diverse techniques; Exact sciences and technology; Food Science; Fundamental and applied biological sciences. Psychology; Humans; Imaging; Laboratory Medicine; Mathematical analysis; Molecular and cellular biology; Molecular Sequence Data; Monitoring/Environmental Analysis; Original Paper; Pancreatic Neoplasms - pathology; Proteins - analysis; Spectrometric and optical methods; Spectrometry; Spectrometry, Mass, Matrix-Assisted Laser Desorption-Ionization - methods; Spectroscopy; Tissue Array Analysis - methods; Tumours
• ISSN: 1618-2642; 1618-2650
• DOI: 10.1007/s00216-010-3554-6

The development of tissue micro-array (TMA) technologies provides insights into high-throughput analysis of proteomics patterns from a large number of archived tumour samples. In the work reported here, matrix-assisted laser desorption/ionisation-ion mobility separation-mass spectrometry (MALDI-IMS-MS) profiling and imaging methodology has been used to visualise the distribution of several peptides and identify them directly from TMA sections after on-tissue tryptic digestion. A novel approach that combines MALDI-IMS-MSI and principal component analysis-discriminant analysis (PCA-DA) is described, which has the aim of generating tumour classification models based on protein profile patterns. The molecular classification models obtained by PCA-DA have been validated by applying the same statistical analysis to other tissue cores and patient samples. The ability to correlate proteomic information obtained from samples with known and/or unknown clinical outcome by statistical analysis is of great importance, since it may lead to a better understanding of tumour progression and aggressiveness and hence improve diagnosis, prognosis as well as therapeutic treatments. The selectivity, robustness and current limitations of the methodology are discussed.