Breast cancer proteomics by laser capture microdissection, sample pooling, 54 cm immobilised pH gradient isoelectric focussing, and differential iodine radioisotope detection

The presence of progesterone receptor (PR) in estrogen receptor (ER)-positive breast cancer is associated with good prognosis and indicates that tumors are likely to respond to tamoxifen. However, ER+/PR- respond less well. In order to reveal the potential molecular mechanism of this phenomenon we sought to identify differential protein abundances between invasive ductal carcinoma cells from cryopreserved ER+/PR+ and ER+/PR- mammary tumor specimens. Because current proteomics methods are hampered in the examination of most primary human tumor samples by extreme tissue heterogeneity, laser capture microdissection (LCM) was used to isolate tumor cells and a sample pooling strategy was developed to analyse small sample protein lysates. Proteins from LCM-harvested tumors were pooled into 4 sub-pools from each condition of 3 tumors/sub-pool, and proteins from respective paired sub-pools were co-electrophoresed by 2-DE using 54 cm IEF over pH 4–9. Abundance ratios were accurately quantified by differential multiplex radioactive ProteoTope method at low attomole level (3.6μg protein per labeling reaction, <180 ng per multiplex protein sample per 54 cm gel). Applying this approach, differentially displayed proteins were identified by mass spectrometry using co-migrating nonradioactively labelled tumor proteins. They include decreased cytochrome b5 and transgelin, and more abundant CRABP-II, cyclophilin A, Neudesin, and hemoglobin in ER+/PR+ tumors versus ER+/PR- providing a possible explanation for differential susceptibility against tamoxifen as a result of deregulated cytochrome b5-dependent metabolism. This study demonstrates the potential of ProteoTope and LCM to enable extremely sensitive and precise differential analyses from well defined primary clinical specimen.